Not dead yet

I realize that it has been a year and a half since my last post. I’ve been doing projects, just not documenting them.
Eventually you will hear my endeavors to repair an old slush puppie machine (propane is an excellent though explosive replacement for R-12). See some summaries on my 50 watt Epilog Legend 32 laser cutter rebuild. And see my build and trials involving producing a Rum and Coke Slurpee.

Cleaning my Extruders

My Replicator hasn’t been functioning well lately. I switched back to some older filament and couldn’t get much out of it before the nozzle would stop extruding. I am beginning to believe that this is a common problem that everyone calls “clogging”, but I know that in my case, the drive gear just slips on the filament and it loses pressure.

One of the reasons that I love the Original Makerbot Replicator is that it is incredibly user serviceable. Getting to the drive gear is a piece of cake with a few hex drivers.

Once I got to my drive gear I could see how dirty it was. I’m not sure if it functionally makes a difference, but I suspected that the powdered plastic sitting inside the teeth was decreasing it’s ability to bite and drive.



For those of you unfamiliar with the extruder design, the filament comes into the top left in the following picture and is pressed between a delrin plunger and the toothed gear at the end of a stepper motor. The gear bites into the filament and forces it into the hole that you can see below it. This is the entry into the hot end where the plastic is melted and driven out of a tiny nozzle to make the part.



Using a small wire brush (Harbor Freight has them cheap) I cleaned inside the gear teeth and dusted off the filament guides. You can see in the lower picture that there is a groove running down the face of my plunger. Apparently this is normal wear. However I will be going the route of using a bearing face to press it into the teeth instead. Fortunately Thingiverse has me covered.IMG_1111Please note that on the back side of the plunger there are a number of tiny tiny washers. Four in my case. And I spent an hour on the floor trying to find them.


Spotty Charging on a Kindle

A good friend of mine had a Kindle that refused to charge easily like a good piece of hardware. You could get the charge LED to come on for a second or so if you wiggled the mini USB plug, but otherwise it was a crapshoot to get the unit to charge for any length of time.

Being quite a hassle, she asked me if I could fix it and gave me one of those pretty smiles that I’m such a sucker for.

Enter the Kindle.

The first 5 minutes was me looking around the perimiter trying to find any hidden screws that might hold it together and experimentally wedging my fingernails under the edge looking for pop latches. Once I got one edge to come up, I went around the entire perimeter carefully popping the back cover off.



The USB receptical is located on the underside of the green PCB that makes up the brains of this little unit. I was impressed at the size of the battery on here, but my job was under the PCB. Eight little phillips screws later and some care to remove the plugs and ribbon cables gave me my first look at the connections.

Here is the offending USB port.

Here is the offending USB port.

Some little reflow unit was a bad boy.

Some little reflow unit was a bad boy.


All five of the pins for the USB receptacle had broken their solder joints. This actually happens somewhat commonly in electronics. The amount of heat needed to reflow the solder paste for the tiny components is small compared to the thermal mass of something as big as a plug. I’ve seen a lot of times where the ground pins never get properly soldered to the board and the receptacle is left wobbling until it breaks its own connections. In this case every single pin had come loose. Looking closely, I think the problem might also be that there was not enough solder paste applied to the pads before the component was set on top of it and the solder just couldn’t flow properly as a result. Either way, I found the issue and I knew how to fix it.


The pins were too close together to easily solder separately. I tried. Unfortunately the only tip on my soldering iron that would decently fit onto a pin was too small to carry enough heat to the part to make a good solder joint. I used a trick originally taught to me by Dave Jones of the EEVblog and just soldered all of them at once. This allowed me to use a larger tip to carry enough heat down to the pins and ensure that each one is soldered well and good. There was just one problem.

The Horror!

The Horror!

Now don’t worry. There is this wonderful material called solder wick, and it pretty much does what you think it does. I applied a little amount to the solder blob I just made and put my soldering iron back to the part. Solder wick works by pulling solder into the fine braids of wire through surface tension and capillary action. The beauty of using it in this case was that those same forces also existed between the pads of the PCB and the pins that were sitting on top of them. The result was that most of the solder above the pins was wicked away while the solder under the pins (i.e. the good solder) remained.

Kindle Connection FinishedNow in the interest of full disclosure, I did go back and touch up the pins individually after the solder wick did its job and before that last picture was taken. Once each pin had been properly wetted with solder it responded to my overly large iron tip well enough to carefully confirm that it was free of solder bridges between the pins.

I cleaned off my excess flux (the secret to good soldering) and did another visual check for solder bridges. I reassembled the kindle and confirmed it indeed worked and couldn’t even interrupt charging when I tried.


And I got another pretty smile too.


Makerbot Replicator X Axis Ribbon Cable Upgrade

I’ve been the happy owner of a Makerbot Replicator for almost a year now. But I’ve started running into repeated problems with the X axis motor. Well not the motor itself, but the wiring harness that goes to it. The wires go through some flexing with every movement of the Y axis and with enough flexing, the copper starts to work harden and break. It’s definitely repairable but requires patience to find the break and the finesse to re-solder it and patch it up well enough.

My solution is to use a wire that’s meant to take a few million bends; A Printer’s ribbon cable.

The donor for this project was an old inkjet printer that had been relegated to the garage for the past 2 years. Inside was a 22 conductor ribbon cable, meant to last for thousands of pages and millions of flexes.

The conversion


The Ribbon is essentially flat copper conductors sandwiched between a plastic support, so soldering to the pads on either end is a bit of a challenge. For this attempt, I cut the wire down to 10 conductors and soldered 5 pairs of ends together while pinned flat to my workbench.

Makerbot Ribbon Cable 2




Once everything was soldered together and I amazingly enough did not have to use my spare 5th conductor I taped up the ends of the ribbon cable with some electrical tape to prevent any shorting that might occur.

Makerbot Ribbon Cable 3



The cable seemed to be just the right width to place into the original holders for the wiring harness. I wedged it in place and reinstalled the new hybrid harness. Just a few checks to make sure that I had the proper length that could move and I tucked the remainder of the new and longer cable underneath the machine near the motor drivers.

Makerbot Ribbon Cable 4

Flange Focal Distance Part II

So apparently my theory on Flange Focal Distance being a factor in my focusing problems seemed to be a dud. It turns out that I can hold the camera actually a fair distance away from the flange mount without any observable degradation in image quality.
In hindsight this makes some kind of sense. As the microscope acts as the camera’s lens, the light hitting the sensor must be relatively parallel.

To test this I actually decoupled the camera from it’s mount and shot lens-less from a ways back in a darkened room.

So going from fully attached to several dozen millimeters out changed my image impressively little.


Microscope Focal Distance 2


Microscope Focal Distance 1

Hand Held










Microscope Focal Distance 4

Hand Held

Microscope Focal Distance 3




Flange Focal Distance

I am not a professor of optics.

That being said I do understand that there’s a lot to learn about hooking a camera up to a microscope and I am woefully ignorant of the minutia.

But this is science and it is full of venturing into the unknown, even if many have already figured it out.

So one of the problems I have been having with the Versamet microscope is that everything on the camera was out of focus compared to the eyepiece. I did a little correction with the fine adjustment focus knob to get around the issue. But being limited to a Canon 5D without live view meant I had to focus everything through the eyepiece. Cumbersome and uncomfortable are appropriate words here. This meant that the photo quality suffered.

Then I discovered Flange Focal Distance. It was something I was vaguely aware of but didn’t know it was critical. When light exits the rear of a lens heading for the image sensor (or film), the light rays aren’t actually parallel. Thus placing the sensor closer to or further away from the lens flange than designed can lead to a fuzzy shot.

From Wikipedia

From Wikipedia

Now it seems that the Leica M mount that originally came with the microscope had a focal distance of 27.8 mm while the Canon EF mount tops in at 44.0 mm. There’s enough of a disparity there to make me think that this may be a factor in my images.

Solving this issue should bring my eyepiece focus and camera focus to the same point. Hopefully it will also help my falloff problem and light up the edges of my pictures. Unfortunately with my microscope out on loan I can’t do anything about this at the moment. I’ll provide an update when I get a chance to tackle this problem.

Silicon under the Versamet

So I wanted to offer an update on the Versamet 2 and the lens adapter. It works!
I even managed to take some shots with it before I loaned it out to a friend of mine with a Canon 7D who really wanted to do some digital microscopy.
I was having some issues with the communications contacts for the EF lens mount on the camera. Modern DSLR cameras actually communicate with electronics in the lenses in order to operate some of their features like Image Stabilization, aperture and focus. My metal adapter ring actually shorted some of the contacts making the camera error out. A little bit of sticky tape solved the problem.
The images were also suffering pretty badly from a falloff in light as you left the center of the image. I didn’t have enough time to track down the problem and the best solution I had was to crop out 3/4 of each shot.

But one of the neat shots that I took in a failed photomosaic was this. This is one of the pieces of silicon from a module that was given to me by a gentleman I met and struck up an engineering conversation with. It’s a failed copy of one of the chips that went onto the Cassini spacecraft that is currently orbiting Saturn.

Wire bonds and some alignment marks
I don’t understand enough about Silicon to give a truly accurate description. But this chip specifically has large feature sizes for even a 1994 vintage. It’s a feature called VLI. As I was informed by one of the gentlemen who helped design this chip, the feature size prevents high energy particles from switching transistor gates and possibly creating a latch up failure.

Think of a bowling ball launched at the door of your house. The ball will likely blow your door right in or cause enough damage that it’s essentially open now. Now imagine that same ball crashing into a 50 foot tall blast door. It may do some damage, but you can’t arguably claim that the door is now open.
It might be a brute force tactic, but it works pretty well.