QT10N occasional x axis fault

We are getting a periodic fault #13 on the x axis. The drive LEDs are indicating an over current situation when the fault happens. The fault does not appear to be related to machine position, sometimes it can be sitting idle and go into fault mode.

I am told the suspects are the wiring, the motor, or the drive.

There are 4 wires at the motor in addition to the two multi pin connectors. They are labeled IG1, IG2, A and B. These four connect with the drive. Excuse my embarrassing ignorance of this, but I thought this would be a three phase motor with 3 equivalent wires. That does not appear to be the case.

Where should I begin the diagnosis? Is an insulation problem on the wiring a likely suspect? There seems to be a fair amount of coolant that makes its way back there, maybe that hardened up the insulation and is allowing a short?

After testing the wires, would attempting to isolate the problem by switching the output wires between X and Z be a good idea? The drives are the same (I think).

I bought a 1000v meg ohm tester to diagnose the motor if it comes to that. I know to disconnect the power and all connections to the motor before testing it to avoid frying something up the line. If I do that and ground the black wire on the motor (anywhere on the motor chassis?), can I apply the tester to wires A and B (i am assuming these are the primary conductors since they are larger than the IG wires) to check the windings? Is that all there is to it? I somehow doubt the motor itself is the problem since this machine is not in constant use. I thought it might be a good idea to check the brushes while I'm in there and maybe get a can of clean air to blow out the holes and any carbon that may have accumulated in there. In my (limited) experience, drive amplifiers are more fragile and likely to have a problem than the motor. I have no evidence to support that in this case, however.

Thanks to all for reading and sharing your thoughts.

You may wish to remove the motor and run it off the machine. Even at idle, the motor might by at 40% current.

Before removing the motor, make sure the brake at the end of the ball screw is working. If not, then the X-axis slide may free wheel the ball screw causing the turret to slide down. Block it up just in case.

The brake may actually be the problem. It needs power to release. If the coil is going bad, then it may not release causing an over current alarm.

If the motor is O.K. and the brake does release, then check how the ball screw turns. The nut or the bearings might be bad.

I hope this is useful to you.

Ah, the brake. Another suspect quitter. I haven't heard that one yet. Would that be a problem even at idle? The motor could be working to hold up the turret and fighting the brake somehow? Is there a test for the coil?

Will check. Thanks for the suggestion.

Your megger voltage is way too high for this motor, or any 240 Volt 3 phase motor. You can damage a good motor with it!!! Use it only on 480 volt motors. With this motor using a handheld DVM from one large lead to ground you will probably see 100K ohms or less. Maybe even as low as 10K if there is lots of carbon in the motor.

Before removing the motor find the encoder. If there is a belt to the leadscrew, the encoder may be on the leadscrew. If the encoder is on the motor you can remove it and run it. Be sure to run a small wire from the motor to the machine frame so the motor will not be "hot" electrically. Carbon grounds inside the motor can cause the frame to be "hot" and if there are broken wires the ground wire can be broken.

If you swap motor leads without swapping encoder leads also you will have 2 runaway axis that the E-stop overtravels might stop them before they smack the physical limit.
I do not recommend an inexperienced tech to do this! There is an issue of the Z axis motor is way too large for the X axis servo drive too.

First check; Does the motor look like it has been very hot? like paint burned? (very bad)

Second set of checks: With power off pull one brush out, note orientation so you can put it back the same way, and not flipped over. Is the end that is seated on the commutator shiney? If yes, that is good, is it least 1/2" long? If yes that is good. There should be 4 brushes on that motor, so it will run with one out. With the blocking to prevent the axis from falling, turn power on and handwheel in the smallest unit of travel away from the block while looking into the brusholder at the commutator a small flashlight is needed. The appearance of the commutator surface should be smooth and the color of a brown penny the slots should be even without burns on the edges. Turn the handwheel slowly (2 people will be needed) and look at the commutator for 1 turn of the leadscrew. I forgot if it is 8 or 10mm/turn.

Analysis time.
If either of the first 2 tests were a no than the commutator probaby will not look like a brown penny surface. You may see black burns or grooves in the commutator, or 1 or 2 comutator bars with burns at the edges. If the brush was less than 1/4" long you may see the copper wire exposed on the end that is seated on the commutator. If this is so and you see grooves in the commutator from this or another brush and burns or a bare copper streaks on the commutator, a rebuild is probably in order.

If things look good power off and reinstall the brush. Power on and examine the drive area.
1. Does the motor hum louder then the Z axis?
2. Step the handwheel as you did previously. Does each step move the same amount rotationaly on the leadscrew? Or will it jiggle back an forth at some spots? It may be quite aggressive and buzz (not good) and may be the source of the overload if it will not stop within a second.

I have seen 150% loads on a machine that had been badly crashed repeatedly by a poor operator. The screw was dimpled from the balls in the ball nut and the bearings were worse. The buzz was quite severe as a position that was partway into a dimple was moved to. There were big problems with accuracy that got worse with each crash.
Jogging the axis back and forth there was an obvious bearing rumble and it got worse when the axis got over the crash area.

Also looseness in the drive system between the motor and the encoder can cause the thrashing. Couplings do wear out and have backlash.

For cabling problems usually the problem is Z axis position dependant, you may find the flat spring from the carriage to the frame broken, Or the wires have chafed, or bad chip buildup and a chip wore thru the wire, or the sealtite broke and the wires inside will be broken also. If there is a coolant drizzle or spray over the wires it will find any insulation cracks and flash to ground. I usually move Z back and forth until X faults out, if a visual inspection looks OK, and then start by looking for a ground in the wires or between wires in the motor cable. Be sure to disconnect both ends of the wires first, with power off of course.

Lastly the servo drives do go bad with what you have described.

I can't see the brake as a problem if it faults out when not moving.

Post a reply when you get it figured out so others can benefit from this thread.

Bill

I checked the motor more closely today. There is no evidence of burning on the exterior. The multi meter showed the resistance from both leads to ground at about 14.5kOhm. This is at the low end of your range. I removed one brush and found a pretty good amount of free carbon but a brush length of more than .5". Would it be a good idea to remove the brushes (noting orientation) and blow it out with clean air? Could this kind of carbon accumulation cause the current overload situation that I'm getting? Thanks, Bill for your help.

Carbon is a problem

Yes, the build up of carbon could be causing your low ohm breakdown to ground.

It would be best to clean out the carbon by taking only two opposing brushes out at one time (assuming there are 4) noting the location and orientation as you mentioned. To properly clean out the carbon use clean, dry air in one side while you place a vacuum on the other side. This will help get more carbon dust out of the motor rather than get some out and push the rest deeper into the motor.

Given the wiring you mention on the drive the motor is most likely a shunt wound rather than a permanent magnet DC motor. IG1 and IG2 would be the field voltage and the A and B would be the armature voltage.

I blocked the x, took off the brake and removed the motor. I followed Pipes' advice and used a vacuum and a can of clean air to blow out the motor. After reassembly I checked the resistance again and it was up to 80k ohm for both leads. After doing some cleaning on the backside of the machine I turned the power back on, started it up and it appears to be running normally.

Thanks to Dinosaur, Pipes and Hitandmiss. I hope sometime I can return the favor.

Well I just got the error again. It does seem to work for a longer period of time after being off for a few hours, but after it faults, it will continue to fault again and again. I thought this could be something in the servo amplifier or maybe coolant finds its way down the cables after a little while.

Did the resistance to ground stay at 80K after it has been run?


My first post is a fairly detailed checklist of things to check for common electrical problems. Mechanical problems can not be ruled out yet. It is easier to check the electrical first.
1. Surface finish of end of brush. must be smooth and shiny.
1A. Check the other brushes, any stuck or were stuck (hard to pull out)?
2. Commutator surface finish and color any burned spots?

1, 1A, and 2 are probably the most important items to check.

3. Is the encoder mounted on the leadscrew? or motor?
3A. Has the encoder been wet with coolant?
If the encoder is on the motor you can pull the motor and run it disconnected from the leadscrew. If the encoder is on the screw, and both motors are the same P/N you can swap motors. The qt10 is a small machine and both motors may be the same. If the lathe has 2 servo amps (Both amps in the same chassis are common and outside the scope of remote troubleshooting to explain how to swap them) and both are the same P/N they can be swapped but you must carefully examine the jumpers and switches and change the drives to the other configuration. Make good notes before proceeding.

4. Does the motor hum louder then the Z axis?
5. Measure the resistance of IG1 to IG2 if the resistance is less than 10 Ohms they connect to an internal thermostat. I have never seen a shunt field on these motors.
6. perform a careful inspection of the large A and B wires, if they are in sealtite, is it broken or kinked anywhere along it's length? If yes, you may have broken wires near the break in the sealtite.
7. Do a VOM check of A and B wires. Disconnect both ends and measure A to B, A to ground and B to ground. Should be over 1 Meg. for each measurement.

The troubleshooting objective is to isolate the problem, I try to rule out various common problems, then move on to the less common. Unfortunatly some of the common ones are very hard to identify for the inexperienced, so there are a lot of what appear to be unrelated tests or checks to determine if more tests or checks are needed, and less emphasis on swapping drives etc. Reconfiguring drives can lead to adding more problems, something I wish to avoid.

Background information:

If the commutator is burned or scored the brushes will not make a good connection allowing arcing. Brushes stuck in the brush holders will cause this. The same happens if the A or B wires are broken. The arcing causes malfunctions in the servo drive, leading to an overcurrent condition.
If A or B wires have coolant soaked insulation, The armature voltage can flash to ground creating a momentary overload, evaporating the water ground in the process. then further tests look OK.... for awhile.
Broken wires tend to be Z axis position dependant, but not always, and can be quite hard to find. The quickest way to rule them out is to run temporary A&B wires from the amp to the motor. Note an a-b swap results in a runaway axis.

Mechanical slop from motor to encoder (or an encoder with a bad spot) causes instability or excessive viberation that will result in an overload.

Bad bearings (even in the motor) can cause various problems of overload, instability or inability to hold the actual position.
The Handwheel rotation while looking at the commutator tests and noting the various Items mentioned in my first post helps rule out some of these areas, If you see something inconsistant, Look over the Z motor for a comparison.

Most common problems are in the motor and can cause drive problems.
DC drives are quite problem free compared with AC drives, but do go bad.
Mechanical problems are common, but for your X axis can be hard to diagnose.
Wiring problems are becomming more common as older machines are in common usage and be hard to troubleshoot.

It is tough to think of and document the things I check and observe on what is probably less than a 1 hour service call. Did one Friday and it was less than an hour, it was a burned commutator.

Nuff for now,
Bill

Bill,

I'll put the answers that I know now in the text of your post. I'll have time to pursue the others later today or tomorrow.


Did the resistance to ground stay at 80K after it has been run?

*Haven't checked yet

My first post is a fairly detailed checklist of things to check for common electrical problems. Mechanical problems can not be ruled out yet. It is easier to check the electrical first.
1. Surface finish of end of brush. must be smooth and shiny.

*It is shiny and smooth

1A. Check the other brushes, any stuck or were stuck (hard to pull out)?

*No, all were easy to remove

2. Commutator surface finish and color any burned spots?

*Nice even copper color. It's a little tough to be absolutely certain that there is no scoring though. I thought it looked ok.

1, 1A, and 2 are probably the most important items to check.

3. Is the encoder mounted on the leadscrew? or motor?

*The encoder is mounted on the leadscrew. The brake's on the end of the motor which drives the leadscrew by belt.

3A. Has the encoder been wet with coolant?

*I don't think so. There are some chips in the area, but it seems pretty dry.

If the encoder is on the motor you can pull the motor and run it disconnected from the leadscrew. If the encoder is on the screw, and both motors are the same P/N you can swap motors. The qt10 is a small machine and both motors may be the same. If the lathe has 2 servo amps (Both amps in the same chassis are common and outside the scope of remote troubleshooting to explain how to swap them) and both are the same P/N they can be swapped but you must carefully examine the jumpers and switches and change the drives to the other configuration. Make good notes before proceeding.

4. Does the motor hum louder then the Z axis?

*The hydraulic pump is loud and pretty much drowns everything out, but the X motor sounds pretty smooth. I watched it move while cutting some parts. On back and forth motions (these parts have threads) it didn't seem to have any vibrations in the couplings or unusual noises associated with the ball screw. I noticed a slight ticking noise in long x rapids though.


I'll examine the electrical stuff later today or tomorrow. I'm interested to see what the resistance on the motor leads is now.

If the motor's resistance to ground says above 50K and the brush/commutator condition is as you indicated, the motor probably isn't the problem.
So the focus moves to mechanical and wiring problems.

continue checks 5-7

If check #6 - looking over the wires for obvious breaks and none are found.
You may still have broken wires so the following data collection may be helpful
If the alarm gets consistant ie every 3-4 min of running. try running and watch axis movement.
is Z within 2" or so of the same position when the alarm comes on?
was the axis that was moving when the alarm came, on moving in the same direction as previous times?
Does X act any different than other stops? oscillation, bussing, tick or scraping sound

"I noticed a slight ticking noise in long x rapids though."
Some things to look over while trying to locate the source of the ticking.
While you are looking X axis drive parts over, check the ways or linear bearings and ballscrew for lube? clean and wet with way oil? damp? or 1 or 2 areas are dry? saturated in coolant? lots of chips on or near the ballscrew? some chips/coolant, or no chips or coolant in the area. chips in the timing belt? Chips or coolant on lower ball screw bearing?
nuff for now,
Bill

The ticking noise could be from the balls in the ball screw nut recirculating through the return tubes. Over time, those tubes can deform where the balls come out of preload and go into the tube.

It is possible that this alarm could be caused by poor encoder feedback that might be why it happens even when at idle. Since it is an old machine, the cable in the wire tracks could have had their insulation worn through or some chips may have become embedded.

Dinosaur,
I am thinking along that line too, but have seen lack of lube in the ballnuts and that may be the root of it. Old machines metering units get plugged up and eventually things run dry, and with no oil film on them, the balls just smack each other with a click. I am pretty sure it is a linear guideway machine so there are recirculation tubes on them also.
The qt-30 I redid the lube lines on last year, the X axis had lube lines going to the ball nut and each linear ball assembly. After replacing the metering units and blowing the gunk out of each line the machine ran quieter. I was amazed at the amount of metal flakes that came out of the ballnuts once they got proper lube.

Bill

After running the motor a little, the resistance reads 58kohm through both wires. It was 14.5kohm before blowing it out, and 80kohm immediately after.

i suspect the drive

I am with wippin' on this one. 9 times out of 10 it's going to be the servo drive.

Call Mitsubishi Electric Automation in Chicago. You can send the drive in for evaluation...maybe 150 bucks. Rebuild is around 1500.

Good luck...and let us know what you find.

Greg

I looked over the cables, they aren't chaffed. I think having the drive checked would be a good idea. From a beginner's view it seems like the drive, with its greater complexity, would be more likely to have a problem than the encoder. I've had drive problems on some of my other aged machines.

Just wondering, how intermittant does this alarm occur?

If it's been off for hours it will work for about 2 hours before we get an alarm. They come rapidly after that.

Hi No Fun,

The intermittant problem is the hardest to get repaired. Mitsubishi had problems a few years back repairing the intermittant overcurrent and overload failures. They seem to have improved somewhat. Sometimes they come back with no problem found. And the problem is still there when you put it back in. This can be frustrating if you need the machine running. So.. the next step is to call Mitsubishi at 1-847-478-2500 select CNC tech support They will need the machine model and S/N The part number of the drive with the problem. Describe the problem similar to this: If the machie is off overnight it runs fine for about 2 hours before we get the alarm XX and describe it as your first post does. After powering down and restarting we get the same alarm in 5-10 min ( pick the range that fits the problem) even if the axis is not moving. They will usually ask a lot more questions to determine if the motor or cables may be the cause. The answers you have provided here need to be plugged in there as needed. You have done your homework just have your notes ready! From the tech. you need the P/N (not the same as on the drive) for repair and a notification # that starts with 300XXXXX. Next you will be transfered to the parts dept. There you can review your repair option(s) available and the cost of each. They will fax you some paperwork to fill out and fax back. Then you get the RMA # so you can send it in for repair. Be sure to put the notification # on every one of these forms in the space provided. Expect a 7-10 business day turn around unless you pay extra for faster service. Or purchase an exchange (if available) and they cost more too.

Good luck, and keep us posted.

Bill