Soldering robot V2

My first iteration of the soldering machine was based on Prusa MK3 chassis. It worked well enough but the wiring was very messy and I was not happy with it. It was more of an experimental platform than something useful.

For V2 I decided to use an old MendelMax 3 chassis that I had. As 3d printer that was quite outdated, but the motion platform was sturdy and reliable.

I re-used most of the carriage design from the first version. I added a GoPro fixture, so I can mound an LED light on the carriage itself instead of the side of the machine. In theory this should provide more consistent illumination for the camera.

I was also not happy with the fume exhaust design on the V1. It was using a small 40mm fan that was very noisy, the carbon activated filter piece was very small and it looked very ugly.

For the V2 I decided to put the fan and filter on top of the machine, next to the solder wire motor/extruder. It took quite a bit of experimentation to find the right hose to connect the two. I tried thin silicone hose, that was too thin and the walls would collapse during movement. I tried thicker vinyl hose but that was too rigid and would obstruct the movement of the head. Finally I settled on a corrugated PVC hose – it is flexible enough and would not collapse on it own.

The other experiment was to find what size fan I should use. I tried 60mm, 70mm and 80mm axial fans but they would not provide sufficient airflow, when the filter mount was attached. I finally settled on a 7035 centrifugal fan. I started with 120mm centrifugal fan, but that was too big and loud.

Here is a video of the fume exhaust system in action:


TMC2130 – setback

Well, my enthusiasm for the TMC2130 driver was premature. Faith was not kind and I discovered several design flaws after I ordered a bunch of assembled boards.

First I somehow completely missed the 5V power pin to the driver. So initially it was as dead as a door nail. I also discovered one of the config pins was not properly grounded – but that was not a big deal.

After a few bodge wires the driver sill would not respond to SPI commands. This was a rather frustrating thing and I remember having similar experience with the first PrntrBoard with 2130 drivers. Back then I had to wire a separate clock signal to each driver to get them to work.

When reading the datasheet of the 2130 chip it describes that if one connects the clock pin to the ground the chip would use an internal clock generator, but something is amiss and it does not seem to be working correctly in my version.

So back to the drawing board. I updated the PrntrBoard V2 to have a driver clock signal and now have to order new sets of PCBs.

Quite annoying. If you have a clue why the build-in clock is not working on these I would appreciate the hint.