MechMate CNC Router Forum

MechMate CNC Router Forum (http://www.mechmate.com/forums/index.php)
-   Troubleshooting (http://www.mechmate.com/forums/forumdisplay.php?f=83)
-   -   Machine stops responding intermitantly (http://www.mechmate.com/forums/showthread.php?t=3993)

Darin Tue 22 January 2013 10:53

Machine stops responding intermitantly
 
Hello everyone,
I'm getting a weird error I hope someone can help.

It can be any cutpath. In any random place during the cut for about a second the Mechmate will stop moving and the router turns off. Mach continues to run and the machine is out of sync of course. The router comes back on and the machine goes back to moving like nothing ever happened. Of course the cut and material is ruined.

I don't believe i'm loosing steps due to the router turning off and hanging the machine. When the router stops the machine stops, not hangs up.

My windows is striped of all extras. There is no networking, no anything. No programs running, just mach and this is the same computer i've always used.

My first thought was windows doing some random thing in the background and loosing function to the controller but There simply isn't anything else going on that I can find.

smreish Tue 22 January 2013 13:43

Do you have a UPS on the computer?
This could be a power interrupt

Darin Tue 22 January 2013 14:26

No. I don't I'm bare bones intentionally. I don't even used the hard drive except for Mach and the striped down windows. There's really no files to protect from power outage. All my cutting files are completed and stored on another computer and portable storage.

Thanks for your thought Smreish

sailfl Tue 22 January 2013 14:41

1 Check that your printer cable is attached all the way at both ends.

2 How long of a printer cable do you have between the computer and the BOB.

bradm Tue 22 January 2013 16:41

Looking back at your build thread it appears you have a PMDX-126, and that you are using a motherboard based parallel port.

The symptom you are describing sounds like a low-voltage hardware fault of some kind - the PC keeps on running, but the signals coming out of the machine aren't making it.

My first suggestion is to get an aftermarket parallel card in place. You could also try metering the signal voltages you're seeing out of the motherboard port. It wouldn't surprise me at all if they were down around 3.2 volts, which would make them marginal.

I've been burned enough times by marginal motherboard parallel ports that I never use them. They can seem to work for a while, and then give you trouble.

If that's not it, you need to look for a problem at the PMDX somehow - how is it getting it's power, and is that source reliable. Are there any lights on the PMDX that you can see, and do they do weird things when the symptom occurs?

You could try air cutting some really fast spirals that do a simultaneous move in X,Y, and Z to see if this will cause the symptom.

Also, note that if the signals are marginal, then the length and quality of the printer cable become even more critical.

Surfcnc Wed 23 January 2013 02:49

Darin

Following on with low voltage on the port idea, if that is the issue you may find searching for the term "parallel port booster" a very enlightening search.
Longer cable runs have voltage drop and the industry solution is put a signal voltage booster in line on the parallel port.
Such a device can transform the stability of the signal over many meters of cable run.
Easy fix if that is the issue as it also "conditions" a low voltage output on a shorter cable.

There are many other things that can cause that fault though, noise on the cable from your router or spindle, ground loops, a process on the PC, etc, etc.
The only way to fix it is start ticking off possible causes.

Regards
Ross

Darin Wed 23 January 2013 06:36

Thanks for the help and sugestions guys. Im not using the integrated moterboard printer card but an add on card. Im suspecting the card especially after your responces. Since i posted i wiped the hard drive and did a fresh instal of windows xp. I minamised all the windows functions and installed the latest Mach, tryng to eliminate the pc wondering off to some background function. On the first test the mechmate stoped responding i let it go just to observe and it did it several times in sort of a rythmic ossilation. The problem went for about 5 seconds then it ran normal. I had a couple files i really needed cut and risked the material and was able to get them done.

domino11 Wed 23 January 2013 07:17

Some parallel port cards are still a problem depending on the chipset. This one At Newegg is the same one pmdx sells on their site. Its a little cheaper at Newegg. In case you are looking for one.

Darin Wed 23 January 2013 08:11

That's the exact card I'm using. Has anyone ever had one fail? It seems so unlikely to me the card is bad. I tried to use the on-board printer port just to test things but I cant get it to work at all.

Darin Wed 23 January 2013 13:12

I ordered the new card from NewEgg. I also checked the pin-out voltages. The standing still voltage is 5.04. The step voltage drops to 2.61 when there's movement. The measurements are at the printer port at the pc with nothing hooked up but the meter.

bradm Wed 23 January 2013 13:33

That sounds good and healthy. I'd look at the parallel cable and the power source for the PMDX next.

You might also try disabling your router or spindle, and running an air cut to see if the problem goes away.

danilom Fri 25 January 2013 02:22

multimeter can't measure the real voltage of a step signal, as it measures average signal of the pulses. Only way to measure a step signal is with oscilloscope.

Darin Sun 27 January 2013 16:11

Well I cant see anything wrong with the cable or the BOB. Or the power supply to the bob. The power supply is just line voltage. I've tried everything i can imagine to induce the failure, so I can further diagnose it. I've yet to be looking in the control cabinet when it happens. I sat a starred at it for about an hour waiting for the failure but of course it happened about 2 minutes after I walked away.

I cant imagine how the BOB could suddenly loose power, but it does seem like that when it happens. I'm not sure how to test the printer Cable, any suggestions. The problem is so intermittent it eats up hours trying to make it happen, then once it does I don't know any more than i did.

My plan from here is to try the new printer card when it arrives and if the problem still happens try a different PC. It's a pretty bad situation for my shop, we use the Mechmate every day.

bradm Sun 27 January 2013 17:23

Darin, you must be really frustrated. Since this is a production concern, I think you should go ahead and get another printer cable as well. You can then use a process of elimination.
I'd be tempted to just swap out the cable, card, and PC all at once to see if the gremlins go away. If they do, you can then swap back in one at a time if you're inclined to narrow it down.

Hoping for the best for you.

KenC Sun 27 January 2013 18:14

It is common practice to use a multi-meter to check the printer cable.
the printer cable pinout is here
don't mind the written pin function, just use common sense.
1st check for continuity, if continuity is ok, then check for resistance.
I can't remember what is the maximum ohm allowed maybe someone else can fill in the number.

Surfcnc Sun 27 January 2013 19:25

Darin

If you can, beg borrow or steal another PC and swap it in as a test.
This will quickly eliminate a whole raft of software and PC hardware variables from the problem you are having.

If your machine is mission critical, you probably should consider having a backup controller.
While this might initially seem an expensive waste, that machine can be used at home, to do the books or whatever.
It just has to be ready to use not laying idle.

For others benefit a spare stepper motor, stepper driver and maybe even a breakout board round out the list of spares one is likely to need to keep the machine running.
I only have the first 2 as spares as a breakdown also provides the opportunity to update to newer technology as the break out boards and/or motion controllers improve.

My personal decision to go to the Ethernet Smoothstepper and use laptops as the machine controller was as a result of an intermittent fault with a old Del desktop PC machine controller.
In a few words I have felt your pain.

Regards
Ross

KenC Sun 27 January 2013 22:25

In electronics, most bugs are easy fix, that is when you find the bug. Don't be despair, you will find the bug sooner or later.
Be patient, be diligent & leave no stone unturned.

MetalHead Mon 28 January 2013 05:43

Darin

You may want to setup a video camera (or 2) to record what is going on. I may have missed this, but will it shut off doing "air cuts" as well? You may also want to consider that you may have a feed power issue. You can hook up a meter to the feed power and again record the meter with a camera to see what happens at the point the machine shuts down. I do not know how your business is setup, but if you or your neighbors have purchased any new high power equipment lately, power or EMI may be an issue. As far as a test PC, you can order a decent used desktop from ebay for less than $200.00.

Darin Thu 31 January 2013 10:45

Thanks for your posts and suggestions everyone.

I've traded out the computer the problem is still there. I don't use video cameras much so I camped out watching again. When the glitch happens its as if someone turned off the cabinet. Everything goes off.

I dug in deep after seeing that. I rewired the door buttons and changed the main contactor. I checked every line voltage connection in the box. I found a ground wire off and thought that might be the problem, it wasnt. My friend/employee suggested the surge protector I was using might be causing the problem with a split second power block, still the problem continues plugged directly to the wall. Changing computers has brought about a ton of new issues to the point I cant even run a test now. I've ruled out the computer so my next step is to put the original pc back in place. Then I may try to temporarily bypass all the e-stops, door buttons, main switch and contactor, so I can try introducing things in to isolate the issue. I've also been trying to monitor the line voltage as the problem happens but so far I havnt had my eyes in the right spot in the right moment. I need a video camera savvy friend I guess.

Please keep the suggestions coming.

smreish Thu 31 January 2013 11:19

What is your debounce setting, set at? If you getting some 5khz or other transient, it might trip your e-stop circuit.

Better question, when it stops does MACH 3 show that an e-stop was enabled?

MetalHead Thu 31 January 2013 11:39

The fact that the whole cabinet goes off tells me it is not a setting. Is the PC plugged into the same strip as the cabinet? Also in your FUSE BOX check for a loose Neutral terminal as well as make sure the breaker screw is tight.

Plug the PC into the same strip as the cabinet and if it is a power feed issue atleast the PC will shut off also and quit screwing up panels.

smreish Thu 31 January 2013 11:53

....missed the part about the WHOLE cabinet shutting off.

Darin Thu 31 January 2013 13:31

The PC and controller are plugged together. The PC never shows any sign of a power outage. My thought was the power could be down a short enough time that the pc power supply absorbs the outage but the main power relay loses its hold momentarily. Mach never shows any sign of anything happening and the cabinet is down just long enough to hear the router start to slow down, and it comes right back to life. The machine losses about 1" of travel running at 200 ipm. That was my thinking to support a power supply issue anyway. I have my doubts about the power supply being interrupted, but as soon as I get the old computer in place I will run some tests on different circuits and definitely check the breaker and neutral connections.

I'm needing to re-visit the threads about replacing Milwaukee router bearings too. No rest for the weary :)

I'm strapped for time right now. We're getting product ready for an out of town show this week end so it will probably be Monday before I get to work on this anymore. I just sold some more work to cut on the Mechmate though so it is a major priority.

Thanks for your insights everyone

domino11 Thu 31 January 2013 14:25

Try this at least for the router bearings. :)
Milwaukee router 5625 bearings replacement

domino11 Thu 31 January 2013 14:26

Do you have a schematic of your control box?

Red_boards Thu 31 January 2013 17:10

Darin,
I think I had a similar problem. Nearly drove me crazy. I plugged the PC into a separate circuit and a got a UPS. Didn't help.
I did a lot of things with software and eventually replaced the PC power supply (the box with a fan inside the PC) and the problem went away. PC power supplies are pretty cheap.
Red

bradm Thu 31 January 2013 17:37

Restating some things from above to make sure the assumptions are correct.

1) There is no actual evidence of a power loss in the control box. There is an an observed behavior that is "as if" the power to the control box were gone for a second.

2) Not only do the stepper motors stop, but there is also an audible change in the spindle, as if it were also shut off.

3) An entirely new PC was tested, and showed the same symptoms.

Presumably the router is controlled by a signal from the PC. If it isn't then we're definitely looking for power source problems, like loose neutrals as mentioned above.

If the router is controlled by the PC, then I think we're back looking for low voltage signaling problems.

Having the main contactor drop out and come back without you noticing the thunk seems unlikely. Similarly, I'm having trouble envisioning a flaky Estop that re-enables itself (in fact, that would be kind of scary, no?).

So, musing on the low voltage piece of the puzzle:

I poked at the parallel port, but it seems to have been exonerated. We haven't yet heard if you've tried a new parallel cable or not. It occurs to me that you might have insufficient ground connection between the PC and the PMDX (this is spread over multiple pins), or a lose or dirty connection; marginal wires in the parallel cable, or perhaps a condition where the ground plane is getting energized, which would behave similarly to being disconnected.

I'd do a close physical examination of the parallel cable and connectors; try wiggling the parallel cable looking for broken connectors while the machine is running, and try replacing it if you haven't already.

Presumably there are a bunch of LEDs in the control box. It would be useful to know if any strange patterns appear during a failure.

That's about all I can think of at the moment. Unless you've got someone with an MRI machine on the other side of the wall that keeps turning it on at inopportune moments.

Surfcnc Fri 01 February 2013 01:25

Darin

A bit of my favourite brand reductionist trouble shooting might be in order.
Particularly I would eliminate the router from the mix and try to elicit the fault without it powered up.

A router can generate a lot of noise esp when the brushes are very worn.
The electrical noise is created from the arching between the worn brushes and the contact made of copper they run on.
So a machine running well can reach a certain point in router wear and all of a sudden start giving problems.
Even machines that are side by side can cause the other to have problems.

You were not specific about what other problems the replcement PC caused. so what happened exactly ?
Revisiting my suggestion regarding the parallel port booster, it may be time to ensure your signal strength is given all the help it can get to eliminate that as a variable.

Food for thought anyway :)

Regards
Ross

Castone Sun 10 February 2013 11:19

Z touch software
 
If you have any type of Probe software remove it ,along with the wire to your board. I have it it cause a similar problem on my machine.

Tom Ayres Sat 15 June 2013 17:50

Darin did you figure it out? The suspense is killing me.


All times are GMT -6. The time now is 02:14.

Powered by vBulletin® Version 3.8.3
Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.