It's my Xylotex box? New repeatiblity issue? (?)

I'm stuck with this issue I have been having.
It's one I've just been getting while doing nearly the same task I've been doing for the past year. Basically it's just 3d roughing then 65,000 lines of 3d profiling.
But recently I've noticed the machine slows down or almost stop in some areas, and it's not follow the paths correctly.
It's cutting erratically, but it's not obvious enough to say it is until some time passes. I don't hear any stalls, or see anything, but it definitely goes off.

What's even more frustrating is between this I did a small model, maybe 30,00 lines of code, and it cut out fine! Furthermore, after thinking it was a fluke I did a larger file, and the same issue happened on basic roughing, and finishing! It's a brand now 525MW with an intel SSD, new ram, and so forth with 15K latency with my opening 20 windows of random applications.

I don't think it's mechanical.
The small 12" by 9" working area machine runs happily at 120 IPM (I cut at 85 to 95) with it's 6TPI screws with 40 or so accel, backlash isn't an issue, and that leaves the drivers since I just replaced the computer thinking the old one was the issue.

Could my Xylotex Drivers be giving me strange issues other then an obvious and loud stall? Can I get drifting or so forth from them? (I doubt it since it's kept clean, it doesn't get hot, but what do I know?)

I'd appreciate any ideas on what could be wrong before I panic, then buy a G540 with a new PSU, and spend my little bit of profit money I've been saving.

i'd look at the power supply , and thermal temperatures on the driver chips , if it becomes worse the longer it runs then thats the area i'd look at first ,
or fit some pc fans to cool the drivers etc .

Do you have dual shaft steppers? I know the xylotex controllers can have resonance problems.
It may be possible there is a zone between normal running and stalling, where there may be intermittent lost steps.
If you could fit dampers, it would take care of 1 "what if".

It does seem to be repeatable still, but I don't have limit/home switches to be truly sure. I'll jog it to one edge, run something, and see where that goes.

Other have complained about Xylotex using thermally conductive tape, but mine have epoxy. I gave a heat sink a light tug, and now I'm stuck fixing it with arctic alumina, and can't find the epoxy kind around here. (a dab of super glue on the edge with alumina in the center?)
The heatsinks do not get warm, the PSU gets a lot of airflow, and overall thermally things seem good.
I will look for a different parallel port cable, and see if that changes anything.
I've never had issues with resonant frequency recently, I used to but ever since I upgraded to 6 TPI screws I've never had it be an issue.
Perhaps to rule it out I should run it slow enough so that my velocity stays constant, and see what happens.

It does seem to be repeatable still, but I don't have limit/home switches to be truly sure. I'll jog it to one edge, run something, and see where that goes.

This is where I would start, find out if you are losing steps. You can set a dial indicator on your axis before you start your program use the G53 as John suggested. Then run the axis around a bit with g1 and g0 moves and then return to the home position as it should be the same.
Try a few times to make sure it is correct. Then run your program and see if the machine returns to the correct location.

Then as you suggested try a slower speed and test again. Also try adjusting your acceleration.

If the machine does always return to the correct location you are not losing steps and will have to look elsewhere.

Rick G

I lowered my acceleration from 50 to 40 (I thought it was 40 but it wasn't) and the issue seems to be gone.
I am also going to lower my maximum rapid speeds down, and I am taking the machine apart then realigning everything to minimize friction. The machine seems to drift when I really push it and the friction can sort of accumulate.
I've never had this sort of issue (random stalls that aren't easily repeatable with jogging or simply running a program, then seeing an OBVIOUS stall) before with my machine in the past year and several hundred hours I've been using it. I'm rearranging everything in my cramped workspace, finishing a dust shoe, I'm installing some HAL limit switches, finding some correct wire that can be shipped to Hawaii(and a brush strip that ships here), and building a new (less ghetto) enclosure. Hopefully then I can start "production" and then by the end of the year see where I am at when it comers to buying another machine. I think another smaller one with G203Vs and the higher end things would be nice, or perhaps take the plunge with a R&P 4'x4" table.

Anyways...Thanks for the help and tolerating this 20 year old who occasionally jumps the gun too quickly

Glad to hear you got things sorted out.
It is safer to be a little more conservative with acceleration and speed and not have to worry about missing a step in production.

Rick G