Tidying up the USAXS electronics

Head-on view of the two motor controllers, and some (still) messy wiring.
Figure 4: Head-on view of the two motor controllers, and some (still) messy wiring.

While messes of wires spread over multiple desks allows complete freedom to play, eventually things will have to be put into a shipping-ready package. Here’s what we did with the USAXS electronics…

Partially filled USAXS control DIN rails, without the messy wiring...
Figure 1: A (tidy, unwired) set of control electronics for the USAXS instrument.

You might remember previous images of USAXS electronics, for example that one shown in Figure 1. We now needed to tidy all that up in a neat, shippable package. Fortunately, we’ve used DIN-rail-mountable components wherever we could, so we just needed a framework.

Figure 2: one side of the electronics box with the serial connections and power supplies.
Figure 2: one side of the electronics box with the serial connections and power supplies.

We found several wooden boxes from the 1980s, with a contemporary “rapid prototyping kit” in them, essentially a Meccano for adults. It has profiles, gears, linkages, chains, hinges, sliders, ball-bearings and axles, and was used for making quick rigs for testing and measuring. All components are made from (anodized) aluminium, and have survived time well. So we nicked that, and built, with a bit of help from my colleagues, the supporting framework for the electronics.

Figure 3: The motor controller side of the box with the (yet unpopulated) panels for motor connectors.
Figure 3: The motor controller side of the box with the (yet unpopulated) panels for motor connectors.

Figure 2 and 3 show two angles, one of the power supply and serial interface side, and the second of the motor controller side. I didn’t quite get as far as I wanted before the pictures were taken: the eleven motor connectors have not yet been mounted on the white laser-cut panels on the side.

Head-on view of the two motor controllers, and some (still) messy wiring.
Figure 4: Head-on view of the two motor controllers, and some (still) messy wiring.

Once that is done, I’ll have to try to fit the Raspberry Pi in. This Pi is controlling the instrument through python / jupyter notebooks. Most of the electronics can be reached over RS232 using the ethernet-to-RSXXX converter, but unfortunately there’s something amiss with the RS485 connection: I can’t get bidirectional communication going with the two motor controllers, something’s wrong with the signal levels. A temporary fix (nothing lasts as long…) is to use the USB connections, but it’s not nearly as neat.

The wiring isn’t quite so tidy yet, but I think that’ll improve once the connectors are on their panels. Some heavy tie-wrapping should also improve matters. So far, though: progress!

 

Share

Be the first to comment

Leave a Reply

Your email address will not be published.


*



*