Noi encoder input any more on 7i77?

Well, I've just spent five hours trying to figure this out, and I think it's time to ask for help

I've been working on my 5i25/7i77 combo in my milling machine getting "extras" like SSRs that control air blow off and startup of the spindle drive working. I had to take a break for a week while I waited for some parts. During that time I opened up each encoder's cover (they're amt102v encoders) and changed resolutions from 512 ppr to 2048ppr. I've been wanting to do that to have some more precision for the PID loops to work with.

Apparently I changed something else that I don't remember, because now I get no input from any encoder. I've checked power to the encoders from the 7i77 and it's ok, field power is ok, the enable inputs for the servo drives work, and commands to the drives work. I know this because I've had each axis run away or give me a following error, or in one case both (X ran away in one direction and gave me a following error in the other). I've got FE set to 0.010 for slow moves and 1.010 for rapids.

I've tried manually moving the motors and the encoders in linuxcnc show no movement. I've also tried connecting a different brand of differential encoder to the input for ENC0... still no pulses read.

I'm attaching my configuration along with a dump of my hal pins and the relevant lines from dmesg.txt.

Edit: Forgot to mention, this is a 7i77+5i25 kit, I'm using the base firmware and the cable that came with the kit.










If someone can see something immediately wrong I'm interested to know. Otherwise I'll be back at it tomorrow.

Thanks in advance,
Erik

Things I would check:

To determine if this is a hardware or software problem:

Check encoder 5V at the 7I77 encoder terminals
Check encoder signals at the 7I77 encoder terminals:
Check differential swing (measure between A and /A etc)
and 90 degree phasing between A and B

In linuxcnc, check encoder raw counts with watch or HALMeter

I checked encoder 5v power at the 7i77 encoder terminals... specifically TB3 encoder 0, found 5v.

Check encoder signals... what should I use to do this? I have a single pin logic probe, might be able to get my oscilloscope working though I'm not that good with it.

Differential swing, oscilloscope?

No raw encoder counts are coming in, the count doesn't change in the HAL watch window when I move the motor by hand.

Also, I did connect another (known good but different type) encoder to the terminals on TB3, no counts were showing up there either.

Erik

To measure differential signals you measure between the A and /A pins
and between the B and /B pins. You should be able to jog the axis a bit by hand
to get both A and B in both states. The differential signal should alway be greater
than about 2V

Got it, thanks to your suggestions above.

I decided to try an external 5v power supply just to see if it would help. My encoders don't need much power, but I thought maybe the 5v rail in my PC power supply was flaky or something. The external supply didn't help directly, but I did notice the "power good" LED on the supply was flickering. It came on solid when I disconnected it from the 7i77. So, I theorized a short that somehow wasn't completely killing the 5v power on the card.

I unplugged the Y and Z encoders and tried just X, and it worked again. Added in Y, both still worked. Added Z, nothing worked. Aha!

Turns out my fiddling with the encoders had created a short on the Z axis of some type. I pulled the encoder, and re-wired the connections taking care to ensure that there was no possibility of a short... the shield on the cable I'm using for my encoders may have been the culprit, it's a very fine mesh and may have been pushed around a bit in there. I covered it with tape and reconnected things.

Once that was done, all three axes came back to life in the next test.

I suspect my old meter with its 1.5 hz display update speed was masking some intermittent power issues when I tested encoder power before. It might be time to get more familiar with that O scope or else get a new meter

Thanks,
Erik

Thats the problem with intermittents,
The voltage may well have been 5V when you measured it...
Glad it was nothing serious (Other than being hard to debug)