Skip to content

Posts from the ‘Musings’ Category


3D Printer Webcam

After doing some testing with webcams I’ve come to the conclusion that you need one wide-angle of a brute to get the job done without having to end up setting the webcam up 30+cm or more away from your print bed.

Awkward to say the least.

What I wanted was to put the camera on one of my delta’s towers and have it catch the whole print bed. Impossible.

I duplicated the setup on my bench so this is what it can see… The webcam, standard el cheapo USB version on the left edge, printed 50mm LED light housing sitting on a 220mm bed.

IMG 0966

Through the web cam, this is what I can see…

Webcam Normal

Ah yeah, that’s useful. More like a macro cam. I had to keep in mind that these are meant to sit on the top of your monitor or laptop and grab your happy smiling face for those online chats. Nothing more.

I couldn’t find a wide angle webcam but I did notice that a lot of eBay sellers had some “wide angle” lens attachments you could use on your cellphone to photograph wide angle scenes. Cheap enough, wasn’t worried about the quality, so I ordered one.

IMG 0969

The description:

Angle: 175°- 180°
Magnification: 0.33X
Lens Construction : 3 Element 3 Group
Max Diameter: 25mm
Length: 15mm
Wide Angle :
Angle: 130°
Magnification: 0.67X
Lens Construction : 2 Element 2 Group
Max Diameter: 20mm
Length: 11mm
Macro :
Magnification : 10X
Min Object distance : 10-15mm
Lens Construction : 1 Element 1 Group
Max Diameter: 20mm
Length: 15mm

The “wide angle” one is fairly useless but…the fisheye one hits the mark!

I merely held the one marked “Fisheye” in front of the webcam lens and got this:

Webcam Wideangle

Probably some edge distortion, but at least it covers a fair portion of the print bed compared to the standard webcam.

I plan on printing up a simple mount to fix it to the front the webcam.

While I’m at it, I should advise that if you get a web cam, make sure you get one that allows you to adjust the focal ring. I have a couple that are fixed focus and a pain to adjust, whereas most of the budget ones I have allow easy focus adjustment.


Soldering Gear


A word that can strike mind numbing fear into first timers. I kind of remember learning how to do it, from a ham radio operator no less. Frustration for me, laughter for him. When I was 10 years old.

Now admittedly most parents wouldn’t dream of letting their 10 year old grab hold of a hot soldering iron because, well, they’re kids. And you never know what could happen.

Fortunately I was born in an era that believed if you grabbed something hot and burned yourself, you instantly gained the experience to never do that again. My parents bought me a wood burning kit for Xmas when I was 10, and with it came the responsibility that I wasn’t going to “burn the house down”. To their delight, I’m sure, it was never an issue.

Soldering came to me because I was a model train nut in my pre-teens. Running wires and switches for tracks, sidings, lights and so on with nothing more than electrical tape was, really, more of a hazard than my wood burning set.

This blog won’t teach you how to solder. Neither will too many of the online videos or “instruction” pages I’ve seen. You want to learn to solder, find someone who can do it well, beg or bribe them to teach you. One on one. Makerspaces are a fabulous resource for this.

The dead giveaway for someone who lacks soldering skills? They use the gun or iron as a “paint brush” to “flow” the solder around the connection. You don’t wipe a solder joint with an iron to flow the solder. Unless you’re sculpting.

Guns, Irons, Stations, Torches

When you’ve got soldering under control, and don’t expect instant miracles, you may want to actually buy some soldering gear of your own.

If you’re under the impression that you can buy a 25W soldering iron and you’re good to go, my best advice would be save your money. Go see a movie. Why? Because the pleasant experience of the movie will last a heck of a lot longer than a 25W soldering iron experience will.

So let’s start with the cheapest thing that will melt solder..and the thing to avoid buying if you can…


First, Weller makes some truly great products. A pencil styled soldering iron isn’t one of them. Think of these more as a fabulous wood burning tool and you won’t go wrong. True, they will solder, in some cases, but as you get better with soldering, you’ll toss this thing. Main drawback? No temperature control adjustment.

Next up we have soldering guns. Like this one:

Weller 8200n

Runs 100 to 140 watts, easy tip to replace. You can use it for almost any gauge of wires from 14 to 26. You won’t use it to solder any components on a printed circuit board. The tip is too large. If you’re building a 3D printer, the one place you need this gun?

IMG 1218

That’s right. When you soldering the wires on to the bed heater. The bed heater is a big chunk of aluminum with a copper pattern on it. The metal dissipates the heat so fast from a soldering iron, only a gun will work. Plus the gun has to match the area. That’s a 220mm plate with a 100/140 watt gun. If you have a large metal plate, like 300mm, you need a 220/260 watt gun.

These old Weller guns last, well, at this point…here’s a snap of me at 15 when I was fixing my guitar amp (yeah on the kitchen table)…that is the SAME gun in the heated bed photo above. It is 50 years old and it still works just fine and the only repair was I put a new plug on the end of the power cord.

Mel Soldering

Plus in a pinch when I didn’t have a replacement tip (they do wear out), I’ve grabbed a piece of house wiring, 14/2 (NMD7), stripped the insulation off it, bent it in the shape of the original tip and it works, just fine until I can get a replacement tip. Which they still make…

As to why I am fixing my amp, I couldn’t afford to take it to a tech so I learned enough electronics that I could do my own repairs.

On the list next are soldering stations. To me, even a lousy one of these is better than a pencil iron.

Stations vary in quality. A lot. Typically a Hakko (genuine) is the cream of the crop, everything else is a wannabe. However, if you don’t do a lot of soldering, you don’t need a top of line station either. Again, match what you want to do with the gear to do it.

What you want in a station is first, replacement tips, with a chisel tip of 1.2mm or 1.4mm being the standard tip for component use. There are scads of tips to pick from if you have special requirements. Order a few extra when you buy the station. If they stop making tips for your iron, the iron is a conversation piece.

Second, you want temperature control. And not some analog gauge that looks like the black/green/red one from an old battery tester either. A digital read out one that has a knob to control not push buttons.

Below is an example of a relatively inexpensive station that will do a decent enough job and not put you in the poor house. Notice the place that holds the iron? It has both a sponge and a wire shavings area to keep the tip of the iron nice and clean. Oh, and don’t forget to wet the sponge before you use it. It’ll last a lot longer…:-)


Last on the list is a torch. And again it depends what you’re doing. Like plumbing. Torches are great! Working on a printed circuit? Not so great. Pay attention to what the pros use, get something along that line.


Of course, having the melting tool isn’t going to do anything unless we have something to melt. In this case, solder.

And all solder is created equally right? Nope. You need to match the solder to what you want to use it for. Unless it’s an emergency or some stop gap measure in which case, yeah, we’ll use whatever we can lay our hands on.

Lets start with the big stuff…as in plumbing…typically 60/40 or 50/50 (lead to tin mix). You’ll need some solder paste, clean joints and a spool of this stuff:

50 50

And a torch. This is big heavy stuff, takes a fair bit of heat to melt it.

Moving on, we get to electronic stuff. Printed circuit boards, wiring, finer detail stuff. And in a pinch, automotive as well.

Solder Type

Here things get a little complicated because there are two very different types of solder in use. One is the age old lead/tin and the other is lead free. There are pros and cons for both so pick a camp on your own choice. One thing to keep in mind is that if you’re repairing something that was made with lead free solder, you need to use the same.

Solder Diameter

Soldering diameter is also important. You want something with a .31 or .32mm diameter if you’re doing PCB work (or general smaller gauge wiring). Buying by the pound, Kester for example, is usually the best way to buy. But it’s not cheap and if you have to mail order it, keep in mind that it is a pound weight…

Quick Review

I recently got a tube of MG Chemicals solder, lead/tin, to try out. My normal brand is Kester 44 sn63/pb37 (tin/lead).

The first thing I asked myself is who in world decided to ship solder in a tube? So I thought about that for a minute and I think that they are saving on packaging. Normally solder comes on a plastic spool, inside a cardboard box. By shipping it in a tube, for a small quantity, they move away from that problem. Cost effective, don’t know about the plastic tube in landfill though…

Course the day I buy a small quantity of solder isn’t going to happen any time soon.. So the first thing I did:

IMG 2360

Yep, pull it out of the tube and coil it up. And I have to say it does make a nice sized coil. But it got me to thinking… MG sells solder wick, and I readily admit I have used wick about 3 times in fifty years), in a “spool”. I’d suggest the plastic tube would be better served with a simple spool like the wick. But I’m not a marketing guy so…just what I’d prefer.

IMG 2361

As I mentioned before, I’ve used Kester solder for a long, long, long, oh heck, since I started.

When using the MG, I found that it doesn’t flow as easily or nicely as the Kester does. And really, 63/37 solder is 63/37 solder. Thus the difference has to be in flux.

MG Chemicals says, “RA-like flux core” and “It generally exceeds J-STD-004 and J-STD-006 specifications

Kester says, “The high mobility and fast-spreading action of this flux results in more reliable production line soldering. 44 is classified as ROM1 per J-STD-004

Whatever all the heck that means. Me? I’m just an end user and I like the flux wetting action of the Kester by far.

While I really like the new MG Chemicals PLA, the solder? True is does do the job, but …meh.


Printer Vibration Iso – Part 2 et al

Before I get into the meat of the topic, I’ve noticed something rather interesting with the Arduino IDE and the Marlin firmware, at least on a delta printer.

During my tests with the firmware I’ve found the RepRap Smart Controller (an LCD2004) is very unresponsive when I was using the button on the controller. The whole printer “worked” fine, but, it was obvious that something had changed.

I initially thought that one of my “mods” in the firmware had affected it. However after going over my notes I couldn’t find anything that should affect the response of the rotary knob.

Purely by accident I was making yet another change to the firmware, when, much to my surprise, the rotary knob was back to fully responsiveness again. Say what?

I had inadvertently run the 1.6.5 version of the IDE and compiled and uploaded the firmware. Ah ha! I quickly ran the 1.16.13 version, re-flashed the firmware, ah yes, dodgy again.

Couple more tests and I found any Arduino IDE after version 1.6.8 caused the rotary to be flaky. I have no idea why, if it’s something in the IDE that has changed, if one of my libraries isn’t happy, or what, but it warranted a post-it note on the printer to remind me…

Printer Vibration – correction…

After my initial tests with my db meter a few days back, there’s been something gnawing at the back of my brain. Trying disparately to get out. Apparently. See, the 82db just didn’t ruddy sound, skip the pun, right. In my recording studio I’m well versed with sound check levels and something was wrong.

Tonight the penny dropped. I was metering the printer again with some new feet I made and I was sitting at 54db. What? How the? And then, as I said, the penny dropped. So for you audio geeks, here’s the scoop.

When I measured the original sound level previously, I was standing 1 meter in front of the printer. Cause sound measurement levels are supposed to be 1 metre away from the source. Tonight I was standing about 45 degrees off the front, but still one metre away. After scratching my head for a second or two…bingo. It hit me.

Or rather it didn’t “hit” me. What didn’t? The air stream from the fan. On my printers the fan blows directly towards the front of the printer. All db meters have very sensitive microphones and what I was measuring was the air pressure from the fan that blows 8CFM. If I put my hand in front to block the air flow, of course the sound level dropped.

But from the side, yeah, no air stream. Room was 50db without the printer doing any print job, and 54db with printing.

Live and learn. Again.

Feet Version 1

Here’s a recap of the original concept feet. I threaded the center of the vertical 2020 beam with a 5mm tap, screwed the mount to the corners, added the cute dollar store balls.

IMG 1987

Within a couple of days I found a couple reasons why this wasn’t a terribly bright idea. The first is, the damn balls roll. When I whack the stuck off the print bed, the whole printer wants to move because the print is stuck there. I also found the when printing with any speed, there’s a lot of movement going on. No, not so much it’s going to roll off, but enough that I didn’t get any warm fuzzies.

Lastly, there was the quality of the balls themselves. We’re talking dollar store stuff. I had visions of one of the balls splitting in half and the printer looking like the leaning tower of Pisa. Or worse, on my floor. In pieces.

This, of course, lead to a rethink.

Feet Version 2

It sort of started with, gee, I wonder what the balls are made of. Out came the box knife and I halved one. High density foam rubber. And not a lot of air holes in it either.

As I stared at the half, there was a spark of, ingenuity? Okay, so I didn’t want to throw out the pieces..I designed a holder for them.

IMG 2010

I used the same 5mm screw location from the last test set, added some double sided tape in the holder and mounted one on the corner.

Feet Mount

True it does stick out somewhat but the center is directly under the corner and the print is using a 50% infill.

Foot In Action

Now if the ball splits or whatever, the most the printer will lean is about 15 degrees. Plus with the design the printer doesn’t roll around or shift it’s weight on faster print jobs.

Just for the heck of it, I thought I would add three more cups for a total of six.

Iso V2

Safe to say that once you have the corners done, adding more doesn’t do much. Perhaps makes it a touch more stable but not by any margin that I could measure.

The other printer I have, I stayed with the corner design only.

Iso v1

Of course this printer has more weight pressing down on the corners than the one with six support points but really, I don’t see any difference in the amount of noise being transferred into the cabinet below.

All I hear now is stepper motor noise.

Isolation Enclosure

One of the disturbing things I see is 3D printer owners building all manner of cabinet enclosures to assist in printing the tricky ABS, or trying to keep the noise from the steppers at a level where they can get some sleep at night.

Why is this disturbing? Pretty simple. Electronic components are designed to work within a specific temperature range. The majority of commercial components are designed to work from -30 – +70C.

The ATMega2560 CPU is designed to work -55 to +85C. With a heated bed, in an enclosure, warmed up to 80C+, the not so earth shattering news is the component is not going to last as long as if it was in the free air.

On the printers that are designed to be enclosed, there’s no doubt a cooling fan for the electronics and a vent out of the enclosure. However, these are brand name printers, not the knock off junk one finds on fleaBay, BangGone, or AliExpleatedDeleted…

Secondly, the RAMPS board has some MOSFET’s on it and while those handle heat quite well (depending on the MOSFET on the board), adding some 90-110C heat to it, you know, just to keep it warm, isn’t going bode well over time either. Plus if you have under gauge wire for the bed, a bad connection to the terminal block on the RAMPS board, yeah, not going to bode well. At all.

Food for thought any way.


Delta 3D Printers – Mirth, Myths and Madness

3D printing is not new, it’s been around in some form or other since 1989 or so. What is new is the introduction of 3D printing for the hobbyist, we poor unsuspecting souls. This has been gaining momentum since 2011.

A quick search of the net will result in the fact that there are over 1,000 different 3D printers available today and a small percentage of those will be delta style printers. That small percentage is the basis for this blog entry. Dated, today. Stardate whatever…to go where no man has gone…oh wait, we have gone there.

Delta’s owe their heritage to Johann the person who provided the jumping off point, not to be taken literally, for the delta’s we have today. His first prototype was the Rostock, followed by the more refined Kossel.

While it seems like age ago now, a mere short 18 months ago what I knew about 3D printing could be written on your fingernail, in large type and the nail would still be plainly visible. I suffered from D.A.A.P at the time. What’s DAAP? Dumb As A Post. To put it into perspective.

When I started I understood there was some hardware and software to make the thing work. Software I assumed would be fairly easy because I’ve been writing software since 1979. That was a bit of a mistake looking back, but at least I had the background to figure things out.

The hardware, that looked amazingly easy, turned out to be not quite as straight forward as I assumed. Thus I decided I’d better have a good backup plan before I “built” any type of printer.

As it happened, David, one of the members of Fraser Valley Makerspace had cobbled together some extrusion, electronic bits and mechanical pieces, and made them available through the Makerspace as a Kossel kit. He had a working one, so…I decided it must work and bought one.

The Madness

There are seemingly two kids of people who want 3D printers.

The first are those who want to use it like a toaster. Take it home, plug it in, briefly read enough of whatever documentation there is to make toast and use it. Of course there’ll be some burnt bread, but there’s a better chance of it working at the on set, the learning curve will be mainly filaments and software.

My first piece of advice, if you’re this first type, buy a Cartesian printer. Brand name. Make friends with other people who have them, even better if they are local to you. Check out any local maker spaces, they make great resources. Something eventually will not make sense and/or break, you need a backup plan to keep the toaster toasting.

If you must build a kit, find someone who has built one that can help you build it. As a friend of mine put it, two heads are better than none.

The second types are the tinkerers. My ilk. Folks who love to build stuff, take it apart, tweak it, make it work better, change it, redesign it, use it and educate themselves to no end on it. Whatever it might be. There is as much joy in building as there is using for me and my ilk.

My logic at the time was that if I got stuck at some point in the build, I needed a parachute to help me over the rough areas and that would be where the Makerspace came in.

I had zero clue at the time I’d be dangling off that rip cord for more than a month with the ground rushing up at Mach 3.

I owe my sanity in my initial build to one of the FVM members, Rob. While he’d hadn’t built one of the kits himself, he did own a Ditto 3D printer and had single handedly got the mini-Kossel working at the space. Thus when I got stuck or didn’t get something, Rob to the rescue. A lot. He become, unbeknownst to him, Rob Rip Cord.

Even with a well equipped workbench, micrometers, scopes, power supplies but I have to say that I found I needed a metric tap and die set, metric drill bits, metric hex wrenches, metric hex screw drivers. I did a lot of shopping on fleaBay.

Plus my kit didn’t include any preformed wiring harness, no Dupont ends already in place, no power supply, or power supply wiring, thus for a fair chunk, I was on my own. However, having built circuits since I was 13, it didn’t really present much of a challenge in many respects.

There was a manual that I downloaded. It was supposed to be a guide on how to build the Kossel. Pretty evident from the start that the downloaded manual was for a different printer. Everything from the nuts and bolts to the extrusions were different. Sort of like building a Chevy with a Honda manual.

While I thought this lack of instructions must be unique, in hind sight I have now found scads of printers in kit form that clearly state, “If you don’t know how to assemble one of these, don’t buy this because we offer no manual or limited technical support”. Of course the price of the kit reflects this but still…not what a first time builder wants to start with. Me? I suffered DAAP, I was undaunted.

I finally reached the frustration point where I was either going to toss it out, or find the humour in it. Fortunately I found the humour.

IMG 0119

And while I’d like to think that I know what I am doing when it comes to building things, yeah, not so much for a 3D printer. Mostly due to the fact I didn’t know how the thing worked in the first place. DAAP.

I did everything backwards, wrong, or just dumbly. Heavy on the dumb. Hey I even lost some of the ball bearings out of the linear rails because I didn’t know they just slid off the rail…oh yeah. A comedy of errors abounded.

Yet, I still remember the first print…and no, it’s not supposed to be a lace doily cube. It’s what you get when you don’t have the foggiest notion of what you’re doing but you do it any way.

At this point, I wasn’t believing in miracles, I was relying on them.

IMG 0188

So I churned out a few of these stringy things, and they might have made good scouring pads for pots and pans, but something was amiss and I didn’t know what it was.

The Myths

And of course like every desperate individual I turned to the internet for help.

I was still suffering from DAAP I should add.

Using the net without knowing much is like using Dr. Google as your personal physician. There are thousands of DAAP sufferers like myself whose opinion and help is about as useful as a milk bucket under a bull.

Thus it was finally through the Makerspace gurus that I learned something new. A thing called “bed levelling”. My prints didn’t adhere to anything because, well, my basic point of reference didn’t exist. At least in this world.

My Kossel kit had an auto level probe, but for the Marlin firmware version I had, it seemed like every time I levelled, just before the print started the firmware tossed all the bed info it just collected and proceeded to print yet another scouring pad. Handy. Not. Auto level working? Not.

After learning many tweaked the firmware to adjust for the bed level I decided against that avenue. My logic was that if anything happened to the firmware I’d be doing it again, i.e. bugs in the levelling routines or whatever, well, yeah.

I decided to do bed level via mechanical style.

When I inquired how anyone levelled their beds manually, I was greeted with tales of woe, daring do, silicon snake oil and just plain dumb luck. I had copious amounts of dumb, just seemed to lack the luck.

The tragedy was that you loosened the top end stops with the micro switches on them and moved them up or down less than a cats whisker and hoped you got it right. For all three towers. With a flat blade screw driver. 1/8″ thick. Like that old saw, “Measure it with a micrometer, mark a line with chalk, cut it with an axe.”

I looked at the problem, pulled out a drill bit and did my first mod, on the 7th day. The ‘x’ marked the spot, put in an M3 x 6 bolt, the head triggered the end stop switch.

IMG 0199

What used to take days of fiddling manually to get right, took all of 30 minutes, 3 minutes to set the scripts in Pronterface. The only firmware tweak required was if the center of the bed was concave or convex. You had to change one of the rod lengths and re-level. Again, do it a couple of times and you can do it in your sleep.

For the X/Y dimension size, print a cube, measure it, calculate how much too big or too small it is, change a number in the firmware, level the bed if needed, reprint, check. Having built a number of Kossels for others, worst case from power on was out .21mm, best case build was out .06mm. You need that kind of accuracy to print a.. ah… Pokemon. Or something.

Eventually the FVM kit would be upgraded to include the screw carriage adjusters and better belt grips. Huge progress.

However the myth still exists today that delta printers are next to impossible to level, or will take days to do. So you must have auto level.

The real truth is, if you understand (or take the time to understand) how to do it, you’ll find you can manually level the bed, exactly, once. And unless you change the hot end, diagonal rods, effector or something that will change a dimension in the build area, you don’t level it again. Period.

I printed over 900 hours on my first Kossel before I changed the nozzle and did a re-level. In 15 minutes. Big whoop.

The fact is all 3D printers come with their own little bag of issues. Be it the slicer, firmware or how you communicate with the printer. Pros and cons. Plan on doing a lot of research and look for actual A/B print examples not opinions in a forum.

Lastly it helps, especially if you build a kit, to be an information sponge. Obtaining little bits of info from others with the SAME kind of printer as you and then TESTING that bit of info to see how it fairs on your printer will enable you to produce better prints.

Even after logging over 2,000 hrs on my first printer, I am still learning and testing what I can do to turn out better quality work.

The Mirth

You’d think that just having a printer that works would be joy enough. Truly it is very satisfying, but I’m going to step back a moment here and think out loud as to WHY my Kossel kit was more than I’d even hoped for.


For the two FVM kits I built for myself, my thanks goes to the FVM’er, David who, either by shear good luck or design managed to cherry pick most of the best parts for a Kossel build.

The one shallow item in the kit was the full graphic display (12864). As a software programmer I know how CPU intensive just updating these displays can be. I think I had mine on the Kossel for about 3 days before I tossed it in the dumpster (along side the auto level probe setup) and replaced it with a Smart LCD 2004 controller.

The rest of the kit, stellar. Just stellar.

The first is the linear rails. These are premium items that will give hundreds if not thousands of printing hours. Nowadays you see what I call the cheapskate bearings like the Rostock used for the carriages. They are big, burly, have mass, and will not wear well over the long term.

Second is the carriage pieces that I had a hand in modifying. The current ones hold the belts so much better and have the adjustment screws.

Third is the massive stepper motors in the kit. These are 1.68A steppers, probably overkill to the nth degree compared to the whimpy ones I see now, but they run great and cool.

The 2020 open beam extrusions do a good job of keeping the frame rigid, even when when printing up to the 210mm height on mine.

Then the stepper drivers. Out went the old A4988’s and in came the DRV8825’s. The 8825’s can deliver more current and/or finer stepping if you need it. Nice to have that additional overhead.

Heat bed. Many printers still don’t offer or come with a heated bed. My Kossels have a 220mm heated bed, large enough to print everything I’ve needed so far.

Top drawer extruder. Three redesigns later, the current one is the best design I’ve seen and I’d select it over any of the metal ones I’ve seen.

So…thanks FVM, David and Rob, from Rocky and Bullwinkle (and me of course)…

Twins Printing


Black is BAACCKKK…MG Chemicals

In my ever expanding quest to try out new filaments, I’ve always found that if you want to see how any filament performs, at the worst, just try out a roll of anything they make. In black. Seriously.

I’ve read that black filament is the toughest to print because of the dyes go into the production. Or magic. Yeah, one of those for sure.

Regardless of which of my delta printers I use, black has always been a bit of a duck shoot when it comes to getting good quality. Any other colour, no issues. Thus, must be magic. Yep.

As I sat staring at the two brand new unopened black spools of filament on my workbench I was undecided which I should try first.

The first was PLA, the new “improved” brand from MG Chemicals.

IMG 1911

The second, PETG. Truthfully PETG is my compromise to avoid printing with ABS (which gives me royal headaches) but it is so darn stringy and unpredictable that I haven’t printed with it much. Some would call that being a smart coward.

IMG 1912

Of course, I used common sense and did one at a time. And pigs can fly. Forget the common sense thing… I loaded up one of my deltas with PLA and the other with PETG. On the PLA I had to print out some tank treads for a robot I’m building so, 15 treads at a time. In black.

The treads are from Thingiverse Caterpillar Tracks. Although labeled as a work in progress, I’m doing it any way. Although I’d really like to build a fairly large Wall-E…maybe in time.

IMG 1903

My settings, 0.3 layer height, 50C bed covered in Elmers glue stick, 208C for the first layer, 200C after that. I’ve made 60 treads so far and they have come out absolutely perfect. No strings, no separation, just way better than I expected.

On the other printer, I decided to make something to hold my SD cards and flash drives. With PETG no less. Again, merely a design I found on Thingiverse called Moby_Stick

IMG 1910

My settings, 0.3 layer height, 50C bed covered in Elmers glue stick, 224C for the first layer, 222C after that. Used Slic3r’s built in support for the whale tail. Not the default supports but “rectilinear grid”, which I have found out is far better than the default support patterns.

On the whale the supports literally fell off when I chipped it off the heated bed. You got that right, I had to use a metal spatula to get it off the bed.

I’ve never had any issues with adhesion with Elmers and a 50C heated bed. I’ve read many a horror story from others about getting PLA or PETG to stick to the bed, heated or otherwise. With every solution from 70-100C heated bed to hair spray on painters tape, to Kapton tape. Me? Borosilicate glass, heated bed, 50C, Elmer’s stick.

Perhaps with a delta because the bed doesn’t move the glue stick approach works fine. But I’ve done the same with the standard Cartesian printer I had here and it stuck just as well. Thus the problem must lie elsewhere.

After the prints, I measured the diameter of the filaments. In hindsight it should have been done first but, well, in my excitement I kind of missed doing it. None the less, both the PLA and PETG measured out 1.75 on my micrometer.

Last year I used a number of different MG Chemical PLA filaments, which I will call “old stock” now. While colours like their gold or white were superb, some of the others, not so much. But this year, things are not the same old.

MG Chemicals has changed their products on Amazon. The new stock has cardboard sides and is labeled “Improved”. They aren’t kidding. I’ve noticed a few complaints about the size of the center hub not fitting some spool holders. Since I use spool rollers, spool center is a non-issue.

No idea what kitchen they are using to create this new improved brand of filaments they are selling, but so far in my tests, I’ve found that I don’t need my go to brands any more. I can grab a roll of MG, thread it up, and I know exactly what it is going to come out like.

When I add in the fact that MG is less expensive than what I usually buy, well, that’s just icing on my cake, sprinkles on my donut, gravy on my fries, well, you get the idea….:-)


Put a sock on it!

E3D would like you to put a sock on your hot end: E3D Sock No, really. I’m not kidding.

Silicone Sock Fillets

Seems there was this REDDIT thread where someone did this a while back and it’s been the single most requested item since that point. People have be clamouring for it. Not sure exactly what people but by reading the rest of the announcement you’d have a better understanding.

The socks come in bundles of three. Silicone, even high temp, custom poured, massaged by virgin ants, will eventually succumb to the high temp of the hot end (call it the Priestess end) and be sacrificed. Poof. Yep, you need more than one.

E3D is even including these socks with every new hot end they sell. You know, so you can keep your filament in your…I mean in check…safe PLA’ing and all that.

And really what 3D printer enthusiast doesn’t want a clean nozzle? Take a look at the photo of what your hot end looks like before and after being protected with a sock on it:

V6 with and without silicone

That’s right kids, if your hot end looks like the one on the left, you might want to have it checked. By a professional.

And SHARPER prints! The laws of heat transfer tell us that if your hot end is moving so slow above the printed object that it’s melting and distorting the layers as it goes, you’ve got more than a hot end problem. You got a SPEED problem. Big time and this little blue bandaid isn’t going to lower much of anything. Except your bank balance.

E3D even clearly shows how you install your sock to prevent dribble and just the tip of your hot end is visible.

Sock placement

Now perhaps Cartesian printers suffer from this hot end malady, but on my delta, with 1887 hrs on it, I barely have the beginnings of the disease… Whew, lucky huh?

IMG 1713

I do use tweezers when my hot end is getting up to temp and if there is any filament drool I whisk it away. Possibly that’s the reason mine doesn’t look like the one at E3D. Or maybe E3D was just going overboard for the shock factor.

Fascinating to see a market sort of created where there never was a market before. You know, because. Of. REDDIT. Where all the “experts” hang out (“ex” – the unknown quantity, “pert” – a drip under pressure). Like going to “Dr. Google” for medical advice.

Maybe I’m just old fashioned or not in the know, but apart from my tongue planted firmly in my cheek as I was penning this, I have no idea why a silicone sock for a hot end even exists. Maybe it’s the solution for those without some common sense. Like E3D. Who. Thinks. This. Has. A. Market.

[Addendum] It was pointed out that if you have fans blowing around the hot end, the sock can act as an insulator to maintain better control over the temperature of your nozzle. As opposed to wrapping your nozzle in insulation and Kapton tape (which is not nearly as inviting to lusty filaments). So perhaps in this case, a sock does make some modicum of sense. And I’ll leave it at that.


I built a Kossel

Yep. That’s right. I built a Kossel. For anyone that knows me, this news will not come as any surprise since they know I built one in June 2015. It took me three weeks. With many hours every day. A number of good stiff drinks. Long quiet walks on the beach.

Although advertised as a “kit”, in all honesty, it was a box of parts with a link to an online manual that didn’t reflect the parts that came in the box. Which made the assembly and calibration one of the most frustrating things I’d ever experienced. Apart from wanting to heave the whole thing in the garbage bin (not sure which parts of it would be adequate for the recycling bin), I started to find the humour in building it.

IMG 0119

As I remember the build, two things were obvious. First, this wasn’t a kit by any stretch of the imagination. Second, although I’d label myself a “maker”, what I knew about 3D printers on a scale of 0 to 10, was running at about -5 (that’s right MINUS 5). Mechanical I know some of, electronics I know more of.

My stubborn and “want to know what makes it tick” attitude determined that I would indeed learn the thing inside, and out. Thus I didn’t give up. I got it working the absolute best that it can work. I’ve even bling’d it out with my own add ons.

So fast forward to today. Ten months down the road. Over 1100 hours of printing time. And the printer, as of this minute, it printing something else for me. That’s right, in spite of it all, it’s working like a champ. Who’da thunk.

Back to the topic, I built a Kossel. Actually, I’ve built TWO of them now.

I printed out a complete set of parts for the FVMakers “kit” in MG Chemicals GOLD PLA (MG has some VERY nice deep and vibrant colours in PLA!). Rather than just build the thing, I decided to do a photo build of it. Not just “this part goes here” but building it the way a real kit would be built. I.e. you build in subsections and then assemble the subsections.

The fully assembled and calibrated Midas is for sale from FVMakers (if you’re interested ask them about it by name) so I didn’t built it for myself. I built it for someone get as many hours of enjoyment from as I get from my own.

In two evenings the complete mechanical sections were assembled. Two more nights for wiring and some bling and it was done. Bling? Yeah. Kossel with power supplies hanging off somewhere or graphic displays laying on the desk are what I call sloppy. So, make it neat, make it compact, bling it on.

With the gold PLA, I decided the easiest way to differentiate it from my other Slic3r/Firmware settings was to “name” it, therefore it became “Midas”. It printed it’s own name as a matter of fact.

Kossel Midas

A HUGE advantage of building a second one is that you know where the weak areas are. And you pay attention to them to make damn sure you get it right the first time. The first stumbling block is getting the basic PLA parts printed accurately to start with. I printed these on my own printer and although I admit it’s not 100% accurate, it’s within .07mm dimensionally.

As I was building Midas I never looked at a build manual. I just put together the way it made sense to me that it should be assembled. As in logical. A much more enjoyable build. Besides I’ve had mine apart and together so many times that was just easier knowing what part had to be installed before some other part.

Secondly because of my own and others input, this kit has been upgraded in a number of key areas. Let me point out that those areas are extremely critical. Take the bottom frame corners. Taller and more robust. Take the extruder, extremely efficient compared to the original one. The carriage holders for the Traxxas rods, fabulous. Plus more.

In the process of building the Midas I paid attention to the areas where any inaccuracy is going to come back and bite you in the butt later on. Like the Traxxas rods. I have a custom jig that I made that does nothing more than make sure the Traxxas rods are equal in length to each other. Each rod is a perfect 301.92mm.

Why is this important? Those rods hold the effector, that’s the plate the hot end hangs from like a bunch of grapes and ensures that one rod isn’t stressing another rod. You want that effector plate to glide around like the marker on an Ouija board.

The hot end should fit snugly in the effector plate and you shouldn’t be able to rotate the hot end once it’s clamped in.

The bed rests, where the aluminum plate and Borosilicate glass sit, should be sitting flat on the 2020 beam. Peak underneath to make sure that things are sitting level on those pads.

Getting away from assembly for a second, let me talk about tools. Yeah. If all you own is a flat bladed screw driver, forget building one of these. Or plan on a number of evenings at a maker space where they do have the tools.

And what will you need? Metric hex wrenches, normal and ball end (2 & 3mm), long ones too! Metric tap for some of the fittings, metric drill bits (1.5mm through to 5mm), DVM (digital voltmeter), alligator lead, jeweller screw driver, soldering iron, 63/37 leaded solder (cause it’s better than unleaded in my opinion), crimper for the Dupont connectors, wire strippers/cutters, extra wire (16 gauge) and good lighting so you can see what you’re doing.

Interestingly enough, in spite of the fact that Canada went metric in the mid-1970’s, you’ll find it next to impossible to buy metric drill bit sets. Instead they sell the non-metric sets. You can buy taps and dies in metric. If you need those things, eBay or a maker space are your best options.

Back into the assembly portion, one of the most common displays that comes with the 3D kits is termed the “Full Graphic Display”. Essentially a 12864 graphic monotone LCD with a daughter board to connect to the Arduino MEGA 2560. They refer to this as a “smart controller”, all evidence to the contrary.

With this graphic display, you can’t just write at some point on the display and update that specific area. Nope. You have to refresh the whole display each time. So you’re writing a lot of pixels out. Of course it doesn’t look like that’s what you’re doing because in the background that’s what the u8glib Arduino library is doing for you. And it only goes so fast.

Herein lies the underpinnings of the rub. The computer is sending GCODE, the Arduino is figuring out where in space it has to go and extrude something, then it has to update the display so you know it did it. Now to be honest I have no idea why it updates the X and Y positions. They don’t mean squat to me, the Z is the only one I’m interested in.

To continue, we need to spin the graphic fan icon of course. Which is pretty much useless because there’s no feedback loop from the fan to the controller to even tell you have the fan connected. Hence, “smart controller”, yeah, not so much.

When the printer gets to a tight area, specifically a bunch of tight curves, it stalls. There’s just so much that the little Arduino MEGA can do, so, rather than just skip the display, it stutters the hot end for you. Yeah. Handy as heck.

This is why the rage a while back was to add a SmoothieBoard or some heavy weight processing power to keep it munching through the data. Which of course, introduced a number of its own problems. No such thing as a free lunch.

To remedy this problem there are three avenues open.

One, toss the full graphic “smart” controller and put in a non-graphic controller like the 2004 LCD.

800px Smart Adapter 9

The second, although not quite as foolproof, print from an SD card. Even then it can stall slightly on tight curves.

The third is to modify the firmware. Look for the XYZJerk values, normally set at 20 or so and drop them down to 5 or slightly less. If you drop them to 1, you WILL solve the problem, you will also entering a waiting game for a print or the completion of the next ice age. Yeah, it slows it down that much. The whole printer is spasticated.

In the case of both the 2004LCD and the 12864 “full graphics smart controller”, they are current pigs. if you’ve ever looked at one of these displays and notice it flickering, there’s a reason for it. And it’s like this…

To power the Arduino, that 360 watt 12V power supply feeds 12VDC to the Arduino Mega. On the MEGA is a 5V regulator that accepts the 12V coming in and drops it down to 5V so the MEGA and any attached devices can use it. The MEGA uses a modest bit of current, but when you tack on the display it goes up. The problem with getting the incoming 12V down to 5V is that anything above about 7Vdc is dissipated in heat. And the hotter these regulators get, the less efficient they become, until they overheat and go south.

The Arduino developers will tell you to run them with 9V DC adapters if you must and don’t advise using 12V. Seems the REPRAP PCB designers missed the boat on that one. No surprise, it’s open sore.

The instant you plug in a computer via USB the MEGA automatically switches from that 5V regulator to use the 5V coming in the USB bus. And the display stops flickering. So…want the MEGA to last a long time and only want to print from an SD card? Plug the Mega into a USB powered hub, use a 5V USB adapter (like you charge a phone with), or leave it plugged into a computer. Mine happens to be connected to my Astroprint box which is supplying the 5V to it.

The more you print and dive into the workings of 3D printers, the more you can see weak areas in them. You can spend a lot of time innovating little things that could be better.

I’d hazard a guess a lot of people jump into 3D and drop out of it just as fast when the problems rear their ugly heads. If you’re in that leaky boat, one oar, and a 200 foot waterfall dead ahead, check for a local makerspace. In general these folks will go out of their way to help you get running again, and probably educate you in the interim. Probably save your sanity….

My Kossel update….today @ 1143 hours of printing time total, my temp for the hotend was all over the place. I wiggled the thermistor wire and it came off in my hand. Seems all that shaking around on those little thermistor wires took its toll. Fortunately I had a spare so within the hour it was back printing. Hence, spare parts are a must…


Memory Lane & Bunny Trails

It’s the darnedest thing sometimes. You’re working with some “new’ technology and all of a sudden, bingo, you’re transported back who knows how many years to something from your past.

Happened just yesterday. I was looking for emulators that ran on the Raspberry PI and next thing, I’m at Adafruit’s web site looking at a Mac OS 7 emulator.

Mini Mac Pi

Running on a 320 x 240 TFT display inside a PLA 3D printed case. Talk about RETRO…

Now I remember getting the Mac in 1985. I was actively programming the TRS-80 Model 4 at the time and the company who was selling my software was doing the documentation on a Mac. When I sat down at the Mac in their office to do some documentation changes, I could see that the Mac was a great system for doing the docs so I went out and bought one. I had zero interest in programming a Mac at that time.

Then I wondered what other emulators did the Rasp PI have. Naturally I looked up to see if a TRS-80 emulator was around for it. Which of course turned up this post.

Ah yes the TRS-80. I started with a Model I in 1978, writing in Basic. Ugh. Needed a printer. My first printer was a behemoth Model 15 teletype (from WWII to give an indication of age) with a machine language driver you had to load every time you booted the computer.

M15 KSR 3a

Course this was a Baudot printer which means you didn’t send it ASCII, you sent it five level code, that tickled a relay at a whooping 300 baud. And it printed. Solid steel, heavy as heck. Noisy. Mine didn’t have the keyboard.

Which is how I got started programming in machine language. Someone wrote a driver for the TRS-80, non-relocateable, and if you wanted to change it, you needed to learn how to PEEK and POKE commands. Not to mention learning how to read machine code.

For whatever reason Z80 machine code was fairly easy to learn for me. I never wrote another line of Basic code after 1980. I wrote Z80 assembler until 1992.

In the mid-1980’s I decided to write a program for profit. I had done all the market research, I saw the need, I wrote the code, marketed it. I sold three copies. Wow. LOL. The problem was twofold, first I went for profit, second I had no intention of ever using the program myself. Since then I have never written a program that I did’t intend on using myself. I.e. money has never been a motivator behind writing software. Granted I’ve sold a lot of software since then too, but I needed it and therefore use it myself, apparently others needed it too and have bought it from me.

One of the most successful programs I ever wrote at the time was my BBS (bulletin board software; a precursor to the internet that wasn’t around at the time) and a terminal program called FastTerm. As of 2011, some were still using my FastTerm program in emulators to connect to DEC systems. Or whatever. I dunno. Last FastTerm I did was 1991 if I remember right.

Someone even wrote blogs about how to use my YMODEM protocol in FastTerm. And the screen shots, yep, that’s me. Notice the play on the Mac menu/desktop system.


But I had the Mac and in 1988 I started programming for it. At the time that started because there wasn’t exactly a lot of terminal programs for the Mac, let alone ones with X and Y modem protocols so as a bit of a challenge I wrote one. It wasn’t like I didn’t know how it worked.

Course this started me programming the Mac and I’ve been programming it ever since.

I did write 3 or 4 programs for Windows over the years, but the programming environment wasn’t exactly friendly, API’s were limiting and it got in my way more than it helped me get the task done. Thus I never pursued it, just stayed with the Mac.

Of course when you start programming so long ago, you get a work flow and it tends to stay with you. Versioning Control was something I never did because I was programming before it came in vogue. Heck I was programming before ethernet got to be a mainstay. When I write software now, I can create umpteen versions of a piece of code in a day or hour for that matter. When I did investigate VCS, I spent more time making notes about what version did what than writing code.

So for me, I start the day with a “Save As” new version number on the filename. If that version works like it should, I do another Save As, increment the version number, carry on. In my mind I can rewind or reopen any version prior, but the way I debug as I write, I’ve rarely ever had to do that. Maybe that comes from writing machine code in the early days…

Now I write piles of code for the Arduino because it’s the part I love best about the whole thing. Code controlling something in the real world. A visual example of working code if you like. This is no doubt why I have absolutely been enthralled with my 3D printer. I can design something and then turn it into a real world whatever with my 3D printer.

For as much as technology has “evolved”, coding hasn’t changed much and frankly neither have some of the protocols. Serial for example. I was programming modems via that in 1980. Shows how robust it is. Actually electronics hasn’t gone off the deep end either. True you have micro controllers in places where discrete components used to live but Mr. Coffee has been around a long time too, so in many cases, same old, same old but with new skin on it. Lift the hood, same old.

So that’s how memory lane can descend into bunny trails in a matter of minutes. Smile worthy though.


Slic3r Not Always Nicer

Today I needed a part printed (making a new effector mount), so I sliced one up, sent it to my Astrobox (wireless Kossel controller) and started the job.

Second layer and the extruder went bananas. Retracting and injecting at warp speeds, filament grinds going all over the place. Which didn’t do much for the print. Ruined it in a heartbeat.

I thought, bad upload. I resent the sliced file, started the print again, same thing happened in exactly the same place.

Ah ha! A damaged slice. I resliced the object, sent it, same problem.

Time to sit back, sip some coffee, ponder all things…

I loaded the STL into Slic3r, then clicked on the LAYER view. Nope, nothing there was wrong. Changed the density of the infill to 50%. Checked the layers, nothing wrong again. Last step was to change my layer height to .2 (mine normally defaults to .3). I saved the Gcode and sent it to the printer.

Same problem. Oh wait, I didn’t check the layers after the .3 to .2 adjustment. So back to Slic3r and do all the steps and then finally change the layer height and checked the second layer.

This is what showed up.

Slic3r Error

Now where in the float did that come from? Changed it back to .3 height, perfect. Back to .2, all messed up. The GCode was completely screwed up too so no wonder the extruder went on medical leave.

All the layers above or below the second one were correct.

Thus let this be a heads up. I can’t always blame the printer, server, or STL file for a screw up. Sometimes it’s the Slic3r that’s not nicer. Use the Layer View, might just save you a bit of a head scratcher.


The printing of a Heffalump

My Kossel printer seems to work best when I don’t ignore it. In other words I try to print something every day. Might be a quick item or not. Might be to test out some of the settings I’ve changed in Slic3r. Might be because there’s nothing good to watch on TV so I watch the tool path…

Lately I’ve been experimenting with trying to lessen the strings that my PLA seems to want to leave behind. Which means I’ve been mucking about in Slic3r and seeing what the settings do. Let me rephrase that, I know what the settings do, I want to see how that affects printing.

Reading through all the posts on the web one comes to the conclusion that not too many deal with a delta printer to start with. Last stat I saw was that 20% of the 3D market might use a delta printer. So finding relevant information is somewhat of a challenge right off the bat.

The issue with delta printers is that they tend to blob and string. No, that’s not a 60’s dance craze. The “suggested” settings for the Kossel was something like this:

Slic3r Retraction

The problem comes from the Bowden tube. We’re shoving filament into it and it flexes to provide a constant pressure into the hot end. The longer the Bowden tube, the more pressure/flex you have at the head. When you get strings, you need to relieve this pressure. Therein lies the start of the quest.

A lot of 3D users tend to attribute stringing to temperature. Though it’s true temperature can affect it, during my testing, that wasn’t the real root the problem.

What I found through a lot of experimenting is that 5mm of retraction is about right. Too much further than that and you run the risk of hauling the filament up into the cooling section, where it solidifies and causes a hot end jam. Something you don’t want to experience.

Retract on Layer Change is a good one to have enabled. It sucks the filament back in as the head moves up to start the next layer. And it works like this. When the printer gets to the end of the layer, the head will pause for a heartbeat as the extruder rewinds the filament for the jump, then it jumps and shoves the filament back into the head. This is where the first problem can show up as a blob just before the jump. Might be a small blob, and dollars to donuts if it’s a blob, you’ll get a string as the head jumps.

The option that is supposed to address this is “Wipe While Retracting”. Keyword being “supposed” to. My experience is that it makes more strings. Why? Because the hot end was sitting on that last bit before the jump and as it moves the filament is dragged out of the nozzle. In other words the retract happens over the distance of the jump. Too late to be effective.

At this point I think I have read every bandaid remedy there is to “fix” this. Guys will lift Z, increase the retract length, add extra length on resume and so on. For the retraction speed, some slow it down to 20mm/s while most tend to leave it at 40mm/s. Some even enable the “volumetric” retraction in the firmware. All with mixed results on a delta.

Like a good sheep I went through all the settings, trying all the suggestions and walking away scratching the ole noggin. You know the old saw about doing the same thing wrong over and over but expecting different results? Yeah. Time to get outside the box.

So I went back to what the problem was from in the first place. The Bowden tube. The pressure. I needed to relieve the pressure in the hot end. Fast. Ah ha. Speed. I don’t profess to know how fast a non-delta printer can retract, but the Kossel can be very fast. So…one change…


Set the speed to 80mm/s. The filament I have been fighting strings with suddenly has no strings. Zip. Nada. Zilch.

The extruder is happy as a clam, doesn’t heat up, when the filament comes back in it doesn’t miss, and surprisingly enough it works really well. Time to try it out on a small detailed print.

Enter the Pachyderm. I spotted a design of an elephant on Thingieverse:


Admittedly I haven’t been this close to a elephant since I worked with them in my teens (Shrine Circus) but it was small, detailed and fit the printer bed. Obviously not a life size version…

Kossel Ele1

I’m using a .2 layer setting, I have a 120mm fan on the side of the Kossel that’s outside of the picture frame. This is a small print and unless the filament is cooled quickly it’s going to end up in a mess.

Kossel Ele2

So that’s a bit of the elephant done. If you look into the body fill area you’ll see some strings. I know I said strings were all gone. They are. This is a 5% Honeycomb infill and at the speed it’s going I found it always strings. Using a 20% of more infill and those strings would be gone.

Kossel Ele3

What’s really important is that there are NO strings between the different sections of the Heffalump. You can see some saggy parts (and trust me, real elephants have some saggy parts too) around the legs.

Kossel ele5

Without a doubt the only time an elephant would fit in the palm of your hand. What I hadn’t realized when I started the print job was that this elephant was articulated. With a tiny screw driver I gently pried away the legs (the head moved freely off the get go) and all of a sudden:

Kossel Ele6

There’s Jumbo standing on my hand. I came away from this printing with a lot more knowledge than I had at the start. Retract speed, print layer, print cooling, and so on. So yes, Jumbo probably cost about 28 cents to print but he brought a lot of valuable information with him.

But the best part? I don’t have to shovel up after he’s “done”…:-)


Kossel Layer Example

For printing on the Kossel, unless it’s something special, I normally slice with a .3 layer thickness. This generally gives good speed and acceptable quality. Kind of a win-win if you like.

When I am doing something that is more detailed or I want a nicer finish, I might print at a .2 layer thickness.

Rob, one of the FVMakers, mentioned he was printing with a .1 layer thickness and I think that’s what got me thinking about it. But I had some serious doubts at the same time.

Consider this, we’re squirting PLA or ABS from a .4mm nozzle, that merrily swings from a mathematically nightmarishly controlled upside down tripod, driven by an 8-bit CPU, that we’ve levelled on the print surface with a piece of 20lb bond paper. Of course we expect maximum accuracy with a setup like this and we should have it! Or not.

If you’ve ever looked at the bed maps from any delta printer, you’ll see high and low spots all over the map. Even if said printer is perfectly calibrated for “flat”. Or whatever flat implies.

Conventional thinking says you print at a layer height that’s about 60-75% of the nozzle size. Thus a .4mm nozzle is usually the happiest printing at .3 or so. But if the bed is calibrated flat fairly well, you can indeed print at .2mm.

Truth be told, I only had the auto-level probe installed on my Kossel for 3 days when I first built it. I could see so many failings with it and the software, I’ve always manually levelled the bed. Takes me about 15-20 minutes and I don’t do it very often.

Tonight I thought what the heck. Throw caution to the wind. Let’s take this baby for a spin and she what she’ll do.

I blame Rob, at least partially, for this lack of judgement on my part. What? Well, yeah okay I spread the blame around so if bad stuff happens it’s not 100% my fault. I know I didn’t fool anyone with that so just let me have my dillusion.

I have a small print (80mm x 60mm), Slic3r creates it’s usual spasticated tool path, printed at .2mm.

Google point 2

Notice as you move around the eyes follow you? See I was trying to distract you. Did it work? No? Blame “he who’s name I mentioned previously”. Okay, so blame me. It’s easy to see the layers. Kind of looks like a relief map of some canyon actually.

The same print, but this at .1mm layer.

Google point 1

Can you say blended? Ignoring the cats whiskers I didn’t bother cleaning off, the transitional lines between layers are minor. Although the eyes still follow you…weird huh.

Flipping the print to the top side you can see the crevice is not as cleanly defined with .2mm. I am using a macro lens here so bare that in mind too.

Google point 2 Top

Take the same print but this time with .1mm it looks very smooth. And if you look closely at the circles around the eyes, not the bags under them, you’ll see they aren’t really circles at all. Nope. STL doesn’t have circles. You see polygons which is the best you can hope for with STL.

Google point 1 top

I admit that when I started the .1mm print I had my hand on the mouse for the job kill function. Turns out I didn’t need it.

When I read all the ads stating that these filament droolers are capable of 100 micron prints I lumped that into the same boat as a used car salesman’s pitch. Makes for good ad copy and never thought about it again. Till tonight when I remembered “he who will not be blamed” had seeded the idea in my cobweb encrusted mind. Then I printed out these two objects.

While my objects are purple, you can colour me amazed.…you know who…;-)


Octoprint, Astroprint and Wireless – Oh My!

One of the not so nice things about owning a 3D printer is the fact that it’s tethered to your computer. With a USB cable. Probably a short one. And if your computer goes south as you’re working on it at the same time as you’re doing a print job, well, it does tend to expand ones vocabulary.

In some cases a tethered 3D printer means you need to clean your stuff off the table so dinner can have a place to be served. Providing your have an understanding spouse. Who wishes you had a faster printer. And to save you the time, no, this will not work as a ploy so you can BUY a faster 3D printer. Nice try sparky. Don’t ask me how I know this.

According to the USB 2.0 specifications the cable length can be 6ft 7in to 16ft 5in.

My experience with cables is that unless the cable is well made, i.e. properly shielded and decent sized wire inside, you want to stay with shorter cables. But, I’ve seen a 10 foot cable with a USB extension cable jammed together to make a 20 footer plus in some cases. Followed by a question of why the printer “seems” to have an issue printing. Hint, it’s not speed, it’s noise and signal degradation.

I see a lot of long USB cables that would serve better as jump ropes than for reliable data transmission.

If your printer setup is in that particular kettle of sardines and you need a longer cable, do yourself a little favour and look for an ACIVE USB cable. I have a 20ft active cable that works just peachy.

Of course the dream of many is to get the printer released from the computer altogether. As in wireless. And that’s the gist of this article.

When I started down this rather 4×4 rocky road, I had a great a mythical vision and a tankful of dumb. A dangerous combination to be sure but since when did that ever deter anyone?

I started by looking at the problem. Wireless. I have lots of plain RF and RF UART modules. We’re talking short range for the most part. Should be a snap. The Arduino MEGA in the Kossel has spare serial ports so I could shoe horn into the USB port one (TX0/RX0) or I could use some of the unused pins on the MEGA for raw RF.

I modified the Marlin software and stuck these in.


Insert first rude awakening here. The modules I used are one way only. Turns out that the software in Marlin sends an OK after it processes each line of G-Code. When the host that’s sending the GCode doesn’t get the OK, it waits. You know, like forever. While I was burning some of the dumb from my tank, I also found that these things can manage 4KB per second.

To put that speed in perspective think dial up modem speeds. From about 1982. Good. Got that out there.

The Kossel wants 250,000 baud. Time to look for something that will do that. Next up, I tried Bluetooth (HC-05) and I actually got the thing working. Course there was no way in the real world it was going to work at 250K so I started to crank the speed down.

Two new problems reared their ugly behinds. The first is when you get a speed that the Kossel will run at, you find that there’s a delay in the transmission rate in Bluetooth. So you send GCode, the MEGA immediately (it seems like it anyway), sends back OK. The HC-05 does not like to be pestered this fast. Better for slow speed mice and keyboards. Secondly, there’s a buffer in the HC-05 of 20 bytes.

Rude awakening number 2. But I’m on a roll and really burning through the dumb fuel now!

To make this short, after trying numerous other RF “things” I came to the conclusion it is like the engineering triangle. You have cheap, fast, reliable. Pick any two.

Abandoning RF, I decided to tackle a 3D print server idea. Repetier has host software that you can run on a computer that is connected to your 3D printer but it means you are tossing a full desktop/laptop computer at the problem. Fail.

Which led me to the Raspberry pi.

An all in one Linux based SBC (single board computer). Dreamt up by some guys in the UK as a solution to a problem that they seen. I mean all you have to do is connect up a keyboard, monitor, power supply, and a powered USB hub. Heck, a second hand computer has to be in the same ballpark for cost and a heck of a lot more friendly than Linux/Python and Scratch. It’s like “instant water!”, just add water, stir and you’ll have….wait for it…water.

I picked up a Rasp pi B+ Model 1. I didn’t even know there was a model 2 with double the ram and faster speed.

Try to imagine what happened when I put Octoprint on the lowly pi and watch it stutter PLA. Now apparently this combination works for some, but at the same time trolling through the various forums it certainly doesn’t work for a lot of people. I am one of those.

Jason, one of our fellow FVMakers, was kind enough to loan me a Model 2 Rasp pi to try out. Stuck Octoprint on it and was printing in a matter of minutes. Admission, Linux is for nerds, I am a nerd.

See, at this point, I was down to less than half a tank of dumb. Things were indeed looking rosy. Ah ha. I thought I had the solution. Nah, just setting myself up for the next head banging session.

I ordered a couple of Rasp pi’s MODEL 2’s from Adafruit, along with various Wifi adapters.

Then I got a WooToo powered USB 3 hub from Amazon so I could power everything. You read that correctly.

The pi runs off 5V input but cranks it down to 3.3V internally. The USB ports on the pi are severely underpowered and I’m told there is a fuse on the input to the pi that makes sure you don’t draw too much current via the USB ports. I don’t want to find out.

So how do you know you’re sucking too much power? Turns out the pi, if you plug it into an HDMI monitor and boot it up, will show a colour rainbow icon towards the top right of the monitor. I had a Wifi adapter, USB keyboard plus mouse and I was seeing this “blinkin'” icon during my tests.

The Raspberry pi only has 600ma of current available and that is shared for it’s four USB ports. So the design team on the pi made some rather goofy choices since USB standard is 500ma per port. But hey, it’s Linux. So hackable. You can modify the pi’s boot up config to give you 1.2A. Course at this point you just might need a fan. To cool the pi’s voltage converter because I don’t think it was ever designed to handle that amount of current.

Finally down to a reserve tank of dumb, I finally sat down and started to read specs. Because the internet is full of such accurate information.

If wireless 3D printing is your goal, here’s what I found works, reliably. And fast. Notice which one is missing? Good.

There are other Wifi adapters for the pi. I tried several different ones, all of them a LOT smaller than the Ourlink one. First thing I found was they drew more current and the range on them was pathetic. Great if your router is in the same room, not so good otherwise. And really, these dongles all cost about the same anyway.

The Kossel doesn’t plug into the PI’s USB ports. It goes into the powered hub. I found when I had the Wifi dongle and Kossel plugged into the pi, the LCD display on my Kossel would flicker. I don’t see this as a good sign.

Now that I finally had all the hardware, it was time to try out the software. First I tried Octoprint. Download it from a precompiled IMG (image), dumped it onto a microSD card and good to go.

Right. In an alternate reality.

See the catch is the microSD card. Although some SAY it will work on as little as a 4GB card, forget it. Get a 8GB card and make darn sure it’s a CLASS 10 card. The Rasp PI site says class 6 is a good trade off, but I found the class 10 cards work better than the class 6 cards do. The main difference is the class 6 card works slightly better when dealing with lots of small files. Whereas the class 10 handles larger reads and writes better. GCode files aren’t all that small…

Some class 10 cards also sport UHS-1 (3) as in ultra high speed. I had a lot of trouble with some of these cards and I can’t find any info from Rasp themselves that say they even support this higher speed card.

The next order is an SD card formatter. No, not the one that’s built in your computer. One from the SD manufacturers. And if your card supports it, use the option to optimize the card format.

Good the card is formatted. Now you need the downloads:

Read the text on the site, if you’re using Wifi you need to edit a file on the SD card after you write it.


Astroprint (astrobox)

Oh wait, how do you get those images on the SD card? Right you need more software…

For Windows – Win32DiskImager

For Mac users – ApplePIBaker

Essentially select the image, select the card, write. Make a coffee while it’s doing it.

If you wrote Octopi and you need Wifi you’ll need to edit a file on the card.

In my case the easiest one to get running is Astroprint. It starts up as an open access point, you connect it to, configure it to connect to your Wifi router, reboot it. Do a few settings, good to go.

Octopi, not so easy. Lots more settings, not mobile or tablet friendly. But there’s vids on setup works a lot like Repetier Host.

I’m not going to review the two packages, but I’ll tell you that Astroprint is a fork of Octoprint. The main difference is in three areas. Complexity, control, and features. If you want MAX control and feedback from your print job, Octoprint. Hands down. Octoprint will even do time lapse videos of your print job, not to mention realtime video streaming of your job (about 5 or 6 seconds lag).

If you want to send a GCODE file, start the printer, then go about doing something else. Astroprint. Hands down.

Both of them are easy to access via your LAN or WAN if you want to open up a tunnel in your router and use an access password. They use Apple’s Bonjour (zeroConfig, mDNS). This is built in every Mac out there, but you need to install it in Windows. If you use iTunes in Windows you probably have it already installed. Or you can download it directly here

Running your browser and typing in astro.local or octopi.local will bring up each servers control pages. So you can control the printer, upload files to the sd card to print, or directly to the SD card in the Kossel…well, Astroprint doesn’t always like the SD card in the Kossel so I don’t have one in mine.

It’ll take a few seconds to connect to the pi server, if you need it faster, I don’t know what you’d do to speed it up. If Octoprint is busy with a print job, I found I can make it stall for a split second when I connect. I have a hunch that because it’s streaming live video that sucks all the CPU power up from the pi.

Astroprint does no do live streaming but offers you the option of plugging in one of those old USB cameras like Logitech makes and selecting to do a still capture at that moment.

A popular feature of Octoprint is the time lapse photo ability. It takes a series of photo shots as your print is happening. These look great on YouTube when compiled into a video but beyond that I don’t really see much of a use for it. Plus the jpeg streaming that the pi does tends to tax it’s wee little processor. Even the PI 2. Enter overclocking for the PI. I don’t do it.

Using either server via a browser is simple, but Astroprint is far better designed for mobile browsers. Octoprint doesn’t have the same capability. Thus controlling your printer from a phone or tablet is easier with Astroprint.

One of the better reviews I found online for Octoprint vs Astroprint is this one. The thing to keep in mind that it’s more like an arms length review. There’s no lifting the hood to see what makes it tick, nor any real world tips or techniques. Or things to watch out for. I like to kick the tires on these things and see what rattles.

I ran each one for a month and put both through their paces. In my humble opinion neither one of them is perfect. Both have flaws in different areas. Therefore you have to examine each to figure out exactly what the most important parts are for you and then use that setup.

In addition I found that you really don’t want to leave them running 24/7. I have my PI 2 setup along with the Kossel on a power bar. When I want to print, I power them all up at the same time. Both Octoprint and Astroprint work more reliably for me that way. In not doing that, I found the Octoprint would lose it’s connection via USB to the Kossel, and Astroprint would screw up the controls if I cancelled a print job. REBOOT….:-)

There are still things I am trying to figure out, one is the amount of SD card space left when you send a design to print. Octoprint makes this easy, Astroprint I have no clue. It says nothing about available space that I have found.

Lastly, for whatever reason, printing from the computer via a USB cable is faster than having the PI with the servers to do it. It makes zero sense at all to me why this should be, but timing the same print job always takes less time coming from the computer. Just weird.

There are other 3D print servers that run on a PI as well but not free of cost. Like PrintToPeer. While the software appears to be open source, you need an account with PrintToPeer so need an internet connection where you send your STL file, they slice for you, then send it back to your printer for printing. Judging from the web site it looks to be aimed more at commercial applications than home users.

Setting aside the cost of whatever a PrintToPeer account is, connecting to the internet so I can control a LAN based 3D printer sitting 10 feet away me means they must have found my empty dumb tank and possibly a few others and overfilled them for their own use. Words “self reliant” have been lost on these guys. I just don’t get it. But they aren’t the only ones doing this net/cloud based stuff for your local control. Honeywell and Accurite do as well. They should pay you for doing their market research.

Hopefully you got something from all this info, because I’ll tell you straight out that the water’s pretty murky when it comes to print servers. There’s a lot of time when you feel like your row boat only has one oar. Hence when you see the caption that says, “Untether your 3D printer for $35”, you want to do a oar count.


IoT – ESP8266

IoT or Internet of Things has been gaining momentum for a while now. The first I’d heard of it was last year when I was introduced to a Spark Photon. A 3.3v Wifi enabled controller with some I/O and a cloud based programming software solution.

My initial reaction was disappointment. First off, 3.3V, my world revolves around 5.0. A new “language” or “script” to learn and being tied to a cloud for developing for it. So with three strikes against it, it wasn’t something that I was getting any warm fuzzies from. Another solution looking for a problem to solve.

Then along came the Photon, Spark Core and a plethora of Wifi things for Arduino (well, micro controllers in general). The Wifi things caught my eye since most could be a access point or client. I think I ordered almost every variety of them that came out.

Thus everything from the USR-232_WIFI, CC3000’s to ESP8266-01. They all shared pretty much the same communication route. That being of Hayes modem AT style commands. Yeah, Hayes. From the early 1980’s. I got the feeling we’d come full circle on that dog and pony show. Ugh.

However the ESP8266 was so economical that guys would devise all kinds of circuitry to restart them to maintain connections, or spend hours to make them more reliable. Finally adding a LUA interpreter because reliable the early ones weren’t. But they were cheap. You just knew folks would keep hammering at them.

As it turns out the ESP8266 is a fairly powerful micro controller with WiFi capability. So from the first sign of it in August 2014 until now, there’s been a LOT of people beating this thing into submission. For them, my hats off to you because we are now at a point where I can actually use them without spending time and energy debugging them.

So what’s the specs on the 8266?

  • 32-bit RISC CPU: Tensilica Xtensa LX106 running at 80 MHz
  • 64 KiB of instruction RAM, 96 KiB of data RAM
  • External QSPI flash – 512 KiB to 4 MiB
  • IEEE 802.11 b/g/n Wi-Fi
  • Integrated TR switch, balun, LNA, power amplifier and matching network
  • WEP or WPA/WPA2 authentication, or open networks
  • 16 GPIO pins
  • SPI, I²C,
  • I²S interfaces with DMA (sharing pins with GPIO)
  • UART on dedicated pins, plus a transmit-only UART can be enabled on GPIO2
  • 1 10-bit ADC

There are two devKits 0.9 and 1.0 that make connection and programming MUCH easier! Starting with version 1.6.4 of the Arduino IDE you have the ESP8266 core libraries built in. You program the 8266 just like it was an Arduino, even the blink sketches.

GitHub has instructions and information…really once the IDE boards manager has installed the ESP8266 and its variants, using it is pretty simple.

I’ve built a couple of projects now where at one time I’d have an Arduino, ethernet board, Wifi dongle just to get WiFi. Now I do the whole thing with the single devKit 0.9 or 1.0.

I’ve connected LCD’s, DHT sensors, RTC’s, switches, and even NeoPixels. All work flawless. If there’s one area that you can get caught with, it’s that it’s still a 3.3V chip and jamming it with 5V signals is not good if you want longevity from the board. Use a level shifter or voltage divider if need be.

In closing, now that I have used the ESP8266 12 and 12E devKits, I can’t imagine using Wifi any other way. I feel we finally have a product that’s worth of the tag, IoT.


Nixie Clock Kit

Apparently I have an affection for clocks. Of all types. The latest is is a bit of a throwback to when I was.. ah.. a lot younger..

Specifically a nixie clock. Nixie tubes are those cool looking neon tubes with numbers inside:


The tubes take 170vdc to “light” up a number so there’s a bunch of leads on the bottom of the tubes. Digits 0 to 9, two decimal places and an anode. Appearing in the mid-1950’s, they fell out of favour in the late 70’s but were still being made in Europe and Russia. For a while any way.

Now that no one makes them any more, they are becoming more valuable and the driver chips for them even more so although there is a Russian knock off that’s around. None the less a search on eBay will turn any number of sellers of them (usually in Europe someplace) and of course all manner of tubes. Some removed, some NOS (new old stock) not to mention a slew of different types.

My favourite Nixie have always been the IN-14’s. Large, easy to read, I like them. My first “digital” voltmeter, a Heathkit IM 1202 used Nixies:

IM1202 front1

Used it for twenty plus years before going to LED. Kind of wish I still had it…

I decided that rather than design one, I’d start with a kit. of course my idea of a kit and what others idea of a kit is can be vastly different. When I think “kit” something like Heathkit or Tamiya comes to mind. Lately my “kit” experience has been quite different. More like a box or bag of parts and an online PDF.

There’s a number of different companies producing Nixie clock kits and I chose PV Electronics in the UK for one reason. I liked the concept of an easy to replace Nixie tube because I know they don’t last forever. PVE sells on eBay and I ordered one of their kits. It arrived in short order consisting of some PCB’s and a bag of parts.

At least someone had the forethought to put a URL on the parts bag so I could find the assembly manual online.

Thus armed with a PDF I started to assemble the kit. At the same time a letter shows up from the eBay seller with a fuse part inside. Apparently it was missing from the parts so it was a good thing I’d delayed a few days before starting to build the clock any way. Now that’s what I call support. Rare indeed.

My 55 year observation is there are people who can write assembly manuals and there are people who shouldn’t. To their credit PV Electronics falls right in the middle. The information is in the manual, but it’s not organized in the best way that’s going to walk you through the build. I.e. the parts are on one page, the result on another, and no outlines on the result picture to show you the area you’re working on. Nope. Just a R12-R28, R32-R34 XYZ resistor and your PCB should look like this.

Considering the quality of the PCB and the parts, the assembly manual was a bit of a downer. But at the same time, not all that surprising.

I spent considerable time examining the silk screen of the PCB looking for where the parts go.

The resistors are metal film types and if you can’t read the colour codes or have access to a resistor tester you’re up a creek.

Hence right off, this is NOT a kit for a novice in my opinion. Certainly it’s one thing to know how to solder but to get this kit working I had to do some trouble shooting for defective parts. Parts that I have replacements for and know how to trouble shoot. Whereas a novice would be up that creek again…

The complete assembly took me just over 3 hours. Mostly spent flipping back and forth on my iPad reading the PDF.

At the end of it, I fired up the clock and it worked. All Nixies were perfectly. But one of the RGB LED’s was missing the blue component. I traced it to a faulty RGB LED. I replaced that and the clock worked as it should. But I have these components in my parts bin so not a big deal for me.

The heart of the clock is a PIC chip. Which means it’s a software clock. It is adjustable. How accurate is it? Compared to the lowly DS1307 or the more robust DS3231? No idea. No pun intended but time will tell.

The kit does sport a serial port so you can connect up a GPS to it, program in the offset to where you are and it will compensate. This is the route I took. I got one of my GPS breakout boards that I use on some of my Arduino projects and wired it, via a jack, to the Nixie clock board. It works absolutely flawless. Even if you forget the GPS chip when you move it around (I took it to some meetings to demo it), the Nixie will tell you when it hasn’t sync’d in a while. I plug in the GPS and with seconds, it’s running perfectly on time again.

Also some of the parts have changed since the kit was done in Dec 2014. A 15pf cap was replaced with a 33pf cap. No doubt to help with the software clock. There was supposed to be a piece of paper in the parts bag that told me this. Keyword was “supposed” to be. I never saw one if there was.

The “super” capacitor has changed value so if you go looking for the one in the assembly manual you won’t find it. You’ll find one that’s physically identical but a different value.

That’s one of my pet peeves with kits. If there are substitutions, update the assembly manual. Simple. Do it.

Here’s a shot of my completed build.

IMG 0214


IMG 0213

Thats right, there’s LED RGB lights under the tubes that you can create custom hourly colours with.

I still really like the concept of the tube sockets but I’d say it’s a kit aimed at more experienced builders.

My brother looked at it and I got a “cool” thumbs up from him. He loved it…


Apple Yosemite – Kitty Litter – er… Sandboxing

I upgraded one of the development Mac’s here at the Wanch to Apples latest OS, Yosemite (10.10). If my brain had been firing on all cylinders I’d have made darn sure that I also had a master install of Mavericks before I did so.

Yeah, well, to make a long story short, I didn’t have a Mavericks to fall back on.

And it’s not that Yosemite is actually all that bad, except for a couple of areas. Specifically display and printing are the problems.

The first issue is the display. Obviously someone at Apple loves that flat, pastel, washed out, pre-kindergarten look of the OS. I am not even a remote fan of that person(s). Matter of fact, I think I’d vote for flogging in this case. Any amount of Googling and you’ll find people evenly divided between the “love it” and “butt ugly” camp. Of which I could run for president of the latter.

It wasn’t until I happened to visit an Apple store that it suddenly made sense to me. Heck, didn’t even require a “Genius”, not that I’d actually label many of them as that. Nope, one look at the 5K display on the iMac and yep. That’s why they did it. It actually looks fairly good (still won’t give it more than that because I’m not a photo guy) and at least the writing is legible on the screen.

Here at the Wanch there are several Apple Cinema displays. Sizes range from 20″ to 30″ and those monitors have always been, quite simply, stunning. Right up to the point when you look at Yosemite on them. From that point on you swear there’s something wrong with the contrast and clarity of your rather expensive monitor.

Fear not, mere unwashed mortal, it’s a nefarious plan by Apple to encourage you to buy into their rectal…I mean Retina displays. Because we do nothing but look at glorious waterfall photos all day. There, I can remove my tongue from my cheek.

Admittedly the 5K displays look great, but at the cost of making every Cinema display look like crap. Sorry, that’s not really an acceptable trade off in my mind.

I should have seen this “hi-rez” coming judging from the iPad leap into the retina canal. But I didn’t. My dumb.

Now if some nice Apple employee has a nice shiny Mavericks install they want to get rid, do email me. I’d appreciate it. Because then I’d make a dual boot Mac.

The second area I, and apparently more than just a handful of others, have been fighting with is printing. I have a nice colour laser, Xerox 6015N, and when I try to print multiple copies or a page range I have, since I install Sam (you know that Yosemite cartoon), I get a “Job Cancelled – Filter Error”. And here I thought Windows messages were cryptic.

I contacted Xerox, was told that they don’t have a Yosemite driver, no idea when or if they will. Sure makes you all warm and fuzzy with their products. I scanned every message board I could find, looked at the console error messages. It looked like where the Xerox drivers were installed was different than what the OS wanted. Beyond, I had no clue.

Plainly the issue started with Xerox, and Apple finished it.

Tonight I was wandering the forums again, because the one Yosemite Mac I have actually did print without an error until tonight. Had to get this fixed. And then I found it. Turns out it’s Apples sandboxing. The fix is two lines in the Terminal app.

sudo sh -c 'echo "Sandboxing Off" >> /etc/cups/cups-files.conf'
sudo launchctl stop org.cups.cupsd

I immediately searched for Sandboxing and printing, came across a number of others who used “Sandboxing Relaxed”. No, I’m quite happy to be printing finally, as are a number of Kodak users from the sound of it, so Sandboxing can stay off.

If you’re a little worried about security, although I’m not sure how you can use the words “Internet” and “Security” in the same sentence, then try the “relaxed” option first. Me, I have work to do so what I don’t need is some OS getting in my way of that because “it” thinks it’s doing something useful.



Garmin Approach G6 – Golf GPS

Since Callaway is dropping all support for their GPS systems in September 2014, my original uPro is destined to go the way of the Dodo bird.

If you only play local courses and have less than 50 of them, the uPro will still work fine. However, no new courses will be added and support will be gone. Not that there was any significant measureable level of support to begin with. And Callaway even thanked me for supporting them in the same email they informed me that I was about to own an orphan product and suggested I should check out their club offerings.

So maybe it’s just me, but I don’t get any of the warm fuzzies from a company that treats me thusly. Therefore, thanks Callaway, but not a chance.

I’ve been shopping around for a replacement golf GPS and wow. Does that take some concentration. A good site for some information is over at Critical Golf. Obviously I have specific criteria when it comes to a golf GPS.

  1. Unlimited Courses
  2. Unlimited Updates
  3. Mac friendly
  4. Ability to keep score
  5. Accurate
  6. Good customer support

Previously I’d owned a Golf Buddy. Bit of a nightmare to use on a Mac. I’d tried a SkyCaddie but I’m not big on the annual subscription but now I may not see that as a bad thing like I once did. Maybe it encourages fixes and updates.

In the end I settled on a Garmin G6. My feeling is that Garmin should know GPS, mapping, software and fit my prior list requirements. They get a good solid 4 out of 6. I was worried about the touch screen, there was no need. The only drawback to it is that you need to keep it away from stuff, like tees in your pocket, that will touch it’s screen. I keep mine in the netting on my ClicGear cart. But I can’t toss in a golf ball or it bumps the screen.


True, I could lock the screen, but nah. Couldn’t be bothered. I got used to the rather cartoonish looking graphics quite quickly. I have yet to find a local course, even new ones, that are not on the GPS. Even the executive Par 3’s are there.

Connecting it up to my Mac with a USB cable and it mounts as an external drive. Theres a Micro SD slot on the bottom of the GPS but the only reason it’s there is because Garmin repurposed the case from another GPS product. So it does nothing.

Pressing the buttons to get a green view, or score card is easy. Touch screen areas are easy to use and the GPS has been rock solid for GPS signal.

All, a very nice GPS. But it falls flat on its butt when it comes to the scorecard and customer support. First off, I keep score without handicapping, so it’s strictly a stroke score for me. If, some how you forget to enter in a score for a hole, the GPS adds one in for you. Now you’d think it would make the mistake of say, oh, giving you a par. Yeah, well it doesn’t. It gives you 255. That’s right. 255 strokes on a hole. BUT that 255 doesn’t show in your total score.

You figure that one out. I gave up. Here we have this nice GPS, when we advance to the next hole, if the score is 255, it should have the brain to prompt you for the proper score. I’m sure I’m not the only one who’d like it.

So…fail #1.

Due to the fact that I’m a Mac software developer, I get more into things than a lot of people do. Actually probably more than most do. So what follows is what I found by looking under the hood. That’s the kind of guy I am.

The GPS comes with Mac software that serves two purposes. It updates the maps, it updates any firmware. There is no software that gives you a score card, stats or anything else. There’s some 3rd party stat software from Approach Stats. If stats are where you need to go, check them out.

I’m not much on stats, more on score cards. So I modified my uScoreCard app to read the files from the Garmin G6. If you have a Mac, Garmin G6 and you want an app, email me and ask. No charge.

So lets look a little more under the hood.

Inside a subdirectory on the GPS (it’s just a flash drive to the computer) there are your score cards as XML files. If you examine this file, there’s some interesting stuff in it.

First off: Distance Walked: 3.93 Miles – Elapsed Time – 4.14. That’s right. It tells you how far you walked (in yards, I just converted it to miles) and you get the elapsed time of the round. For the round time, keep in mind that it starts when you hit the PLAY and select a course. So it you set up the GPS for the course before getting to the first tee, that time will be in there.

Secondly, “Mel – 10 out of 13 fairways hit – 10 greens hit in regulation – 34 total putts“. Thats right. Fairways hit, GIR. However the GIR is CALCULATED. There is no where you can actually enter in that you got a GIR. If you are playing a par 3, hit onto the green, then proceed to putt off the green, have a bad chip, then finally chip back onto the green and finally two putt in. You didn’t get a GIR for that 6 stokes.

Is that right? Or not. I did some research and read that in the prior situation, the first hit was a stroke, once the ball is on the green, every stroke after that is a putt. Even if you knock it off and chip back on. The chip is a putt. I have no idea if that’s buried in a rule book or not, I found the info on a forum. So, take it for what it’s worth.

For accuracy, the GPS, for the most part is right on with the yardage markers (1-2 yards at maximum). One of the nice features is that you can move the pin for the hole to the rough location on the course to get the real yardage. I do this once in a while and I match it with a friends laser range finder. Works nicely as it remembers the new location until you advance to the next hole.

But hold on, let’s step back to the XML file…we have a local course here that has two different par’s depending on the tees your playing. The mid tees make the hole a par 4, the forward tees call it a par 5. There are THREE of these dual par holes on the course (Surrey Golf Club). The G6 gets one of them correctly. The two on the back nine show as par fives, with a back nine total of 38.

The frustrating part is you can not change it. Your strokes will be correct, but your pars and GIR’s just went out the window on two holes.

The finishing hole on Surrey Golf Club is a par 5, 559 yards. Yea, it’s long. I’m a senior and it seems forever. No amount of anything can get me to a wedge into the green. Always at least a mid-iron. Ugh.

When I examine the XML file for that last hole, this is what I find:

G6 18

Now if you’re anything like me, you’d LOVE to see a par five at just over 92 yards. It would make an Eagle look doable!

But no, it’s wrong. 100% wrong. And. Not. Even. Close.

Which leads me to the conclusion that the reason Garmin doesn’t have any software to show these “stats” is because the XML file generation is a mess. They DO have a “scorecard html” file on the G6 itself that you can run from a browser. So long as the browser is not Safari on the Mac.

G6 No Safari

But you can use Firefox:

G6 Firefox

Notice the 255 on the 15th. Yea. I forgot to put in the score. I can go back on the G6 and edit it I think. But I usually just do it right on the G6 once it mounts on the Mac.

Then notice the 10th and 15th are Par 5’s? From the mid-tees they are par fours. I found no way to change this.

Lastly, if you add up the strokes, notice I got a 79? But there’s a 255 on the 15th? Really want to trust your stats with a company who has seemingly bad math? No? So what the G6 does for a 255 is gives you par. If you want pars, just don’t enter in any score. You’ll be a scratch golfer in, oh, 4 hours…

So what does my software do, for one it fixes the par’s, and adds correctly. It will also tell if you have a 255 in there so you can edit it.

Surrey Golf Club ~ Main Course 2014 06 17

I admit that I use the same sort of calculation that Garmin does for GIR’s. My preference would be a box to actually enter that.

This information has been emailed Garmin. They are noticeably unresponsive. I think they’d just prefer that I go away and stop trying to hold them accountable for their blunder. As surprising as it might seem, Callaway had better support than Garmin does. But Callaway is getting out of the GPS business so that doesn’t say very much for Garmin. Does it.

So, Garmin, guess what? You get a fail for this half baked product. Now, scurry back to the oven and finish it properly.


Under The Microscope

I’m doing more SMD work (that’s using very small surface mounted electronic parts) and checking my work, even with a headset magnifier is difficult. I’d purchased a “budget” USB microscope from fleaBay last year but apparently you don’t get a lot of quality for $12. Surprise…

I’d read that the 5mp cameras with a 50-500x magnification and controllable LED lighting are the best way to view your stuff, so, I ordered a USB microscope:

USB Microscope

Of course when it showed up, the first of note was the total lack of Mac support. There was a MINI CDROM for Windows though and I did manage to install that in VMWare Fusion under WinXP to make sure the thing worked. The software is sort of “iffy” though with a non-intuitive interface. But then I find my Windows application tend to look cartoonish and sport functions they don’t explain. Anywhere, including the “Help” menu. Begs the question of why put in a Help menu in the first place.

For the Mac, there are some software choices, none of which I could find that take advantage of the two push buttons on the microscope. Namely SNAP and ZOOM. I was sort of disappointed until, I realized that when you’re hitting those buttons at a high magnification, you shake things. A lot. Thus, they are really kind of useless.

I found that the Macs PhotoBooth and Quicktime could use the “USB Camera” so all wasn’t lost after all. The third party software for the Mac didn’t really appear to have a lot of options I’d use any way.

While this microscope had good clarity and ample zoom, it also had a problem. A flickering set of LED ring lights inside. There is a POT on the cable that allows you to adjust the brightness, hopefully to get a better photo or image. If I wiggled the cable where it went into the microscope head, the lights would flash on and off. Like an intermittent connection. Not good.

I naturally contacted the eBay seller who offered me $5 to buzz off. Or ship it back at my expense (about half of what I paid for the scope). Eventually we settled on a higher amount and I thought I’d just build my own LED lighting for it.

This thought lasted an entire afternoon. Indeed. It bugged me. So, rather than leave things alone, I ripped it apart to see what the problem was.

I started by taking apart the pot that controlled the LED brightness. Not much to it.

IMG 4559

Getting the main unit apart. I got the silver ring off first (it’s around the unit by the Snap switch in the first image). Once I got that off, the end with the cable in it is sort of glue welded and it was easy to pry off the cable end.

IMG 4560

There was a metal clip that looked like it was trying to cut the cable in half that I removed. You can see the indentation in the cable. I used a zip tie to replace it. When I wriggled those wires the LED’s would flash. I thought it might be the main board or something so I started to look inside:

IMG 4562

I thought Behringer was the only company that used glue to hold little pieces in place. Nope. The blue and yellow wires are from the POT. I checked as best as I could for a bad connection but couldn’t find one.

It was when I started to look closely at the pot, I found the Blue wire was a cold solder joint. I resoldered both connections anyway and now it works just fine.

I put it all back together, used the Handyman’s Secret Weapon to hold the cap on (duct tape).

Previously I said that using the buttons on the unit are pointless because you shake too much (or I do). The best way to do it is to set it up on it’s El Cheapo stand, prefocus it, then slide a printed circuit board underneath the lens and don’t touch the microscope at all. The Mac Quicktime application works better than the PhotoBooth one. In PhotoBooth images are flipped and reversed.

My advice if you get one of these, see if you can find one with a VERY sturdy stand. I’m going to make a stand for mine out of wood. That will hold it very steady. And really that’s about all you need. It actually works quite well and the LED’s brightness control is very handy.


PMPO – Peak Music Phantom Output

Having a signal generator on the test bench helps with trouble shooting some of the audio circuits I assemble from time to time, but it would be nice to have a signal tracer as well. As in audio signal tracer.

Initially I had the idea that I’d assemble a token amp around one of those low power TDA chip amps. However when I started to cost out the plan with a case, pots, parts and so on, I thought I’d better check out the “other” source.

Of course, what better source for mislabeled widgets than the eBay supported off shore market. I didn’t need anything with a whack of output power, actually didn’t even need stereo inputs.

I immediately found a plethora of these styles of “iPod” amps for the low cost of about two Starbucks lattes:

IMG 4499

This thing is a MA-170. Whatever that’s supposed to infer. Digital Cinema Sound or DCS (darn cheap sound). Looking at the photos of the rear I found it looked like most of those car stereo power amps:

IMG 4500

Both the back of the amp and the eBay listing claimed this was a 2 x 100 watt amp. Which, incidentally, you could run off a 12VDC 2A wall wart. There was no familiar and standard RMS rating, just this PMPO moniker.

PMPO stands for Peak Music Power Output. It’s supposed to be a measurement of the peak power you could get from the amp in ms (milliseconds). Which is a pretty useless figure because even the shortest musical note lasts longer than the allotted PMPO measurement time. So, write down any number you like. It doesn’t mean squat any way.

You can also lump PMPO in with PEP (peak envelope power), amp ratings in IEC (Behringer used to do this a lot). Folks, if it doesn’t say R.M.S., you have no idea what it is.

If you’ve ever done anything with car stereo power amps or probably any amp at all, you’d probably already know that a 12V 2A supply isn’t going to generate 200 watts of anything. Except maybe warmth when it goes up in smoke.

Of course I knew this going in, but hey, I wanted a budget amp, so that’s exactly what I expected. And got.

When the box finally showed up, it listed the actual specifications on the side of the box. It clearly said, “2 x 4 Watts RMS”. Obviously a slightly smaller, and more realistic number than the 2 x 100 claimed elsewhere.

I contacted the seller and suggested their listing facts were incorrect. No, I didn’t outright say they were lying, I was actually quite nice and gently suggested they update their listing with the accurate information. To my efforts I received a reply that the box was “just any old box” they used for shipping and had no bearing on the contents. Considering the artwork and specs printed on the box were an obvious match for the contents, ya, they were lying.

Therefore, as Dave Jones from the EEVBlog says, “don’t turn it on, take it apart.”. I did.

IMG 4501

Removing the front splined knobs was the first trick. I swear someone must have pounded them on with a sledgehammer. I have a special tool for removing knobs on guitars without leaving scratches and there was zero clearance under these knobs to even use it. I pried the crap out of them with a screw driver then took off the four front mounting screws.

IMG 4502

You can just see an LED to the left of the volume pot, it lights up the translucent ring to show power is on and maybe, smoke to follow…

On the extreme right edge you can see the TAD chip “clamped” to the wall of the case.

IMG 4503

Slotted holes to allow for some “wiggle” room of course. You have to unscrew those two screws, let the clamp plate fall inside the housing to get the thing apart. Because the TDA chip doesn’t fit tight against the casing of the box, they put an aluminum spacer in there. Something else to contend with when you reassemble it.

IMG 4504

Looking at the PCB, there are as many jumpers as there are components. The 4558D op amp for preamp/tone controls, and the TDA chip on the right edge. That stuff that looks like silicon grease on the TDA chip isn’t. It’s silicon. Off shore construction seems to have found more uses for hot melt glue and silicon than Red Green has for duct tape.

IMG 4505

If you had to scratch build this circuit, it would cost more than buying the whole thing as is. So in that respect it’s kind of nice.

However the fact remains that the TDA is a TDA7266sa chip. An obsolete chip and its data sheet says 2 x 7 watt. Even the data sheet doesn’t say 7 watts RMS, PEP or PMPO. Nope, just 2 x7 watts. So you roughly figure RMS is about .707 times the power, well, surprise. You get 4 watts.

When I pointed all these facts out to the eBay seller, well, apparently I am invisible. I am quite happy with the price I paid and the product I got. But trying to pass a scooter off as a Harley is never going to work for me.



Today marks our 44th anniversary. I have trouble remembering the day after yesterday and yet, we both remember that day from 44 years ago. When you’re married at 18 (Carol was 16), you have NO IDEA what life is about, nor what you’ll face over the years.

We’ve had ups, downs, circles and bliss. Still found time to have a family, and we get together with the “kids” quite often and enjoy those times immensely. Thus, long marriages can work but you both have to be willing to work on them.

Since we’re both retired, we’re sort of in that “nesting” phase where you make things around you as comfy as you can. Because in reality you’re at that time of life where you can enjoy it. You know, if it snows outside, who cares. You don’t have to go anywhere and drive in it. If you have a sleepless night, who cares, not like you have to be somewhere at 6am in the morning.

Our family room has been our media room for 20+ years in the house and we’ve gone through a couple of furniture changes to make it more comfy. Last go round we put in a rocker, two stylish leather chairs. Yeah. Not so bright. But it worked for a few years.

We went shopping for “what would work”. We started with those theatre seats. Ones with power reclines built in, massage rollers in the backs, and hidden storage compartments. I’ll say right now, those things are BIG, HEAVY and PUFFY. A few thousand bucks a seat and really, the majority not something you’d want to sit in for more than a couple of hours. We have 12 feet and I could fit “maybe” three of them (narrow ones) in. Doesn’t work if you have the family over…

Then it was off to look at sofas. Nope, sofas don’t work. Hey, lets check out sectionals!

Off to Norwalk, Urban Barn, etc. About seventeen different places. Norwalk had one that would basically fit wall to wall in the area we have:

Before Sectional

Norwalk had one of those more luxurious ones, at three times the cost of any one else (and it was on sale; $10K). Nice but not $10K nice.

Being the smart cookies we are, we came home with the measurements and laid out the size with painters tape on the floor. I figure we had about 16 square feet of floor that we’d have to vacuum. The remainder would be covered by sectional. Good thing the Norwalk one was too big…

Ah ha! There’s the key. Find something that looks good, feels comfy, and doesn’t dwarf the room with it’s size. And there’s the rub. These things are made by places that use the cookie cutter approach. As in they make 15 different styles all exactly. The. Same. Size.

We found ones with lousy construction (no hardwood just MDF or something), poor padding, seat cushion zippers you could see from the front of the sectional (hello Sears, who designed that blunder), cheap fabric, and on it goes.

Finally ending up at Urban Barn (thanks to our daughter for that suggestion!). We found one we really liked, and went back to see it a couple of times. It was a little hard on my back. But we found another one we’d overlooked prior. Perfect. Course you can’t buy it because it’s a custom built one. That’s right. A local furniture maker, Stylus in Vancouver, makes them for you. You pick the fabric, pieces, foam, and so on. They make it for you, deliver it, and set it up.

An excellent warranty, local company, local store. Really, doesn’t get any better than that for us.

We’d ordered at the “cut off” date to ensure we had it by Christmas, so when they delivered it today, it was a bonus for us!

So what did I get Carol for our anniversary? Chocolates, flowers, dinner out? Nah, got her a sectional so she can snuggle in the corner and play Airport City on her iPad in comfort!

After Sectional


Not the baker…

As I did last year, I’ve signed up for golf lessons through the fall and winter at the local academy (NorthView Golf & Country). Andrew & Rosie have as well.

And like last year, because the sessions are drop-ins on each Saturday, we all take turns making dinner. So Andrew, Rosie, me, Carol and new to our sessions one of my golfing buddies, Dan.

Of the “cooks”, Andrew and I are the least talented. Which, of course, can make for some interesting dinners. The undeniable talent in our “group” is Rosie and surprisingly, Dan (being a confirmed bachelor may have something to do with that).

Although I can make meals, what I do enjoy doing from time to time is baking. I remember growing up as a kid, it was the baking that always fascinated me. Probably because it always had something to do with sugar. Yum.

Regarding nutrition and baking, when I was kid, we had meals from the five basic food groups: boxed, canned, frozen, instant and take out. Baking wins with me.

Rosie makes some buns (we call them Rosie buns) that are to die for and so, well, I remember making some dinner rolls. Yeah, a long friggin’ time ago but since when would that deter me. I decided to try making some. I got out my recipe books, noticed that although the ingredients were basically the same, completely different ways of making the sweet dough. Doh! Of course, I took the easy one.

I didn’t add as much flour as suggested and even at that I think I might have added a little too much. It’s a texture thing. I don’t have the feel for it because I haven’t done enough of it. And kneading the dough was awkward until Carol showed me how because it was obvious I remembered eating better than I remembered making.

Waiting on the DOH! to rise took, well, a couple hours. Note to self, fast rising yeast next time. And then forming them into buns and waiting another hour for them to rise and finally, into the oven at 375 for 20 minutes.

Not The Baker

According to the recipe it was supposed to make 24 rolls. I got 17. Which is about what I expected. I remember when I was making cinnamon buns it was supposed to make a dozen, I got a nice big six. Yes, because my idea of size and the book is evidently quite different! They’re thinking something the size of a small tart, I’m thinking something the size of a softball. Different.

In my haste, providing you can actually make DOH! in haste, I neglected to put in the salt. Didn’t seem to make any difference one way or the other. So. Whatever.

The next thing I decided to do is make an old family tradition. We call it a “Trinkle” cake. Basically a white cake base with a brown sugar, butter, coconut icing you put under the broiler to render. The “trinkle” I came from the long grain coconut when I was a kid.

I made the cake and then made the icing and darned if the whole thing didn’t turn out very tasty!

Trinkle Cake

Lastly, about 25 years ago, when I was actively working for a living, “Capt’n Billy”, one of guys I was working with brought in a meatloaf to share with us. This may seem funny, but I was 35 at the time and I’d never tasted meatloaf in my life. My mom NEVER made it. So this was a new experience for me. I got his recipe and we’ve made it numerous times over the years. It’s a little different than most recipes I’ve seen in that it uses saltine crackers.

The other day at the grocery store we spotted some ground buffalo. Buffalo is one of those meats that I’d take any day of the week over beef. The trick with buffalo is that since it’s a very dense but low fat meat, a little goes a long way and you can overcook it because it doesn’t have a lot of fat. Doing a buffalo roast is very challenging to do if you’re feeling adventurous.

We bought a pound of buffalo and I made a small meatloaf with it. I did add some extra oil to the mix because the buffalo doesn’t have a lot of fat. Apparently the buffalo we bought was a couch potato because we actually got some fat off it. Not anywhere near what comes off beef, but more than I expected.

For the recipe I kind of followed my eyes and nose when I was adding stuff. Little of this, little of that. Dr. Fronkenstien sort of mixing. I’d made enough meat loafs I know what goes and what doesn’t.

And the whole thing shrunk down in the pan just as much as a beef one does. Which was quite surprising as well.

BUT it was fabulous tasting. Oh my goodness. Loved it. Something I will make again!

Now there are those who can find six jars of something in the cupboard and turn them into a gourmet meal. My hats off to those few. I don’t aspire to do that but just to get comfy working around in the kitchen. You know, without poisoning myself or setting off the smoke alarm.


Chair Legs

My daughter bought a couple of comfy living room chairs last year and at the time, there was a problem with one of them. One of the seat frames was warped about 1/2″ on the front corner and it caused the front leg to hover off the floor. But she didn’t want to return the defective one. Because. Why? Because. Now quit asking.

I suggested something like this as the simple solution:

IMG 4288

She evidently didn’t go for that “look” and asked me today to “fix” it and also test move a sofa. Yea, how many people you know want to “test move” a sofa. All we wanted to do with the sofa was see if we could manage to get it down stairs in her townhouse. We didn’t carry it all the way, but I think it will go. So, sofa, so good.

IMG 4276

The chair is made from something that closely resembles wood. At one time. Perhaps many glued and stapled pieces thereof. Perhaps a wood recovery project someplace. The legs are probably the strongest part of the chair since they are solid.

IMG 4277

There are threaded inserts in the legs for the bolts to cinch into. And for the most part, they seem sturdy enough.

I used my Festool table (it’s flat) and set 4 pop cans on it to act as “legs”. I wanted to see what the gap was in the left front corner. You can see I easily stuck a pencil in the gap and there was still room. Thus, almost 1/2″ as I originally measured.

IMG 4283

As I was pondering how to level the legs without sawing them and then refinishing them, the wife noticed that the legs had plastic buttons underneath. Probably to stop marring on the floor. Although I doubt that would work considering the quality of the buttons. But they were fairly thick.

Grabbing them with some slip joint pliers and giving a quick tug dropped three legs by almost 1/4″. I then inserted all the legs in the chair bottom, put some weight on the chair and used a small impact drill to set the bolts for the chairs lowest and most level position.

No, it’s not a perfect fix, but it sure beats cutting and sawing, staining and finishing. Besides the chair wasn’t all that expensive to start with. On the bench, the end fix is easily seen. The buttons removed on three legs, felt pads installed on ALL legs and it sits level. Least it will be less stress on the, dare I call it that, frame? What ever.

IMG 4284

But once on the floor, it’s not all that bad:

IMG 4285

Mission accomplished…although I still think the original fix I suggested is more “artistic”.



The single most addicting thing I have on my work bench is…wait for it…an Arduino.

MEGA, UNO, and MINIPro just unleash the TinkerToy complex in me. I feel like Simon Bar Sinister. That’ll give you something to mull over.

Since I’m an avid golfer, I’m also a sucker for weather stations. I have two AcuRite’s, and recently just bought a third, er I mean turd. Say what?

That’s right. As far as I’m concerned, it’s a turd. But it’s a trending turd I’m seeing so…The one I purchased is this one, ACU-RITE® 0135CCDI:


Comes with AC adapter, screen is always on so you can see it (you can turn it off), according to the package it connects to your “pc” and you can read the data on your computer or your iPad, iPhone. Since the fine print required my glasses that I did not bring, I couldn’t read the print that said, “Windows”. Don’t you hate fine print.

Thus, I don’t have “Windows”. I have five Macs. At any rate, I’ll explain how the “magic” works, because I downloaded and read the manual before becoming throughly disgusted and returning it.

AcuLink System Requirements: PC with always-on Internet. Mobile device with app access (optional).

You need Windows, the main weather station connects to Windows via USB. Your computer must be POWERED ON 24/7 to read the data and then sent to a web site (maybe AcuRite’s web site, I didn’t bother looking) where you sign up for an account. When you want to read the data with your iPad or iPhone, you log onto the “web” page and read the data.

So apart from the fact you need a Windows computer, on 24/7, you don’t actually connect to your own LAN to read the data with a mobile device. You wander out on the world wide wait to do it.

Ask yourself WHY would they do this?

The answer I think is, cost. They can hire any number of minions (apologies to Gru) who can write Win code to parse the data or just send it to the remote server. There after the data is stored in some SQL database. When you connect to said web server, the host then serves your data.

Because first off, your bandwidth doesn’t cost them any money and second everyone knows leaving a computer on 24/7 is energy responsible.

I can think of any number of ways that this would have been better implemented, apparently AcuRite doesn’t have that forward vision. Not exactly a progressive bunch of thinkers there.

If at some point in the future AcuRite decides they can charge you to use “their” service, hey, you supported them by purchasing their product, so like any good politician they’ll assume they have the mandate to do whatever they want.

An energy conscious person might decide that leaving a PC on 24/7 is not the best use of energy nor is it free. AcuRite has the solution for this in their AcuLink Internet Bridge. Which they sell you for about the same price as the weather station. Yep, requires an always on internet, still sends it to their site. Big whoop.

The paranoid may like the idea of Acurite handling the data since they can send you “weather alerts”. Because you know that the AcuRite weather alerts are going to be more accurate than anything you hear on the local TV or radio stations. Big whoop x 2.

As I said, this is a trend I see in a number of new devices that I don’t like. No, not the fact you need Windows, the fact that you need an always on internet connection and someone else manages your data. Maybe their idea is to use your data to create the most accurate and awesome weather station the world has ever seen. That’d be the control fetish of the century.

There’s been a number of people who have successfully decoded the Manchester encoding from various weather stations and as far as I know AcuRite doesn’t publish their protocol. You know, because of that control fetish they have. I briefly considered going that route and getting a station that had been decoded. But where’s the fun in that.

So out came the Arduino MEGA2560, DS3231 RTC, a BMP085 barometer and a ColdTears 5″ TFT display with matching shield. This is NOT an economical project. Matter of fact I wouldn’t really advise anyone other than a dyed in the wool tinkerer with doing it.

Having said that, I did it any way because I have that many spare pieces in my parts bins. The DS3231 and BMP085 are standard I2C devices and I have them wired together and they play nice together.

I added one of my Bluetooth HC-05’s so I can connect to the station and do minor programming changes on it (like setting the clock/date or something), or get a CVS listing of the previous 24 hour pressure readings.

IMG 4262

The Coldtears 5″ display is 800 x 400, and the two reasons I use it is that first it’s BIG, second is that it has a font IC in flash ram on board with eight or so different fonts. So I don’t need any software fonts, I just use the built in ones.

So this is what the display looks like:

IMG 4259

I hard coded the lat and long for the sunrise, sunset calculation. I record inHG/mBars from 900 to 1100. If it goes above or below those, well, that should prove to be interesting. I save the full results from the past 24hrs, and every three hours I check for rising, falling or steady. The icon at the top right will change based on the rate of change in that three hour span.

The graphics are stored on a 1GB MicroSD card on the display.

I’m still trying to figure out weather “prediction”. Acurite seems to have some small degree of success with that in spite of the fact that it is nothing more than a best guess.

And while I’m taking AcuRite to the task, I have two identical AcuRite units outside here. One is in the front yard, one is in the back. They are 4 feet of elevation different. Neither one agrees with the other on barometric pressure. The one in the front reads 29.91, the one in the back reads 29.97, my BMP085 is reading 29.86.

I suspect it has something to do with the altitude correction. There is no way to set the altitude on these barometers so I think it takes a flaming guess at it and “adjusts” itself based on the temp or carrot production. Some where. Is it right? No idea.

What I found was you really want to see the trend and having a weather prediction would be nice but not critical. And to put a nail in that, according to my back yard barometer it’s going to rain, the front yard one says mix of sun and cloud. Gee, does that sound like the weather report you get on TV? I think they have 60 different words or phrases to describe rain.

Ultimately, if you really want to predict weather with any degree of success, it’s going to take a lot more than a backyard weather station. Like maybe a satellite photo or two. Or wait a few years until enough people use AcuRite stations so they can do an “Our Man Flint” weather control scheme.


Get the LED out

Sitting around today with a wicked sinus cold, the postie delivered some LED lights, GU10 and MR16’s that I’d ordered from FleaBay. I’d ordered some before but the market has changed slightly on these so it was time to try something new.

What’s new? It’s the word to describe that fluffy white stuff on the ground at Christmas when you have a head cold. Alright, that was Canuck humour. Sorry if you didn’t get it. Or not. Blame it on my sinus…

Scanning through the LED home lighting bulbs on eBay is like wiping your butt with a hoop. There’s simply no end to it. Every seller has a listing that claims they sell the best product. There’s more total Lumens in those listings than the sun emits in a decade or so.

In my opinion, almost every one of these sellers also has an over productive fantasy life. Roughly on par with the honesty you’d expect at a used car lot. Psst, hey sonny, have I got a deal for you…

If you’ve been following my past observations, er rants, you’ll know that I was trying to find an LED light that actually worked to replace those light emitting energy furnaces called Halogens.

Some of you correctly ascertained that I’m an idiot for doing so. And you’d be 100% correct. But since when did being stupid ever stop anyone from trying to reach a goal? Be a lot less YouTube video’s labeled “fail” if that were the case wouldn’t it? Where’s the fun in that?

So I’ll press on here for your enlightenment.

Last year I saw the emergence of 3 WATT CREE (okay, the sellers said they were CREE but you and I know better don’t we). Various sellers jumped on the LED bandwagon and suddenly we had 3, 6, 9, 12 watt “CREE” LED replacement bulbs to fit GU10, MR16 and E27.

The part they kind of glossed over was the fact that these LED’s were all using 3 WATT LED’s and multiple units to give the wattage. In front of the LED’s was a prefocused piece of plastic, to spread out the beam, which they don’t do very well (25 degrees maybe). They were claiming 600-800 lumens. Sure.

The second part they glossed over was that as the wattage of the LED’s went up, so did the length of the bulb. A lot.

A standard halogen is roughly 2″ x 2″. That’s round and length. The LED’s I got all all 2″ round but 5W LED’s are 2.5″ long. If you have a nice holder that has no wiggle room for the extra length (makes you ask yourself why it doesn’t because all the heat should go someplace; halogens are 300C on the bulb surface), these long LED’s are not really an option. What I found is that most ceiling pot lamps have ample room for putting in the 2.5″ LED 5 WATT bulbs. Not a chance with the 12W LED’s.

So what happens when you leave them in for a while? Well, since they do run hot and you might get less than stellar construction:

IMG 4255

What happened to this one? Well, I guess it liked the socket so much that it kind of fused itself in there and then when I tried to pull it out, the three plastic “rivets” simply broke off from being exposed to the heat over time (I’ve had them for almost a year).

So my first advice, apart from the fact this one of the 3 x 3W CREE jobs that runs HOT (54C), don’t EVER try to take one of these out when it’s powered on. That’s a big aluminum heat sink, conductive and you will do the AC short circuit dance if touches the hot terminal. Trust me, your life is worth more than a $9 LED bulb.

Runs hot you say? It’s LED it shouldn’t run hot…well, it does. I grabbed my IR probe and measured a bunch of different LED’s (the multi-CREE ones). 50C to 59C was the range. So yea, hot. No touchy.

The new generation of LED is the COB, or chip on board LED’s. Supposed to be brighter, better beam spread, and run cooler.

Being the LED daredevil that I am, I ordered from two unrelated eBay sellers (actually I have no clue if they’re related or not), first a 5W LED:


And then a 12W uber bright LED:


You’ll notice that the 12W version has a hole through the center. No, I don’t know why. Maybe that’s were the LED baker crawled out. Beam angle spread on these is supposed to be 100+ degrees. Closer to about 70 would be my measurement. Lumens on the 12W are supposed to be 700+. The 5W is 450 lumens. I’m going with the 5W seller because it’s closer based on my light meter and a pair of 50W halogens I replaced with them.

Now, about lumens, or the endless search of. If you have one of those “light meters” from FleaBay and it reads lux and FC, keep in mind that 320-500 LUX is about the same level as average office lighting. So lay your meter down where you want the light and measure it. Don’t stick it up next to the bulb and say, “Wow that’s super bright!”

The further the light has to travel, the more it spreads out and gets darker.

Notice that both bulbs have reflectors inside to spread the light. LED lighting is naturally very focused. Think of a flashlight. How much light actually reflects behind it? Now look at your typical tungsten bulb. The light comes from all over and it’s a source of heat. The light will reflect off the walls, ceiling, floor, grandpas bald head and so on.

Heat? Ah yea. Remember when those CFL environmentally unfriendly things were pimped like the neighbourhood kids ride? Sure you got “light”, and there’s some question about the amount of light, but what you lost was heat. On your typical tungsten bulb, 90% of the energy consumed is converted to heat. Ever stand under a halogen ceiling light? Toasty. And of course, nice in the winter, not so wanted in the summer. So it’s a good thing we don’t need lights as much in the summer. Isn’t it?

And lastly, lets go to “temperature”. Yes there is a lighting temperature. Cool white is 5500-6000K, warm white is 2700-3000K (note that both of these will vary so it’s just for a comparison here). There is a new light that’s about 4000K that they call “natural” light. 4800K is direct sunlight.

All of these LED bulbs tend to vary. The warm will look like a tungsten, the cool will look blueish to you.

I grabbed my IR meter and tested the 5W vs the 12W temps. 5W came in a 43C, the 12W came in at 54C. The 5W reflectors seem MUCH better built than the 12W versions. No, I didn’t rip one apart to verify that. But when I was changing lights I was muscling the 5W versions pretty good and they stood up. The 12W felt less “solid” to me.

We decided to replace the three halogens above the fireplace with three cool white LED’s. The halogens were connected to a Leviton Mural dimmer:

Mural Dimmer

With this dimmer you can set the amount of brightness that the lights will go up to when you switch them on and they will gently fade out when you turn them off. Which brings up some interesting things we found out today…

I’d replaced TWO of the three halogen bulbs with LED’s and then I disconnected the third halogen bulb the two LED’s started to flash about once a second. Looked like an Arduino “blink” sketch. if I reconnected the halogen, the LED’s stopped flashing. Tells me that the dimmer circuit needs to detect some feedback current to keep itself off, or it “fires”. The LED’s won’t do that. So even when the dimmer is “off” there’s some residual juice being used.

The LED’s I’d installed were NOT the ones than can be dimmed. So perhaps this would have made a difference.

At any rate the halogens were didn’t really do much for the painting on the wall, nor did they spread out far enough to even cover it. The cool white 5W LED’s were far better.

IMG 4253

Evidently the wife has always hated these “dimmer” switches we put in when we rebuilt the kitchen a few years back, so I replaced the other dimmer that ran the kitchen 5″ halogen pot lights. Whoa. First thing, with a standard ON/OFF switch the halogen are significantly brighter.

Which lead me to assume that regardless of the setting on the Leviton dimmers, there is no way to set them for maximum during a fade up. When I checked the Leviton manual, you have to double tap the top to get full brightness, or double tap the bottom to get a fast off. Sheesh.

None the less, no more tappy, and really some of the halogens you want full on any way. Not sure why we even put in dimmers.

I then took two LED’s and put them over the sink (wife said standing under the halogens was toasty on the noggin’). Here’s what they look like:

IMG 4258

The LED’s 5W again, disperse the light (warm white) and are not as focused as the halogens. She’s a happy camper now. Depending, I might change the 5W cool whites over the fireplace to warm white.

The work bench area has nine 12W LED lights now on a track light system, four are pointed down directly over the bench, the others spread around the room. So in the photo, that’s warm white on the right, cool white on the left.

IMG 4251

So what used to use 450 watts of power to light the room, now uses 108 watts. That’s a significant reduction in power. And heat.

Note that all of the photos were taken with no flash. Just the available light from the LED’s. When I checked the bench with my LUX meter, I read from 360 to 490 on the top of the bench. And it seems like ample light, for a change. With the halogens, I didn’t measure the light at the bench, but I can say that I NEVER needed any heat on in the room, even in the winter. With the LED’s I do. So maybe…hum, there’s a trade off.

So here’s my observation, for what it’s worth. If you plan on replacing halogens with LED’s, you need COB LED’s, minimum would be 5 WATT units and plan on installing 3 to 5 of them to give yourself the same working light. Keep in mind that there is a large difference in the comfort factor between cool and warm white. The cool white is more clinical.

I measured some of the other house lighting. A standard 4 foot, 2 tube florescent is 400 lux, three CFL’s are barely 200. All measured at bench height from the light source.

The one thing I haven’t mentioned is warranty. This is a $9 LED bulb from Asia via eBay. Really, you can’t use the word warranty and eBay in the same sentence. Lifespan of the bulb is reported to be 50,000 hours. Which you may treat the same as the report of pigs flying.


The LED…is dead…

Quite some time ago, I removed all the halogen lighting (9 x MR-16 50W’s) from my work bench area and replaced them with “equivalent” LED (3 x 3W) bulbs. The main reason I did this was that the track lighting holders have a safety glass that was falling out, yes, not of just one but on all nine. No doubt the few years of heat started to affect the bulb holders.

That and the heat emanating from nine Halogen bulbs is rather volcanic to work under. I didn’t tan, I fried.

The bulbs I replaced the Halogens with were from eBay, touted as “just as bright as” and under $10 each. Of course I was only momentarily deluded into thinking the LED’s 9 watts would be as bright as a 50 watt Halogen. So if you’re reading this and thinking, yeah they’d be as bright, forget it. Not even in the same country, let alone the same ballpark.

I figure, oh maybe a 20-25 watt Halogen. Heavy on the maybe. I’ve seen 17 watt Par38’s that come “close” to a 80 watt Halogen, but the LED beams are too focused so light doesn’t spread much. LED’s need a good lens set up to get beyond the usual less than 20 degrees of spread. I have seen ones on eBay of late that advertise 120 degree spread. I figure the three LED’s in the bulb must be mounted so they point out sideways.

Any way, tonight one of the LED’s starts to “flicker”. Just like an old incandescent. Wow. And then all the magic leaked out of it. Just trying to get it out of the MR-16 socket destroyed it. Okay, I helped it a little…

In the photo you can see the 3 x 3w LED’s. And a LARGE plastic lens in front of them. And yes, they run dang hot too. Not as hot as Halogen but still hot enough to melt plastic.

IMG 4135

MR-16’s are 12VAC powered so I took the screws out of the socket to see if the power converter in the LED’s was shot. Here’s the two token shots of the power conversion stage:

IMG 4137

IMG 4138

With most of these LED’s the big electrolytic capacitor is the part that fries. Not in this case though. in this case, the heat and probably a bad connection got to the LED and it cooked itself.

IMG 4139

Not hard to see the discolouration on the PC board. I left on the two plastic “tear drop” lens and to the left you can see the LED just sort of melted it’s way right into the plastic. Nice.

After removing the rest of the pieces, here’s the raw board:

IMG 4140

You can see the plastic retainer melted and malformed, the burned area and be aware that these lights are not really all that great. In longevity or brightness. But you don’t get much off eBay for under $10 sometimes.

Safe to say that LED’s aren’t going to replace those heat generating Halogens any time soon.

Now there’s some new LED lighting called COB (chip on board) that offers lower thermal resistance, sports 120 degree spread and a 5 watt version, according to specs is 520 lumens. Which is pretty darn good. I see a 12 watt COB LED sports 700 lumens. Now that’s better but still a 50W MR-16 Halogen puts out 825 lumens initially, 900 at center beam.

Maybe someday they’ll get this worked out. Still, I’d take either one of these over a CFL bulb. Any day.


Reverend Club King 290 – Un-Bigsby

Usually in a blog about a guitar you get to read about a mod that someone does to a guitar to make it, look, sound, or play better. If you’re looking for “that” blog, I wasn’t even blogging when I initially did that mod to make my guitar “play” better. So for a brief moment, we’ll fire up the way-back machine to about 5 years ago.

I was in a local music store (Tom Lee) and at the time the manager who knows my playing style and sound that I was after, pointed me to a couple of Reverend guitars they had. He said, “Try it, I think you’ll like it.” At first glance the thing looked like a Gibson/Epiphone to me. But as I got closer I could see a lot of differences.

I’d never heard of Reverend Guitars, but as it turns out, they make some decent gear. No way I’d part with mine.

Reverend Club King 290

Standard 25.5″ scale, bolt on neck, 12″ radius, P-90 pups, and simple controls. There were two colours available. A black one, and one that looked like a orange traffic cone. I grabbed the orange one first and heard from the store front, “Try the other one.”. Being one to A/B things, I plugged in the orange one and I think a strung traffic cone would have sounded better. So I grabbed the black one next. Whoa. Day and friggin’ night. I even checked the headstock to make sure I had picked up the same guitar. Yep. A Reverend Club King 290.

The bass tone contour control was something I’d never experienced before. The guitar has a single volume and tone control, but the bass contour control goes from treble to heavy bass, smoothly. Apart from the ease of which the guitar played and sounded, the contour control sold me on it.

After the time I was of the mind that I really, really needed a whammy bar. At that time, the 290 had no such option. After emailing Reverend, they gave me the info I needed to find a mounting plate and Bigsby that would fit it. No extra drilling required. Nice. So that’s what I did.

You know how you start off with those great ideas? And they never actually work out? This was one of them. Sure a Bigsby whammy works, but it’s a far cry from a Fender Jazzmaster whammy. Thus I found myself simply not using it. Plus it made the guitar a little heavier. No, not as heavy as my Godin xtSA (roughly 3lbs lighter than my SUV), but heavier none the less.

So after four years of putting it off, today I decided was “the day”. I was going to put it back to “factory”. Now that’s an original concept for a guitarist… Here’s the token shot of the start:

IMG 4014

First, I took the strings off. I checked and I hadn’t actually changed those strings in three years. Yeah, I know what you’re thinking. Must have sounded like crap. Actually no. Those are flat wounds, not round wounds. They don’t change much in sound, are easy on the fingers and last forever it seems like. I don’t play rounds except on my acoustic’s and my baritone.

By the way, if you’re doing any amount of string changing, get yourself one of those little battery powered string winders. Slick.

IMG 4015

After removing the strings, I took a shot of the Bigsby and the Vibra Mate it sits on:

IMG 4017

What’s a Vibra Mate? Essentially it’s a custom mounting plate made for a Bigsby so you don’t have to drill any special mounting holes in your guitar. Check them out:

Vibra Mate

You can see the four screws that hold the Bigsby to the Vibra Mate (and don’t go into your guitar underneath). Removing the Bigsby leaves the plate exposed:

IMG 4018

It’s a slick way of doing it as you can see and underneath the plate there are felt feet so the plate never contacts the body at all. I did order a roller bridge with the Bigsby but that was it.

Unlike my normal mode of operation, I’d saved all the parts when I installed the Bigsby so it was a simple matter of putting them back in.

IMG 4020

I also removed the roller bridge and put the stock bridge back on. So now I’m going to say something profound like the sound is completely different. Ah, nope. It doesn’t sound a bit different other than the fact it has new strings on it and feels slightly lighter. Within 24hrs it’ll sound pretty much the same as the last set of flats on it did.

It took longer to get the strings off and restring the guitar than it did to remove the whole Bigsby setup. I was toying with the idea of selling the Bigsby and plate, but you know what? Maybe someday in the distant future I’m going to wish I had a whammy bar on my Club King so I’m going to keep it.

For now, the guitar is hanging up back in the studio and it’s still one of my favourites.

IMG 4021


Arduino Programming

I tend to read the Arduino forum on a semi-regular basis. As such, I find that there’s a lot of interest in the Arduino from people who obviously don’t have the first clue about what it takes to write software, let alone wire up an Arduino. “Blink” is about their limit.

Interestingly enough, a lot of interest is from “school”. Someone has a class project and stumbles across an Arduino and jumps in. To the deep end. Without a life line. And they can’t swim.

Thus a common thread runs something like, “I wanna do this or that and I need help on how to do it”. Generally speaking the “this” or “that” is so vague anyone short of the Great Carnac can in no way answer the question. To most of their credit, the savvy of the Arduino forum do give it their best shot anyway sometimes, but also they don’t suffer fools all that well should it become obvious the person is way out of their depth.

As an Arduino newbie myself not all that long ago and in spite of the fact I’ve been writing software since 1978, I found the documentation for the Arduino typical of a lot of “open source” material. In my opinion, about two cuts above pathetic. Not to mention the scads of libraries and variations thereof that can drive you nuts and poorly or not documented as well.

I was looking through the forum messages and I spotted one about using a temp sensor to control one of those PWM (pulse width modulation) computer fans. Since I’d just finished building one for myself, I decided to read the topic.

And there, it all it’s glory was the best answer I’d ever seen. You might think it’s a little nasty, but really some people actually want you to do all the work for them rather than Google and educate themselves on how it could be done (there’s always examples for Arduino stuff).

So this had me laughing so hard I hurt…

Arduino Truth


Renovations – The Good, Bad and Ugly

Nothing says “renovation” like a house full of drywall dust, floors that look like Hannibals Elephants just marched through and that “lighter than air” feeling in your bank account…

We recently had a bit of a plumbing leak in the ensuite bathroom and decided that it was about time to give it a face lift. We’d had a few quotes and decided on a contractor to do the work. And as anyone who has ever hired a contractor, you know you’re at their mercy and you really hope they aren’t just a “general” contractor.

We started with a 23 year old bathroom:

IMG 1395

The view into the room:

IMG 1396

Which, about 8 minutes after a crew arrives sudden turns into this:

IMG 3686

Ever watch Mike Holmes on TV when he “makes it right”? That’s the guy you want. Although you probably don’t have his budget. In reality, you get what you pay for. It’s not Mike. Or even a remote descendent.

Any way, after two months, we ended up with a look like this (tiled shower, new fixtures, 12mm glass door):

IMG 3878

New tiles around the tub and fixtures:

IMG 3881

And then new tile floor, countertop and cabinetry:

IMG 3880

So the view each morning is this:

IMG 3882

Admittedly I did some of the work myself (and fixed some of the contractor oversights; and wish I’d caught more), I did the baseboard heater, mirror frame, the baseboards and countertop.

The single annoying thing is the floor tile is not flat. How you can lay floor tile and notice it’s not flat is beyond me. Take a straight edge, lay the tile, and it doesn’t have to be level, it has to be flat. A no brainer. Apparently the helper was a no-brainer.

So for all you homeowners that are wondering if you should hire a contractor, my advice is… only use their services for what they are good at. And trust me, it will NOT be everything. If there’s tile involved, ask the tile store to recommend someone. They’ll always have a good idea of who knows their stuff and who doesn’t.

If there’s plumbing involved, get a real licensed plumber to do the work.

For hardwood, be sure you use a guy who does nothing else but hardwood floors.

Drywall is probably within the capabilities of the contractor. Not sure I’d advise trusting them with more than that unless they have professional people they use. Contractors are always good at demolition.

So apart from the floor for our renovation, everything is acceptable. Notice I didn’t say stellar, or, outstanding. It’s no doubt better than an untrained homeowner could do, but it’s just as far from what a pro could do. Yes, the price will represent that.

As they say, the devil is in the details.

The odd part? The contractor seems to feel it’s a good job. We obviously use different measuring sticks.


SwingByte – bites the dust…

Late last year at one of my golf lessons my LPGA instructor showed me a new “gadget” called a Swingbyte. It attaches to the shaft of your club, you pair it up with an iPad (or iPhone) app and it gives you some numbers and records them as you swing (and hit a ball).

After using it for all of one minute we could both see that the way the device is attached to the club was a big fail. It would twist on almost every swing and you’d have to realign and sometimes calibrate it again. So while it worked it was somewhat of a deterrent to use.

If you’re of the mind that these swing “add-ons” are going to give you as much or as accurate information as a Trackman, give your head a shake. A big shake. That’s the same kind of thinking behind buying a driver that will correct all your swing faults. Having the best equipment money can buy is not going to help anyones swing faults. Especially for the long term.

As my instructor pointed out, these are good for one thing. Reference. For example, assume you normally have an open club face at impact. You can make a minor change, re-swing (or better yet swing 3 times), look at the number the Swingbyte gives you. Confirm you’re either going in the right direction or not. Work on it.

I use mine for swing alignment. I.e. reducing the out-in or the in-out path. Do I care if it says I swing 105mph or 85mph? Not in the least. Because in all honesty I have way of knowing if it’s accurate or not. What I could do, if I was working on swing speed is look at the numbers to see if I’m going up or down in speed.

This year SwingByte announced “VERSION 2”. An entire redesign of the shaft “zit”, well really, that’s what it looks like, a new clamping mechanism and an updated app. Plus they’d happily give you a discount if you ordered one sight unseen. That’s called “blind faith”. And at times, as was in this case, misplaced faith.

But hey, for the cost of a couple dozen of Pro-V1’s, I took a chance. In hindsight, even considering I’d eventually lose the Pro V1’s, they would have been a better investment.

So what’d you get for $100 plus shipping?

IMG 3728

It’s kind of a good thing it’s white cause if it was grey it’d look like a slug on the shaft of your club. Some complain that it gets in their line of sight and find it distracting. I didn’t get to use it that long but even so it didn’t bother me.

The new clamping setup, which is promoted to “never” come unclipped, seems like it should do the job.

IMG 3730

In my experience with it, all evidence to the contrary. I’m an 8 handicap, senior and I don’t have a sonic boom swing speed. I had less than 30 hits at the driving range before my Swingbyte broke:

IMG 3732

The clip part that is supposed to wrap around and lock the whole assembly in place is held in place by two aluminum “rivets”. Really. Nice soft malluable aluminum.

When I contacted Swingbyte and supplied them with a photo of the less than stellar parts, they admitted there was a problem with some of the assembly runs and the rivets were not “seated” correctly.

The second annoying part is that the USB cover on the end of the Swingbyte pops open on every hit. It needed to be made of rubber for a better friction fit not the El Cheapo plastic that, luckily, is tethered, but still pops off.

Last and a big fail is the LED lights in the SwingByte itself. There’s a pair of them in there that light up for power and a good connection. if you live in a climate that has any amount of sunlight forget trying to see them. At least without cupping your hand around them like you’re trying to light a butt in the wind. I figure during the testing phase of these, it was winter and dark. In which case they’d look fine.

Swingbyte was very helpful in correcting my unit. I was offered the option to send it back, if I paid for shipping (apparently you have to reward shoddy engineering) or a replacement clamp. I thought long and hard about it because I really could use a couple dozen Pro V1’s. In the end, I decided that I’d take the replacement clamp, if that didn’t plan out and I have no expectations about it doing so, I’d design my own with some of the HandyMan’s best friend. Yep. Duct tape…

On the down side, as quick as Swingbyte is to mail me the replacement part, I do believe they sent it by outrigger canoe. Five weeks to get a replacement part tells me they either don’t have any, they had so many failures they ran out of spare parts, or it really did come by barge or something.

So I’ve used it since I repaired it right? No. It’s full swing (no pun intended) golf season right now. I tend to practice in the off season and that’s when this gizmo will see the light of day again. And maybe by then I’ll have my own clamp setup made.

My advice to those thinking that a Swingbyte will cure all manner of ails, you have two choices.

Buy some more golf balls.


Sign up for some golf lessons where the instructor has PGA accredited credentials after their name. That right. If you’re going to invest in golf lessons, you want to use a PGA (or CPGA) professional. You don’t want a club fitter, or CGTF or USGTF. If you’re going to spend your money wisely, plan on spending some time looking for a PGA professional instructor who is more interested in your game than you are…


Cuisinart Popcorn Maker CPM-900C

Cuisinart has been making kitchen appliances for a long, long time. Founded in 1971, bankrupt in 1989, bought and currently run by Conair I believe.

Popcorn has been around a little longer. Since 3600 BCE or so. Prior to Cuisinart any way. We’ve been making this popcorn stuff a long time.

We used to have one of those “hot air” popcorn makers and it’s the only air one we’ve ever owned that worked perfectly, every time. No doubt a manufacturing fluke. When the “hot air” model finally melted itself into nothingness, we went through a LOT of other similar models to find one that worked. The sad fact is they were better at blowing popcorn across the kitchen and failing to pop than anything else.

We tried a home theatre model, one of those “kettle” corn types. It worked no better and was a brute to clean. It didn’t stay here either.

Hot oil poppers still looked like the best option in spite of the health warnings. We purchased several of those and finally ended up with a Cuisinart Model # CPM-900C. It worked perfectly. For almost one year. Then one clip that holds the heating plate onto the base broke. No big deal, I just held it and popped corn.

Last week, the second clip broke. Hummm…I see a pattern. A quick Google search for this popper with broken clips proved someone at the engineering department at Cuisinart deserves to be fried in hot oil.

None of the stores carry replacement parts, and I suspect a MAIN part like this would be more than the cost of the popper once you add in the shipping and handling charges. And really, you can’t use the popper unless you hold it down in place to ensure the heating portion makes good contact with the electrical connection.

Now, I’m not an engineer, but I figure my makeshift work around is about on par with the original designers capability. So there’d be two of us that aren’t engineers.

IMG 3614

That’s the MacGyver in me. Two mini-clamps, a pyrex cake dish (allows the clamps to clip under the unit and gives air circulation), and presto. Opps, can’t say Presto, that’s another load of hot air…

Actually the design looks a little space agey, the clamp colours even match… wouldn’t be a bit surprised if Cuisinart didn’t copy it.

In the meantime, I only have one word for Cuisinart: FAIL.


Samsung…out of tune..

There’s a large body of electronics providers in the world, but for some reason, I’ve never owned a Samsung product. I haven’t avoided them or anything like that, it’s just that they’ve never made anything that impressed me enough to buy it.

When our aging HP monochrome laser printer sputtered out it’s last page, it was time to start looking for a replacement. While HP used to be the company whose name was synonymous with “laser”, that, unfortunately, was a long time ago. The last few years to stay in business they’ve been churning out economical junk to stay alive in the market. So HP was out of the question (I seriously think they have a problem on their jet direct ethernet cards because those have failed more often than anything else).

As I ran through the usual quagmire of companies, Ricoh, Xerox, Lexmark, Brother I stopped looking at price and started to look at “pages per month” and “toner cart” capacity. Consumables are every bit as important as the printer. For example, if you take a 3,000 page toner cart, work in a fuser, then add in a drum, all at “X” number of copies, what used to be a reasonable cost suddenly doesn’t look so good any more.

Specifications are also “tricky” to read. For example, take the boast of many companies like “20,000 pages per month“. Sounds great. Until you read through the documents for the printer and then, it’s more like 1,000 “typical” usage. The “MAX” is 20,000.

Of course this changed when I watched a YouTube review that Samsung put out where they pitted their printer against, as it turned out, to some the exact same other printers on my short list. The printer was a SamSung SCX5639FR:

SCX 5639FR 01 4

The printer is rated to 80,000 pages per month (actual is 1,000 to 1,500 so where the 80K comes from I have no idea). Either way it’s still overkill for anything I’d ever need but the toner cart is rated for 10,000 pages for the high yield 5,000 for the normal cart. For cost, either of these cartridges is very close to the rest of the market.

Since this is an AIO (All in One) you get scanner, copier, printer, AND fax (does anyone still use fax?). On top of that this printer does accounting so you can have authorized users for this printer and you can track their printer usage. Again, not needed for what I wanted it for, but some might need this in a small office.

More features are scan to email, SMB, FTP, and flash thumb drive. Double sided ADF document scanner, 50 pages capacity. Relatively easy to set up on the network, 250 sheet tray, FAST (37ppm; first page up in 10 seconds), QUIET, 600 watts power consumption but goes back to low power state within 15 seconds.

Any problems with the printer and you can have it email you a report. For any or all of the problems it might encounter.

I picked up the printer from a local supplier for less than half of what it retails for. So another bonus.

And that, my readers is the end of the good news for the printer. While the printer is more than capable, the Samsung software is abysmal. A cesspool of settings you have access to, yet the manual is less than minimum (HTML that requires Explorer 6.1.1 for some insane reason; or PDF that’s worse) and a software install that’s probably one of worst install experiences I’ve ever waded through.

I installed the Mac software first. Five pieces of software get installed. No clue what any of them actually did, nor where they were put. I added an IP printer in the System Preferences and away I went. Printing but no clue how to scan anything. I didn’t install the Fax software. That went out with the last ice age. You need to run Apples “Capture” program to scan from the printer and the scan is a kludge at best. It does work, but it’s not intuitive at all.

Next up I tried installing the software on WinXP. No soap. The software flatly refused to acknowledge the printer at the network address I provided. The only way I could get anything to work was to install the software as if the printer was connected (which it was) and then add a printer and use the “have disk” option.

So like anyone who wanted to use ALL of the functions of the printer, I wanted more in-depth info on all of the SMB, FTP, EMAIL, Flash DRIVE for scanning. I then emailed Samsung and remarked that the “6.1.1 Explorer” to read the “manual” was ah…dumb.

Plus if you wanted to scan to email, you used the keypad like a cellphone when you’re texting to enter in the subject, and email address. Say, didn’t texting with 10 digits go out like the last decade. Nope, that’s Samsung innovation.

I hit the Samsung web site (.com not .ca; my bad apparently), use the CHAT option, got an agent immediately, asked my question, got told that because I wasn’t in the USA I needed to talk to someone from Canada. Polite buzz off.

Except the Canadian “chat” line was about as lively as RIM’s last stock option. Dead. Sorry, no chat today. Special. eMail or call us. I choose email.

After a few days, for support I got this back:

“We are sorry; we do not have complete information regarding SMB and FTP setups and also regarding the software key for the host name. We request you contact our phone support number at 1-800-SAMSUNG (1-800-726-7864) in this regards. They will be available from 8:30am – 12:00am midnight (EST). Customer support is available 7 days a week, 365 days of the year.

The Printer supports a USB Flash Drive on the USB port. However, the hard disk drive is not supported by the USB port on the printer. “

O.K.A.Y. If YOU don’t have the complete information for SMB and FTP setups and what’s required, it begs the question, “who does?”. LG? Ricoh? Brother? Xerox?

I happened across some RICOH documentation and it actually helped with setting up the Samsung. Apparently these guys who make these all in ones tend to clone everything from each other. Except Ricoh writes better documentation.

Thus I figured out, scanning to SMB, emailing a PDF scan (set up an address account in the printer it makes it way easier; you can have up to 50), and FTP. All on the Mac by the way.

Maybe Samsung doesn’t think anyone can read, or that their manuals are not worth writing, or worse; that their products are so easy to use you don’t need a manual.

The printer, I rate a solid 9 out of 10.

Samsung technical support and software, I’d rate a solid 2 out of 10. And only two because one, they responded, and two they did provide the answer to one question.

I also tried to update the printer firmware but the software still can’t identify the printer at the IP address I provided. Big surprise.

I’m convinced that Samsung is just plain stupid and doesn’t care a fig about their customers. I have their half baked attempt at an operating manual and a really lame tech support email to prove it. The next time I buy anything Samsung will be…ah… yep that’s right. Never. You shouldn’t reward stupid.


Lights! Camera! Shadow!

I received a nice shiny new Canon A4000IS to replace my aging point and shoot Canon. While my old Canon only sported a 4X optical zoom, 10 megapixels and fairly good macro for closeups, the new A4000 bet it hands down in both areas. Sort of.

Before I start in on that, if you happen to own one of these cameras and you do any indoor shooting at all, you might find your pictures, for some unexplained reason, develop a distinct yellow cast to them. From what I’ve read, this is common for a lot of point and shoot cameras. It comes from the fact that the “AWB” (auto white balance) doesn’t always detect what type of light through yonder room shines.

Outside in daylight, no problem. But inside with CFL’s, LED’s and halogen, I got heaps of yellow. Even using the auto correction in iPhoto made it worse. Here’s an example of “auto” mode.

IMG 0007

Put the camera into PROGRAM mode and select the tungsten light:

IMG 0017

Substantial difference. While you’re in “program mode” if you’re shooting indoors, change the ASA to 200. My experience is that if you leave the ASA in “auto” mode (this thing has more auto modes than a car dealer) Canon loves to ramp up the ASA to…anything between 100 and 800. Quality starts to drop off pretty badly at anything over 400 in my opinion. So I have program mode set to ASA 200.

Also in program mode, you’ll find that the shots at the settings I suggested will always produce a shutter speed of 1/60. Whereas in anything else, it’s all over the place. Judging from the quality of the enlargement, that’s not always a good place either.

I should also mention that I read a LOT of review sites that show the indoor pictures that look “loverly”. They give you the speed, ASA and focal length they used to obtain the image. After spending a whole evening trying to duplicate their settings with my own setup, I came to the probably not so brilliant conclusion that they aren’t using the point and shoot in point and shoot mode at all. They are getting as close to manual as they can and hitting it with studio lights. As if we point and shoot buyers will have that in our camera bag. I now have a healthy disrespect for those so called “reviews”. Fluff. Just fluff.

Any way, I’m less than impressed with Canon’s processor, the Digidiot 4 or 5 or whatever number is supposed to be state the art. More like state of the “ark”. Sigh…

So, back to the topic of macro. The A4000 has what I’d class as a superb macro and it was designed by engineers who should be taken out back and flogged with the camera. When you get into “really” close proximity to your subject (Canon designers take note: this is what MACRO is FOR), there’s a problem. Not outdoors, but in low light.

Take a look at the typical lens position in macro:

Canon A4000

Now imagine something small right in front of the lens:

IMG 0240

HEY! What’s that shadow? It’s the extended lens. I was about 1/2″ from the circuit board. Can you say, DUMB. Sure you can. I did. Did no one notice this during testing? Really? Guess you’re supposed to limit yourself to taking macro images of flowers in direct sunlight. Swell.

Thus, the question became how to light the subject for macro shots. I had an epiphany and thought I’d give some white strip LED’s I have a try. Temp of them is same as daylight (supposedly) but I left it with a tungsten setting any way.

IMG 0246

You can see the depth of field was about “nil” for how close I was. The LED’s have a distinct blue cast to them but since I turned the flash off, that was the end of the main problem.

I am using “Program” mode and I selected the custom “white balance” and let the camera figure out what it was “supposed” to be doing. Surprisingly enough, it did it correctly. Probably a different engineer worked on that part. Like one who knew what he was doing. He probably got a job with Apple now.

IMG 0273

That fixed the LED’s bluish tint in a heartbeat. The lens was about 3/8″ from the circuit board when I took the photo. Again you can see the depth of field and you notice some lens barrel depth of field at that distance. Not a deal breaker to me though.

Unlike the “professional reviews”, I’m going to show you my setup to get that shot. I promise anyone with the same settings (Program/ASA200/Custom White Balance/1M 5000K LED strip) can reproduce this exactly.

IMG 0277

The “professional” set up is on an 8 1/2″ x 11″ piece of paper. I run the 12V LED’s off my bench supply at 12.2Vdc.

The A4000 is now discontinued and replaced with a “touch screen” version. The ELPH 320 HS. The HS doesn’t stand for “holy smoke or hot stuff” either. When I tried the 320HS camera out before I got the A4000, I found I was tapping, dragging, etc about 85% of the time and using the camera 15% of the time. Getting to any settings in the camera is painful too many times. So for Canon and their idea of a “touch screen” interface, I heartedly give them a complete FAIL.

Ultimately, I’m guessing they moved the lens/flash guy out of that area when they realized he was “challenged”. And they moved him to “interface design” to revitalize him. I got news for Canon if they haven’t clued in yet. He’s still challenged.

How long has Canon been in the camera business? With all the advances in electronics and other technology the best Canon can manage is adding WiFi and sporting an interface that mirrors MSDOS 2.0. I’m really going to go out on a limb and hope the guy who failed with the lens/flash/interface doesn’t decide we need the “ANY” key.

You know, “Press ANY key to continue”…