VistaCNC P4-SE pendant stopped working

I just clarified the reproducible steps because there were a few bits of additional information and/or inconsistencies. My keyboard in the machine has a sticky key or two (need to get a different one, just haven't had time) and the ergonomics are terrible for typing longer messages.

Ugh... I just looked at my VC-P4S.HAL file and the test results I provided in my last post were actually with that set of lines physically removed from the file. I've got a bunch of different directories filled with experiments (different HAL and INI tests), and I pulled from the wrong one. Shoot.

When I insert block #13 back in, I get slightly nicer behavior (pendant understand the difference between machine off and estop) but not appreciably different experiences. The fault still occurs after I press the computer console's machine on button and then I press my physical "start" button. Essentially, then, the problem occurs once COR and Vacc are latched, even through I don't see that the control should really care about that.

When I remove both #13 and #14 (so none of the code shown below), I still get the same behavior... When I press the START button, comms with the pendant cease...

Since I have a config setup for a new aux spindle that is not installed yet I'm going to play with my e-stop and see if I can reproduce the issue as you have. In my case I have the same pendant with the EE-box and the issue seems to be random. I will press the start button, click OK if prompted for the tool change and then stand up and pickup the PS4 in case I need to hit E-Stop and it with got to LINUXCNC or garbage on the screen. Other times I pickup the PS4 to jog the machine or so some manual machining and it goes to LINUXCNC or garbage.

John

. I tried, but I am not currently using the 15' repeater cable they supplied.

John,
I didn't use the repeater cable, as I think I may have mentioned in a previous posting. I discussed this with VistaCNC and I think it is a non-issue. First, I have the problem whether the repeater is in the system or not (temporarily installed it). And second, my EEBox is connected directly to the computer's usb port with a 3' or so cable. It is the same cable that I would otherwise have connected to the repeater, so I think you can successfully argue that the repeater would be an inferior addition, in my case.

I need to check my wiring. I've played around with different times I press the START button and more often than not, it introduces the comm fault. That simply does not make sense... Most of the time, problems like this are self-inflicted. I'm human, which is to say that I probably buggered something up! I'm not proud... I can admit that I've done something wrong; I know I make errors and everything seems to point to that here too. Driving me NUTS!

Alan

You should really check the grounding of the machine and computer.
Don't know if i mentioned it before, but that should be the first step when having such issues.

Grounding of the computer.... well, that is an interesting one, in that I'm not entirely sure how to check that or where to perform it. My cabinet is an industrial one that has an internal mounting plate. My 'computer' is a motherboard that is mounted the the plate (using standoffs in the correct locations, of course). The power supply was mechanically mounted to the same plate. Finally, an American style 120VAC plug was also mounted to the sheet. The power to the latter is switched elsewhere in the circuit. Grounding for the plug goes to a central point (although American plugs also ground through the metal box, which as I said, is mechanically grounded to the sheet.

I don't know of a normal place to attach a ground to a computer. Normally it gets its ground through the plug.

You may be onto something though.... I pulled out my COR relay, which means that the Vacc, and anything it supplies, would not be relevant. When I press the START button then the pendant freezes (this is a newer discovery). I'm going to go back and try that test with the simulator when I get a chance... But it definitely seems like a pattern of negative interaction with the START switch is at play here. I've confirmed the wiring but will do so again.

The start and estop circuits are provided as JPG attachments in an earlier post.

Update:
You may be onto something. I got the simulator to hang once when I pressed START (no COR relay in place). But then it all started working. I went back to the setup I had been using last night, where the failure after pressing START was 100% predictable, and now pressing START has no negative effect. Argh! !%^&^^#@#!@#@$% This is the stuff that can drive someone nuts! Spurious problems suck. That said, grounding issues are right up the alley for problems like that.

Power supply design:
I should describe power supply design... I'm pulling all of my low voltage power from the computer power supply. The only (indirect) exception is my 24VDC, which is a small buck board connected to the power supply's 12V. No particular reason for posting the PS information other than it may twiddle someone's thoughts...

I think I may have found the problem, although it seems a little odd. When I reinstall the COR relay, I can cause the pendant bricking 100% of the time in the simulator by pressing ESTOP. The interesting thing is that the CTL-STOP contacts are open in the simulator case because that configuration doesn't know anything about my hardware.

With the real machine (which does energize CTL-STOP, so pressing START *will* energize COR, I could always cause the fault, of course.

Armed with that information, I got to thinking more about my work last year (I still would swear I had tested the machine with the pendant), and it occurred to me that there was a minor difference in my setup, innocuous though it may seem... Because I didn't have the EEBox (basically a breakout box) permanently affixed, I really was only using the pendant's eStop switch as a software interpreted switch. The second pole did not break the electrical latch chain shown in my schematic....

I decided to replicate that case by removing the wires to the pendant eStop, and installing a jumper on the terminal block to simulate the switch otherwise having been closed. Eureka! The pendant doesn't seems to brick when I press START, and the relays all latch as you would expect, etc.

I am not at the point where I want to concretely claim that this new development is absolutely repeatable, etc. I've been burned by stating things like that in the past... But it sure does seem promising (and suspicious) to me... My guess is that there must be some crosstalk or noise that is adversely affecting either the power or the differential data lines within the pendant cable. I've contacted VistaCNC support, and hopefully they have some ideas. Obviously I want to have an electrical estop chain for the previously mentioned safety reasons.

Anyhow, I'm thrilled with this development. I was getting very frustrated. I hope this is the nail in the coffin for the cause, and now I just have to find the solution.

Alan