View Full Version : Erratic, maybe static behavior

02-14-2014, 10:21 AM
I have a motor issue where I startup my machine by powering up my box (PMDX-126 with the usual Nema 34 package from CNC Router Parts) and everything powers up great and talks to the computer but the motors automatically disable themselves. If I am to turn the power off on my Nema case which powers everything down on the machine side and power it up right away again - the motors will then work for the majority of the day. If you listen to the resistance of the motors when they engage they make a loud pop (not loud enough for me to be concerned about wear) on their first engagement when I initialy power up. On my second power up (because I do this idiot carnival ride every day now) they make a much softer engagement tone and then they work flawlessly.

All of this seems to point me in the direction of a faulty ground, something is grounded out, ground loop... Ground ground ground. Does anyone else have something savvy for me to try??? :) So far I have replaced the PMDX board that connects all my geko drives to the PMDX 126 - I have wired them manually twice, and replaced the wiring since with another PMDX board to see if my wiring was bad. All to no avail. I am wondering if there is some sort of static build up and it is discharging on startup - thus the loud engagement sound. Then on second startup it works because there is no static. Does this sound right? Also, how would I fix this? Considering Earth ground and all that - I'm smoked for ideas right now other than running a separate ground rod for the machine itself.

02-14-2014, 10:35 AM
At what point relative to powering on the NEMA electronics is the computer started up? Is Mach3 running when the Nema electronics are powered up?

02-14-2014, 10:45 AM
I would say half and half - this problem has been going on for about 5 to 6 months. It has been getting worse though, maybe just getting more annoying than anything else really but I don't want it to be a part of a larger problem.

02-14-2014, 10:50 AM
I'd ask Steve at PMDX. Sounds like maybe a charge pump issue?

02-14-2014, 11:04 AM
Not sure, the watchdog message comes up every now and then (only the motors are disabled at these times usually) at the time I thought that was more of a smoothstepper issue. Obviously these are connected issues. I changed my settings to 3.1 seconds which helped during operation greatly. However, everytime I start the machine in the morning I have to turn it over two times like a bad car. Really strange. Of course, it needs to be re-set once or twice threw the day. During operation it almost never stops itself but will shut the motors down between runs. So when I go to move the axis again Mach will say the machine is moving (digital encoders moving) but the actual axis are stopped. Occasionaly (more often now, with no changes to the machine) the watchdog message will come up and I have to restart Mach as well.

02-14-2014, 11:11 AM
Is this a USB smoothstepper? If so, they can be very sensitive to noise.

It seems strange that you see watchdog errors. These I think are normally caused by the parallel port driver, which you're not using. Try disabling watchdogs in General Config. You shouldn't be having Mach3 stability issues with a Smoothstepper, unless perhaps it's the Smoothstepper that's causing them.

02-14-2014, 11:29 AM
Its Ethernet Smooth Stepper. I could finalize the settings and put in MAC Address and such. I was pretty anal about shielded cables and such - EMI was a consideration when I put the box together. I do have two IR switches in my box, but when they are enabled they do not seem to change anything as these are only enabled during operation. The watchdog comes up occasionaly... Should I disable it with Ethernet Smooth Stepper? Why not... LOL I'll try anything right now. Honestly, I'm not sure how much good it does for me anyway.

02-14-2014, 11:56 AM
Okay, there was an update from Warp9 for their Ethernet Smooth Stepper and I got that installed and working in Mach3. Supposedly that had a bug fix for watchdog timeouts so that "should" fix that problem. Good eye Ger21, thanks!

However, I do not believe this will combat my motor issue on startup. I will have to give that a day or so and see if those still engage/disengage whenever they please. I am still in the mind frame that the motors/drivers are catching static discharge or something of that nature as it only does this after a full night of nothing or a long full day's use. I believe there are some CNC Gremlins in there somewhere...

Maybe my power supply is shutting down on the DC side? I'll pull out the voltmeter and check that box today and tomorrow when/if this occurs again. *Sigh*

Thank You!!!!

02-16-2014, 07:43 AM
It has been a day or so now, it seemed as though the time out issues were fixed but now I feel I found the power gremlins in my Nema case. I think my EMI filter/switch has gone bad. Now when I power up the box it automatically shuts down right away without even turning up the box fan to full speed. I belive the switch has been going bad for a while now, and I just didn't know it.

Would this cause the problems that I have been having? Maybe the switch would temporarily drop power and disable the power supply on the DC side but the AC side would automatically power up so I wouldn't notice any issues with it. In Mach 3 I wouldn't always get a time out, just the motors wouldn't work, and the digital encoders would continue working but the actual motors would sit there at idle. I'm ordering another switch, hopefully that will fix the problem. Any other suggestions?

02-16-2014, 08:26 AM
Any other suggestions?

Only change one part at a time and test thoroughly, so when you find the problem, you hopefully know exactly what it is.

02-16-2014, 08:41 AM
That's usually one of my biggest mistakes when this stuff happens. Thanks. (change everything before checking results) As of right now I have it connected to its own circuit, and I fiddled with the EMI filter/switch. The box finally powered up with the drives and all. I will cut a few plywood instead of mantel parts today so we don't loose to much if I screw anything up - I'll test it on longer runs if I feel its safe. I have contacted CNC Router Parts for a new switch, I would rather be safe than sorry. I know those switches are all over, but I can't find one with the proper bolt pattern for my case right now. I'd rather not look for stuff as I'v fallen a few days behind now - it happens though :0 (rarely)

If this fails I will bypass the switch in the case, and install another until something suitable comes in. I may not have EMI protection then. *SIGH*

Thanks again for all your help!

02-18-2014, 09:14 PM
If the system has been powered down for a while, all the capacitors will be already drained. When you power up the first time, the in-rush current can be pretty high while the caps charge. This may be enough for a faulty connector to not carry the current and cause weird things to happen. When you power down then back up, the in rush is much lower because the caps are partially or fully charged.

Replacing the switch/connector/cable may very well solve your problem.


02-20-2014, 12:47 PM

Jesse and I have been troubleshooting this over email as well, and the root cause was that the jumper for the power supply (to set it to 110V instead of 220V) was not installed. This was surprising to me, but in some testing here we have found that the SE1000 will in fact operate (albeit marginally) at 48V even when it is set to 220 and is fed 110. However, with 4 drives installed, this often causes the supply to shut down when it can't produce the requested current.

There are some other issues that we are still fighting through with the drives (possibly due to oddities in the power they have been receiving), but for anyone else using our DIY Nema 34 kits, make sure to set the power supply voltage input to the AC voltage you are actually using.

Best regards,

CNCRouterParts (http://www.cncrouterparts.com)

02-20-2014, 01:20 PM
Good tip! Actually happened to me on a PC a long time ago. Weird reboots and drives not always recognized.

Thanks for sharing.

02-21-2014, 01:27 AM
Replacing the switch/connector/cable may very well solve your problem.

Dead on for my motor problems, for what I can see as of yet. I have yet to post anything else as I keep running into the little power gremlins that "could" cause this.

After Ahren noticed my jumper wasn't installed I did so like a good little soldier and much to my demise it brought out all the gremlins :0 The idea of a simple jumper also alarmed the OCD side of myself and I checked everything again, and again... I finally figured out that I had some bad contacts connecting to my drivers/PMDX-134.
Because I re-wired everything manually and back to PMDX-134 several times I forgot the J4 jumper once as well, which shouldn't do anything other than motor disable any way. I made sure that was hooked up even though it never gets used.
Lastly, I enabled 1/2 pulse Sherline Mode hoping that it will increase traffic if that is an issue. (I have an Ethernet Smooth Stepper)
I'm un-aware of how effective this will all be, but I was able to burn out a small corner post today worth about 2 hours of machine time. Before doing that I was not able to do more than jog at half speed for the last week or so.

Tomorrow I will test it on longer runs - probably doors that should last 3-4 hours a piece. I will know by then and post weather or not my efforts paid off.

02-24-2014, 12:16 PM
Yesterday I ended up with the same issues on the same motor driver. So I ended up ordering a g203V, signal cable, and bulkhead connector in hopes that this will clear up these issues. I am pretty confident that this will solve everything for the time being. I may have stressed my drives so there may be future issues lurking ... :0

Thanks for you'r help guys - and quick response from Ahren as usual - I got a tracking number in less than 10 minutes...