[vsystems]

I've posted before looking for advice on this problem, the one axis stutters and shakes when jogging or when running a job, after a process of elimination of power supply, drivers, and mechanical, it seems the problem is in the controller itself. It is a DSP controller, I'm wondering if this is anything anyone has seen before (axis sits still and vibrates when jog buttons pressed, but when it kicks to high speed it moves smoothly in one direction until it reaches the switch)
And if there is a way to repair this, or if I should look at a replacement controller, and if so, what replacement? Since it's DSP and has an LCD control panel/buttons, what would be able to replace it?
tia.

Update: I've pulled the motion controller, it appears to be a mpc6515

[SkipW]

Just a thought...

Have you tried raising the slow jog speed a little in the software? Like if it is set to 6, raise it to 8 or 10 and see if that does anything?

If it is a controller issue it probably won't make any difference but there's a chance it could be a work around.

[DaveDoesIT]

I assume you have swapped the drive cables over to make sure that the problem stays with that axis.

Are they steppers, open or closed system? If closed or servos try swapping over the encoders. Check the encoder connections.

How many components are there on the board that can be swapped around? As in ICs in sockets? Swap one thing at a time and test until the problem changes.

Often you can piggyback a new chip over an in-circuit one. Can be a risky thing to do though as it may Pffhht something else. In a pure logic circuit it should be OK, but do not hold me to that. Your RISK entirely. Turn the power off and let it set for 1 minute or so before piggy backing if you do it. make sure you have the IC the correct way around. Ground and power may be swapped and jittering may suddenly be the least of the problems.

If you have one good axis then use a scope and digital probe to compare waveforms and state at various points between the two axis. Find where it suddenly goes haywire on the problem axis.

It is virtually impossible to tell you what can be done without being in front of the thing and a schematic in hand. You just have to be as rigidly analytical as possible.

Dave

[vsystems]

Just a thought...

Have you tried raising the slow jog speed a little in the software? Like if it is set to 6, raise it to 8 or 10 and see if that does anything?

If it is a controller issue it probably won't make any difference but there's a chance it could be a work around.

The only controls that I have access to are the control panel ones, there are no further software controls that I can tinker with since it's not mach3 or anything good.

[vsystems]

So far I've replaced the stepper drivers completely, no dice there, I tested all the power supply outs which all were good, tested everything I have access to up to the MPC6515 board. I've read that these frequently experience firmware corruption? So I'm in the process of tracking down and installing the latest available firmware. Strange thing is that the board is stickered with v4.1.2.0 and no trace on leetro's site for anything higher than 4.1.1.0.
I've seen some forum posts mentioning 4.1.3.0 but google isn't helping me.

[vsystems]

How many components are there on the board that can be swapped around? As in ICs in sockets? Swap one thing at a time and test until the problem changes.

Dave

Was going to try this if the firmware thing doesn't pan out. I'll keep ya posted.

[DaveDoesIT]

Did you try swapping the motors and see if it follows the motor or the axis.

I would not worry about stickers.

If it was firmware it would be on both axis. I would not have different firmware for each axis movement.

You need to be one-step-at-a-time analytical here.

Dave

[SkipW]

The only controls that I have access to are the control panel ones, there are no further software controls that I can tinker with since it's not mach3 or anything good.

It doesn't use LaserCut software? My mpc6515 card is ver. 4.0.0.15 and the jog speeds are adjustable in the LC software. That probably wouldn't help you anyhow. Unfortunately I don't have the electronics knowledge that Dave or Zax or some of the other posters here have, so that was all I could come up with.

I was just curious if changing the slow jog speed would have any effect on the problem. Apparently there is a bug in my LC 5.0 and controller combo that prevented me from being able to jog the head at all after I unchecked the "auto datum" feature in the software. I had to change the slow jog speed in the software to 8, then it worked again.

I'm sure that has been corrected in later versions.

Hope you get it fixed.

Skip

[vsystems]

I did switch the motors, and the problem followed to the other motor, so it's not the motor. I did the firmware update and it did to SOMETHING, it removed the slow jog speed from the control panel and now any jog press is full speed, however this still does not solve the problem. That's when i switched the drivers and even with the new drivers in place, the problem still existed. So what am I missing between the drivers and the MPC6515 that's causing the problem?

I just did another test and swapping the pulse cables that go to each of the drivers not only reversed the motors (i.e. up = left, down=right on the control panel) which is to be expected, but it also made the problem follow to the other axis. So I checked the mpc board again, there were two identical ICs in sockets on the board, I switched them around, that didn't make any difference. So again, I'm somewhere between the mpc and the drivers. Perhaps I'm wrong about the power supply? I just don't know enough about how this thing is wired.

[DaveDoesIT]

I will be surprised if it is the power supplies. Did you try swapping those leads to the controllers?

Well you are narrowing it down. If swapping the two ICs changed nothing then it is beyond them and before the pulse cables. Time for that scope and digital probe.

have you swapped those pulse cables over?

Have you examined the solder joints on the boards (top and bottom) with a magnifying glass?

Just keep trying the swaps until you have eliminated everything that is swappable. Then it is time to start comparing signal stuff.

What about a couple fo pics showing the two ICs you swapped and all the way to the pulse cables. It is really hard trying to diagnose this without any idea of what I am looking at.

Don't bother with trying to think what might be wrong. At this point it is now down to hard signal testing.

Dave

[vsystems]

Well I've switched the power cables that are going into the drivers, no change there, and since the problem followed witch whichever plug was on the Up/down (Y) axis no matter how I switched the plugs, I believe you're correct it has to be something to do with the pulse/direction wires that are going into the drivers. as I understand it the UVW wires are to the motor, correct?
I'm currently without a cable for my camera, so I can't get any decent pictures with my phone. But the wires are so tightly wound and wrapped into spinal-cord type bundles with no markings, it's very hard to trace them so I'm doing continuity tests now.

[zax15uk]

From the testing you've already done, it sounds like the problem could be with the cable between the driver and motor.

Zax.

[vsystems]

Would a bad relay do this? It seems the wires I've traced that I've switched around the UVW wires go to a relay which seems to be the device that switches from slow to fast speed after holding down the jog for a few seconds, or at least as far as I can guess.
I've been trying to find a primer on CNC mechanics so I can understand what the pulse wires do and where they're controlled so I can understand what I'm looking for.

[vsystems]

From the testing you've already done, it sounds like the problem could be with the cable between the driver and motor.

Zax.

If that is the case, how might this have happened? since the machine hadn't been used or moved for a few days, then all of a sudden I turn it on and it does this. I'll try tracing the continuity on those wires, but it will be extremely difficult because of how bundled up it is.
I'll post on my results.

[vsystems]

Good news then I'm almost there, by switching the motor cables from the drivers AND the terminals where the motors plug into a coupling (which go deeper into the machine and then presumably to the drivers at some point if not to the relay first) I was able to get the problem to completely switch axis. now the problem is on the X axis.
Zax, thanks for the insight, but I'm not quite there, I have to figure out where those cable ports couplers go and since the problem jumped axis, that rules out the problem being between the motors and the couplers, but instead between the couplers and the drivers.

[vsystems]

Ladies and Gentlemen, I've found the culprit.
It was indeed the relay. I traced the wire going from the motor's coupling, throughout the entire machine only to find it connected to a block which had a clear relay block plugged into the top, I yanked that out and lo and behold, there's a large scorch mark on the bottom and bubbled plastic inside. the part is labeled AIKS ARL4F 10a 28VDC, 10A 240VAC Relay.
Now I just have to find a replacement and I think I'll be back to engraving FINALLY!
in order to get to this solution, I had to remove 64 zip ties.
I counted them....

[zax15uk]

Good news, I hope it all works once you replace the relay.

I would order 2, just in case you have a loose wire somewhere that caused the damage to the relay. If the motor circuit is interrupted while energized it can result in a high current certainly capable of causing the relay to arc and fail - also a good chance it could damage the driver too. Double check the wiring to make sure there are no bad connections.

Zax.

[vsystems]

I really hate to have to bump my post, but here goes... I got the relays in. 4 of them in fact since they're insanely hard to find and I had to use google translate to order them from the chinese manufacturer.
I replaced it and it worked great until i discovered a broken wire on one of my limit switches. So I repaired that, but ever since replacing the relay, I can jog the unit in any of the 4 directions smoothly, but any of the files stores in the unit weren't executing, so I hooked it up via USB and sent new files and also tried to send data from the unit. It's not working, it'll jog via software or control panel, but doesn't execute any programs or datum.
So I tried a few different things, and now the Y axis has died again, exact same behavior as before, but... the relay is good, there's no other noticeable damage to anything else, and I'm ready to post this 700lb piece of junk on craigslist.
Any suggestions on what else might have gone awry?

[zax15uk]

I can't explain why your programs wouldn't run, it sounds like a problem with the controller interface. The job keys are likely wired in to a later circuit.

Since the axis is now acting up again I would check continuity on all the wires to the motor (from the driver to the relay, and the relay to the motor). You may need to move the wiring bundle while checking to detect any issue.

If the fault is just wiring it should be an easy fix, but as I said before it is possible the open circuit could damage other components.

Zax.

[CNCZILLA]

You may think this is rediculous but it sounds like your belts may be to tight, if the axis is drivin by belt. If the servo belts are changed or the motor is replaced or swapped and you get the belts to tight when you put them back on this can cause the axis to sit there and vibrate but when running at faster speeds it seems to go away. This is because on a servo system the drive is trying to hold position, so if the belt is to tight the machine is fighting itself so to speak to try and maintain that position. When the machine is running it is not at rest so you don't see the problem. This is true on most all cnc's, and all a laser is is a cnc with a resonator on it. Check your belt tension and loosen it slightly, you might be suprised at the results. It could also be the opposite and the belts are way to loose but it is usually because they are to tight. Of course if you don't have belts then don't pay any attention to this. Good Luck.