Z-Axis Randomly losing its mind

Need suggestions on troubleshooting My Z-Axis randomly losing its height. I have cut this file 3 times and it has changed depth all 3 times, but in different places and to different depths. Since it is doing it in different places, that tells me its not the g-code causing it. The only thing that seems to be in common is that it is happening after running for several hours. This file takes 7 hours to run on my machine. For the first 4-6 hours or so....it runs perfectly. X & Y Axis's are dead on.
https://www.cnczone.com/forums/attac...d=389334&stc=1


I am running a Gecko G540, 48v power supply, 420oz steppers, Super PID, Mach3

Do you have a cooling fan for the G540? It could be overheating.

I do not have any type of cooling on the G540 at this time. I plan on building an enclosure for the electronics and already have the fan and heat sink, but have not gotten to it yet. I suppose it could be a heat issue with the G540, but figured it would affect all 3 axis's. I am going to put a fan on it and try running it again this afternoon.

ger21.....I think your on to something with your overheating suggestion. I ran the part again with a fan blowing across my Gecko. The part completed perfectly.

In just some simple testing, I ran the system without any type of cooling. Within 15 minutes, the G540 was over 125 degrees. With a box fan blowing over it.....it varied between 90 & 100 degrees, never exceeding 100. I will keep an eye on it but I think you put me on the right track. Looks like I need to get on my control box with heat sink and fan....SOON.

Thanks a million.

The fact that it runs fine for the first 4 hours is the key. Glad it's working now.

Apparently, I jumped the gun in thinking my problem was resolved. I haven't run the machine much the past few weeks but fired it up to build my electronics enclosure. The problem is not isolated to just the z axis. It has happened on all 3 axis (with a cooling fan and temp probe on the G540 to monitor the temp). I have completely replaced the electronics on this machine with the exception of the PC.

New 48v power supply
New G540
New Super PID
New 420oz steppers
New PC 690 router
New cable from PC to G540
New cables to steppers

Cables and stepper bought from Nate @ Fine Line Automation

I ran a part a few weeks ago. 7 hour run time. Ran perfectly

I ran a part 2 days ago. 10 min run time. About 90% complete when Y axis flipped out.

Ran the same part last night and after about 3 or 4 minutes the X axis flipped out.

Any ideas? I have reached my aggravation factor and cant seem to isolate the problem

Thanks

I had a few similar issues last month. Had a 16hour run...funky things happening.

Here's what I did to correct my sporadic positioning issues:

1. Had to clean and relubricate my rails and curtain rods. Yes it's obvious. But when the machine was powered down, and I turned my ballscrews by hand
I could "feel" the odd bit of binding, reversed and then forward again and the binding would stop. Turned out to be a bit of crap got into my linear bearings.
Tore machine down, cleaned all the grease out of the bearings...pain in the...u get the idea.

I bought some SuperLube from Amazon. The synthetic stuff with PTFE teflon in it. Wow the machine quieted right down doing those long traverses.

And my machine is only three months old..so I guess I needed to clean/ maintain it more often than I was.

2. I had a problem with trochoidal toolpaths. The machine would lose step, even at low feeds of 1500mm/min. Frustrating when I've run toolpaths for 6 hours at a time at 3000mm;/min. I reposted my programs with simple straight stepover moves between cuts and away it went. For the life of me I still can't fathom why it didn't like the "high feed"
toolpaths...you'd think that the easy spiral stepovers would be easy on the machine ballscrews with the gentle motion but...hey whatever works.

Since I did the above last month I've not had any issues...

Worth a try?

I think my issue is more EMI related. The lost step issue is completely random. I can run the same part multiple times and it will lose steps in different places. All axis are clean of debris and move easily. Im currently in the process of building all electronics into metal enclosures and using sheilded or double shielded cabling between all parts.

Just to update and close this thread........seems the issue is solved!!!

I ended up enclosing all electronics in metal enclosures. All cables (including the power cable) are now shielded and grounded on one end. The G540 now has a heat sink and fan. Before the heat sink and fan, temps would easily climb above 120. Now the temps remain below 80. We will see what summertime temps in Alabama cause but I am confident in the cooling.

As far as the EMI issues. I think they played a big part in my issues. I live within 100yds of a cell phone tower and only a mile away from both a weather radar and small airport radar.

Thank you for you assistance and I look forward to new challenges with my CNC.