Servos faulting out issue

[Thor]

Table layout:
Powermax 85
CandCNC Mp 3500 (iirc??)
Linux command cnc
Dmm tech drives and ac servos
Dual motor gantry on Y axis (A slave)
Precision worm gearboxes and helical rack and pinion drives in X and Y
With the servos I can easily get 1200 ipm rapids across the table, could be higher but step and direction won't keep up.
Single phase power to building, Phase perfect phase converters for 3 phase power to table and other machines.


When I first built the machine 4 years ago it cut thick stuff no problem 65 amp and 85 amp. Now it seems I can run 45 amp finecuts just fine and 65 amp on 10 ga and under with no problem but anything thicker and one of the gantry servo drives will fault out with either an overvoltage or lost phase error.

After talking with dmm tech about it, I installed a line reactor ont he power feeding the servo drives to try and minimize voltage spikes as it seems as of late anyway my incoming single pahse power to the building is running between 245-248vac and the drives fault at 250vac. Again runs fine with 45 amp tips and the 65 amp tip cutting 10ga is fine but one or the other motors faults out trying to cut 1/4" and 3/8"

Talked with Dmm some more, installed some small transformers on logic input power to the gantry drives, cut voltage there down to 140vac with less noise etc. Still same issue.

I'm pulling hairs trying to fix this to run like it should again. I see 3 possible things not sure what to do.
1. Large transformer to cut all power to table down to 220vac, but so far I only see 3 phase buck transformers and they have unbalanced voltage on L1 L2, not sure if thats an issue or not yet.

2. Torch cable is in same energy chain as all other cables going to gantry, though it is 4" wide and I have the torch cable all the way to one side by itself. Maybe time to get another energy chain and airgap the torch cable farther away from encoder and motor cables?? This is not impossible but some what of a job to unthread it and thread in to a new one. But again it worked fine this way back when I first built it running 5/8" thick stuff on 85 amps etc

3. Rip out all the DMM stuff and replace with clearpath servos? Not sure if they would direct swap or if I'd be remachining mounts and shaft couplers to the precision worm gearboxes or what and would be a major overhaul.

Other things I've noticed: When I first built the machine my torch volts would dance all over with no arc, a few months ago I noticed it all of a sudden reads 0 volts and doesn't bounce around at all anymore. I don't know what the voltage to the building was back when it worked fine as never thought to check it, but the power company has said that 248 I've seen lately is within their tolerance.

Any thoughts from the community here? I feel like I'm losing ground on it and getting tired.

[Thor]

After writing it out and then thinking about it some more instead of just being frustrated by it, I think I'm going to try and separate the torch cable out into its own energy chain with a gap between it and the main ones. Seems like the difference that causes the issues is the higher torch voltage of the thicker steel cutting so seems to make sense that its a cable issue. what doesn't make sense is why it has worked as it is for a few years and now it doesn't, but I suppose the cable shield or something could be wearing down from use.

[BTA Plasma]

You can always lower the step pulse and get faster rapids. The step and direction may need a resistor to common for noise. You will find servo and plasma to be a slightly different setup than stepper and plasma. Proper isolation is crucial.

[Thor]

Any resistor recommendation? Maybe noise there is causing my lost phase error.

[weldguy]

If it worked before and not now check that all your grounds are good and not corroded or loose. I have seen issues like that creep into my shop when humidity get high. Separating your torch lead from the motor leads is a good idea to at least see if that solves the problem. If it does and you want all your leads back together again you can then spend some time trying to eliminate the noise but that's a lot of bs to find out its not the issue. Curious what you find works.

[Thor]

I looked in the Dmm Dyn4 driver manual and the inputs for the step and direction signal are opto isolated and look to have a 270 Ohm resistor already installed internally on each input for step and dir.

I'll check grounds and maybe try tuning the servos again to be sure and see if anything gets better. May also try and see if I can do a cut at higher speed that would maybe use a lower tip voltage and see if that improves anything. If that works than I think I would know for sure that is the higher voltage in the torch cable causing the issue.

[Thor]

I removed the torch lead from all cable trays and suspended from ceiling with bungee's. Still immediately had Y1 fault out for Overvoltage.

So I guess I could put the torch lead back in the trays where it was as that wasn't the issue. Going to leave it for now and try re tuning the drives.

--- So tried manually tuning the drive as the DMM tech directions showd a while back. Using sine wave test motion it immediately bangs and faults to a Lost Phase error. Even with the freq at 1 and amplitude at 200, still a bang like its jerking hard and faults out.

Not sure if maybe I should try a different Dyn4 drive or if need all new cables or motor. I emailed Dmm tech support but they seem to be slow at responding.

[robertspark]

have you tried swapping the drive or motor, sometimes you can just unplug for one drive and test the motor on another drive.

if the fault follows you then it's a motor fault.... if not it's a drive fault ...

the motors in my opinion are consumable and don't last for ever

[Thor]

Ok, finally talked with the DMM tech and we started suspecting voltage fluctuations upstream of the plasma when under load. So of course today when I go to monitor the voltage of the generated 3rd leg out of my DP20 phase perfect it basically runs fine except for one fault right away... I managed to get the parts cut that I needed from 1/4" and have to switch back to some thinner stuff for awhile now.

Anyway I had my multi meter set on min/max and watching the L3 to ground, normal about 212 max when faulted of 224

I then switched and measured between L1 and L3, normal 245 and the max I saw with no fault was 258

So we might be on to something. Atleast I know where to start looking when the issue comes up again. I think I’ll look for some voltage data recorders and try and watch all 3 legs from that phase converter for awhile as well.