Trouble Shooting Universal Stepper Controller with Gecko Interface

Thanks Jon. I’m just trying to take a good look at everything since I’m really getting stumped.

I'm close to 100% sure the encoder works, because if I take the encoder that isn’t working and plug it into a drive that is working, it works as expected. I’ve also tried a brand new encoder and behaves the same way.

I guess I can be sure it’s not a configuration issue since, running the pos diagnostic, it only gets incremented in one direction. So, it seems like it has to be a wiring issue. The only thing is, I’ve traced all the wires in question and they seem fine (determined by a DVM continuity test).

I suspect that the wires going to the Gecko Interface aren’t causing the issue, because if I unplug the encoders from it, the working drives update positions and the bad ones don’t. So it seems like my focus on the encoder connection to the USC. It seems like for this issue, the two most important wires to check are on the USC P3 pins A0 and B0 (or A3 and B3 if using W). Would you agree?

Thanks for the information about checking the encoders for logic, that was what I had in mind asking if the signals could be tested.

I had a pocket scope, but the problem with pocket sized items is they tend to get easily lost! So will give it a shot with the DVM.

If looking at the encoder logic signals doesn’t narrow it down, I’m thinking of replacing all the encoder wiring for the X drive today.

It's worth noting that I’ve trouble-shot (swapped out or bypassed) everything except for the X connector (and it’s wiring) that plugs into J1 on the Gecko Interface (and for W testing J4) and the Gecko Drive itself. Those wires check out, but, all tests have used that common plug so it is suspect. It seems the Gecko Interface isn’t related to the issue as I previously mentioned but please correct me if I’m wrong. Should I rewire or even solder the J1 connections to the board?

Reading from the USC board itself, I am getting what looks like proper encoder signals for both A and B. When I turn the encoder they go below 1V and above 3.5V. It's still wired for W, so I read them at A3, B3 and Z3.

I only have one DVM, so not sure how I can measure A and B simultaneously.

Oh, and resistance between A and B is read at infinity for all axes. I think possibly because the signal goes through differential encoder/decoders? Or it is too small for my DVM to read.

For tuning servos, I recall you once helped me with a Halscope view. I don't recall exactly how though.

Does it make any sense to plot the signals if we can?

3DTOPO wrote:

Thanks Jon. I’m just trying to take a good look at everything since I’m really getting stumped.

I'm close to 100% sure the encoder works, because if I take the encoder that isn’t working and plug it into a drive that is working, it works as expected. I’ve also tried a brand new encoder and behaves the same way.

I guess I can be sure it’s not a configuration issue since, running the pos diagnostic, it only gets incremented in one direction. So, it seems like it has to be a wiring issue. The only thing is, I’ve traced all the wires in question and they seem fine (determined by a DVM continuity test).

I suspect that the wires going to the Gecko Interface aren’t causing the issue, because if I unplug the encoders from it, the working drives update positions and the bad ones don’t. So it seems like my focus on the encoder connection to the USC. It seems like for this issue, the two most important wires to check are on the USC P3 pins A0 and B0 (or A3 and B3 if using W). Would you agree?

Yes, that's right.  Since you tried a new encoder, then it has to be in the wires.  There could be an internal break in the wires, especially if the flex when the machine moves.  Is the X axis the longest cable in the system?  Possibly capacitive coupling between the wires is causing errors in reading the signals, and a longer cable would make it worse.  You mentioned differential encoders.  Do you have a differential to single-ended converter module between the encoders and the Gecko interface/USC?  Could that be going bad?

Thanks for the information about checking the encoders for logic, that was what I had in mind asking if the signals could be tested.

I had a pocket scope, but the problem with pocket sized items is they tend to get easily lost! So will give it a shot with the DVM.

If looking at the encoder logic signals doesn’t narrow it down, I’m thinking of replacing all the encoder wiring for the X drive today.
 

Yesterday I wired a new encoder with a new cable that did not use a differential, but it behaved the same. Also, when I swap an encoder, the differential is being swapped too, and works on other axes.

I just completely unplugged the Gecko Interfaces and is not needed for trouble shooting which helps simplify things for me.

I decided to move stuff from W back to X. I've had the diagnostic pos command running and noticed that both the X and W was flicking from 0 to +1. Double checked my work, shrugged, then just cut all W wires off the USC. That got rid of the chatter.

Now instead of the encoder just increasing the value, it jumps back and forth between 0 and +1. I swapped the X encoder (including differential) with Z and the Z pos was correct.

But since it is behaving differently, I guess I'm going to try with the new direct cable to new encoder.

3DTOPO wrote:

I decided to move stuff from W back to X. I've had the diagnostic pos command running and noticed that both the X and W was flicking from 0 to +1. Double checked my work, shrugged, then just cut all W wires off the USC. That got rid of the chatter.

Now instead of the encoder just increasing the value, it jumps back and forth between 0 and +1. I swapped the X encoder (including differential) with Z and the Z pos was correct.

But since it is behaving differently, I guess I'm going to try with the new direct cable to new encoder.
 

This jumping between 0 and 1 (or any two adjacent numbers) is a classic indication that one of the A or B signals is not getting through.  With say, A stuck at zero and B toggling up and down, that is a valid quadrature state transition between two adjacent position counts.  This COULD be a defect in the USC board, but could be in the encoder wiring.
Your meter should show which signal is not responding.
Jon

I'll go check the encoder readings again - but I just saw some crosstalk from X to W and nothing is connected to the W encoder (or even step/dir).

I manually moved the X encoder and saw both the X and W pos flickered between 0 and 1. Since nothing is connected to the W, would you think it might be time to buy a new USC?