Complete Shutdown

[Shayne Jenkins]

We have always had ghosts in the machine but this one is a bit too consistent. We cut a flexible PVC all day everyday. On our older table(6' X 16' PRS Alpha circa 2007) we cut border so it has the least downtime. The issue is that this machine will stop and the software will close in the middle of a file. Sometimes it does this while warming up the spindle with the keyboard open. Sometimes mid cut. Sometimes 10 times a day and sometimes not one. Non of the other machines in the building do this. Any thoughts or direction will be appreciated.

[ken_rychlik]

Try pulling the ram out of the pc.. clean, blow off, and re-install.

[steve_g]

Have you tried swapping out computers to see if the issue follows the computer?
SG

[Shayne Jenkins]

This is a brand new pc and I have three more exactly like it so i have not really considered it to be the issue, but I will start looking into that in the morning. Thank you for the new direction.

[jTr]

Just to clarify/eliminate one more possibility:
What is status of control computer after the incident. Still running? Powered down as well?
If so, it's just one of the handy new features of Windows 10.
My brand new win10 drafting laptop likes to do precisely that. No warning. No reason. Online searches indicate this problem has been fixed. I disagree.
Fortunately, I have a much older computer running the shopbot.

jeff

[Shayne Jenkins]

The control computer stays on. The Shopbot software closes out.

[jerry_stanek]

Did you reload the firmware with each computer change.

[Shayne Jenkins]

I have not changed out that PC yet. I have an update but I am unsure how conclusive it is. I was examining the Yaskawa VFD for unrelated issues and found that one of the three phase wires going to the spindle was not properly connected. It was just resting against the terminal. I fixed that and also found that the fan on the back of the VFD was packed full of our CNC debri and was unable to turn. After cleaning that out we have not (yet) experienced this shutdown issue. Could the VFD overheating cause a shutdown?

[Shayne Jenkins]

Well, the software continues to shut down. I will try swapping out the PC today.

[Shayne Jenkins]

If anyone is still reading this. I switched out the control PC with another control PC that normally runs another shopbot PRS Alpha and I still have the shutdown issue.

[dlcw]

I'm wondering if the control box is sending something back to the PC telling the control software to shut down? Seems, with all the troubleshooting you've done, that is the only thing I can think of. See if ShopBot tech support can help to determine if the control box could do this.

[jTr]

x2 what Don said. Call shopbot.
Meanwhile, congratulations, you've eliminated the PC from the list of culprits.

Another note: I have a laptop that now shuts off completely out of the blue. Finally realized what's going on - I was giving it a static shock, which kills it instantaneously. Now have a raw ground wire hanging on the side so this doesn't happen during this dry time of the year. Pretty disgusting to be faced with such a stupid problem on a $1,200 laptop, but it is what it is.

Naturally, I can't help but think of this as it relates to your problem. Due to number of machines you mentioned, I have to believe you've consistently grounded them all, but take a close look at the grounding to be sure nothing's going on there or has been inadvertently altered.

Please, do keep us all posted. Of all the bizzarr-o stuff that can happen, I've not heard of your exact issue in the several years I've participated in this forum.

Good luck,

Jeff

[jerry_stanek]

Again did you reload the firmware

[Shayne Jenkins]

I really appreciate all the feedback. Thank you.

Grounding looks to be the next step. I haven't had this issue in the past so I figured it was grounded properly all these years, but maybe I need to take a closer look. I see that grounding the frame to the 120v is the proper way. Is there an optimal point on the frame to ground from? Close to the cutter? Motors?

We have reloaded the software and firmware many times, but I will give it another go. Doesn't hurt anything.

Another potential I saw on Shopbot's website is "Check if the excess motor cables are looped up inside the control box.". They aren't but I do have excess motor cable looped under the table. Is this a concern for proper communication?

[chiloquinruss]

One more place to check that I haven't seen anyone state is to swap out the cables. I have a heck of a time getting good cables here locally. Most of the time it is related to USB issues but there are lots of cables to run one of these machines. I wish you well, intermittent issues are the absolute worst! Russ

[scottp55]

Do a UT in the command line, and read carefully....then give us the "Packet ET" time...give us a better idea of comm speed?

USB cable next to a major power spindle cable...COULD do it(Lucky it's virtually impossible on Desktop:)

DID have problems my first week as every cable was anally coiled and nested Perfectly:( :(
Messed it up and separated and was back in business.
scott

[Shayne Jenkins]

I have swapped out the usb cable. Should I try any other cables?

[robtown]

I have swapped out the usb cable. Should I try any other cables?
After it "crashes" check the windows event viewer to see if it's logging any errors?

[Shayne Jenkins]

Packet ET is between 15 and 19.

I do not see any error logs in the event viewer.

[scottp55]

So comm is OK. Good....But the "Intermittents" are the worse to diagnose!
Had an HDMI that was "Jiggly" on the monitor and tried 3 different cords...now elastic banded and duct taped...2 years no shutdown(fingers crossed).
Keep a log of what you've tried and done for Support.
Sorry I can't help:(
scott

[robtown]

Packet ET is between 15 and 19.

I do not see any error logs in the event viewer.
Hmmm... as far as I know, there’s no way for the control box to tell the SB3 software to shut down.
Just losing Comm would result in an alert box that gives you the opportunity to retry communication or proceed in demo mode.

[Mike Windsor]

I had a similar problem once wit my older machine . I could air cut the file no problem , but as soon as I started cutting the static would somehow back feed through the electrical ground and mess up my computer . I eventually ran an isolated ground to the Shopbot that wasn't common to the computers ground and the problem stopped

[paul_z]

"I do have excess motor cable looped under the table" If it is all coiled in one direction, it forms an inductor (coil). It would only be a micro henry or so but still a coil. if you take half of the coil, flip it 180 degrees, and lay it on top of the other half, the halves cancel out each other. Per your description of your problem, I don't think a coil is your problem but it would still be good practice to eliminate it as a source of problems.

Be sure that your data and power wiring are as far away from each other as practical avoiding mutual inductance.

I would also recommend that you get a cheap AM pocket radio (with an earphone). Yeah, I know this sounds like a joke; It isn't. Tune the radio to part of the band where you don't hear any station. You now have an extremely sensitive RF sensor. If you experiment with it some, you'll find that the H Field (magnetic) antenna is highly directional.

If you have an intermittent power connection you'll be able to here it and locate it by moving the radio around. If you use a single to three phase inverter, you are likely to hear that as well and will soon be able to determine normal operation.

The radio will also detect static discharges as an audio "snap". You might want to record the audio. If the recording captures a "snap" and everything shuts down, you know what to look for.

I have used this technique to debug an avionics computer. It served as another modality of detecting abnormal operation.

The technique has also been used to detect the operation of electrical welders and plasma cutters at considerable distance even when there were no other indicators.

Paul Z

[Shayne Jenkins]

Hmmm... as far as I know, there’s no way for the control box to tell the SB3 software to shut down.
Just losing Comm would result in an alert box that gives you the opportunity to retry communication or proceed in demo mode.

I have tried hitting the E-Stop and sometimes the software shuts down there also. This leads me to believe the box or software is seeing the E-Stop is being pressed. The E-Stop is mounted on a small wood platform that I have tried shaking without triggering the stop so do not believe that it is a loose wire or faulty stop.

[Shayne Jenkins]

I want to ask something that may seam obvious. Everything in the box is ground to the ground rod in the box and the box is mounted to the frame. All of this should run back to the AC power, right? So it should already be grounded back to the circuit breaker. I just want to be sure that if I am adding a ground it serves a purpose and does not make things worse.

[Shayne Jenkins]

"I do have excess motor cable looped under the table" If it is all coiled in one direction, it forms an inductor (coil). It would only be a micro henry or so but still a coil. if you take half of the coil, flip it 180 degrees, and lay it on top of the other half, the halves cancel out each other. Per your description of your problem, I don't think a coil is your problem but it would still be good practice to eliminate it as a source of problems.

Be sure that your data and power wiring are as far away from each other as practical avoiding mutual inductance.

I would also recommend that you get a cheap AM pocket radio (with an earphone). Yeah, I know this sounds like a joke; It isn't. Tune the radio to part of the band where you don't hear any station. You now have an extremely sensitive RF sensor. If you experiment with it some, you'll find that the H Field (magnetic) antenna is highly directional.

If you have an intermittent power connection you'll be able to here it and locate it by moving the radio around. If you use a single to three phase inverter, you are likely to hear that as well and will soon be able to determine normal operation.

The radio will also detect static discharges as an audio "snap". You might want to record the audio. If the recording captures a "snap" and everything shuts down, you know what to look for.

I have used this technique to debug an avionics computer. It served as another modality of detecting abnormal operation.

The technique has also been used to detect the operation of electrical welders and plasma cutters at considerable distance even when there were no other indicators.

Paul Z

This technique sounds like too much fun to not try even if I can solve this issue without it!

[chiloquinruss]

This technique sounds like too much fun to not try even if I can solve this issue without it!
Ahh, now we're havin' some fun! :D Your my kind of CNC'r! :D
It's kinda like MY: If there ain't no blood, it can't possibly work! :D
I seriously hope you find it though, I just hate intermittent issues, they just suck! Russ

[jTr]

Everything in the box is ground to the ground rod in the box and the box is mounted to the frame. All of this should run back to the AC power, right? So it should already be grounded back to the circuit breaker. I just want to be sure that if I am adding a ground it serves a purpose and does not make things worse.

My previous set up was exactly as you describe, PLUS I had a dedicated 6-8 gauge ground wire on leg of the shopbot's frame. I had lost communication issues regularly at this time of year.

New Building, same setup, but the box is on the wall, no longer in contact with CNC frame. One year solid, no comms issues.

I know you're not getting lost comm's message, but still appears static related.

Key point here is the ground should look like branches on a tree as it terminates at every electronic device involved with running your cnc. If at any time two branches connect to each other, in this case the cnc frame and control box, trouble can arise, as that static charge (about 10,000V) may be inclined to travel through the control system vs. straight to ground. It's confusing, as you could run for months like this with no issue, but this is the time of year when atmosphere is dry, plus the fact you're cutting PVC means you're likely generating far more static than those of us who have fought this cutting strictly wood products.

With your current setup as described, a static charge has the opportunity to jump back through the control box as it travels to ground. By isolating the frame, static discharge will have a path straight to your main building ground, vs through the delicate controls.


All of this should run back to the AC power, right?

Yes, it all goes to the same place, but best to give it a path that completely bypasses your control box and control computer connections. Electricity, like water, flows, and it will take the path of least resistance, so consider that chassis grounding cable as a first line of defense "trench" to dump unwanted charges to the ground vs. spilling over and "soaking" the sensitive controls.

I would suggest removing the control box from the frame - set it on some 2x material on the floor temporarily. Connect a dedicated ground cable from the cnc table leg to your main ground. See if this separation of grounding paths alleviates the issue.

Cheap, fast, easy test.



Jeff

[scottp55]

Just remembered a time when SB3 was totally kicked off.....it was a panic situation when I first started...and hit the Spacebar for a Pause like crazy.
Sb3 just dropped off the screen and had to do an E-stop.
Reoccurred when we had a "sticky" spacebar once also.

Found the thread where Ted Hall explained that they were fooling around with multiple/fast Spacebar hits(5?) would shut down SB3.
See hi lighted section of thread below.

Probably not pertinent now as it was so long ago, but have you swapped out keyboard and mouse just in case, and one of them IS "Sticky", and that feature is still active in your version?
Just mentioning.
scott

[Shayne Jenkins]

Thanks again guys. Jeff's suggestions are giving me some confidence in grounding properly. Will report back soon.

Shayne

[Shayne Jenkins]

Yesterday I unbolted the control box from the frame and set it on some plastic 4x4s to isolate it from the frame. Then I ran a ground wire from one of the legs to the conduit of the incoming 3-phase AC. This morning so far I have had 2 "unexpected return fault" errors. My AM radio should be here today so I will start playing with that soon.

[Shayne Jenkins]

So that has not solved the issue. Looking more closely I have noticed three separate errors. "Unexpected Return Fault", "Stop button hit", and the complete shutdown. I have been able to duplicate these errors by hitting the e-stop. So somehow the control box is getting a signal that the e-stop has been hit even though this is not the case. I tried shaking the e-stop while running a file and I could not get the errors. Today I replace the e-stop with a spare one I had laying around. We will see if that helps.

[Ken Sully]

Not sure if this helps, but I was having an estop error on my PRS Alpha. I tried everything including sending the main board to showboat for them to test. They sent it back saying nothing wrong with it. I put the board back in and same problem. I decided to purchase a new board and problem gone. It was $$$$ but nothing else worked.

[jerry_stanek]

how many shopbots do you have

[Ken Sully]

I only have one. What was interesting is another shop botter was having the same problem. We talked several times. He also ended up purchasing the main control board and his problem also went away

[Shayne Jenkins]

Thanx Ken. That sounds like a great last resort.