HCAL, Servo 401 Fanuc 6mb

[Andy Fritz]

Hi Folks,

I have a Hitachi Seiki VA35, with Fanuc 6 mb, DC servo amp A06B-6047-H003 and top pcb A20B-0009-0320, model 10 Yellow Cap DC motor which has given me lots of grief over the years dishing out various problems at random. The latest was servo error 401 on start up with HCAL on X axis. I have had those before on Z and then fixed it by replacing the 1"x1" black opamp on the servo amp pcb. When I got it on X it seemed to resolve itself if the machine and control was powered up a few hours before the stand by button was pressed to energize the servos. I just got a replacement pcb cheap off ebay and desoldered the opamp and replaced it. Now I dont get the alarms on start up, but instead I cant move the X axis. That is, when I move it, the control immediately gives 401 and the HCAL comes on the X axis pcb. I ohmed out all the transistor bridge check points on the amp and none has less than 10 ohm. The higher resistance points are higher than the same points on the Y axis drive. I also checked for shorts between the DC motor cables and the 3 phase power to the amp and it seems fine. Next I swapped the Y and X amps, but did not swap pcb's as I didnt want to readjust the trim pots (I have never done this and I dont have an oscilloscope). The problem stayed with X and Y worked fine using the X amp. This would suggest the X amp is ok. Next I did a continuity check on the four motor cables. All fine. I dont know how to check the pulse coder or if a bad pulse coder would give the HCAL alarm at all.
Is it possible that if the trim pots on the VCU pcb where off, would this give HCAL on moving the axis? Maybe if the gain was set too high? Then again, I have not touched the pots and the pcb is the original that the machine came with. But maybe changing the opamp requires re-tuning the pots?

I guess what I really need help with here is;

1. Suggestions on what else could be the problem.
2. How to check a Fanuc yellow cap DC motor and pulse coder.
3. How to check the tuning of the VCU PCB trimpot settings and other parameters with out using a scope.

I have a maintenace manual for Fanuc 6mb, but none of the servo amps or top pcb's are in the manual are the same as the one's I have.

Any help is greatly appreciated!

Cheers,

Andy

[Al_The_Man]

I just fixed one with identical 401 error problem, no error until an attempt to move.
This turned out to be a build up of carbon on the brush housing and causing tracking to ground when power applied, fortunately I had a stripped motor that had a good brush holder assembly.
I would remove the power connector at the motor or drive end, and with the lowest resolution on the hand wheel, rotate the HW slowly and see if you get the 401, you should get a following error after a certain amount of rotation, but if you do not get the 401, it most likely points to the motor.
In most cases you can slowly move the motor/ballscrew shaft in the direction of motion you would normally see so the encoder roughly keeps up with the HW, if you do this carefully, you can actually do this without the following error.
Al.

[Andy Fritz]

Originally Posted by Al_The_Man I just fixed one with identical 401 error problem, no error until an attempt to move. This turned out to be a build up of carbon on the brush housing and causing tracking to ground when power applied, fortunately I had a stripped motor that had a good brush holder assembly. I would remove the power connector at the motor or drive end, and with the lowest resolution on the hand wheel, rotate the HW slowly and see if you get the 401, you should get a following error after a certain amount of rotation, but if you do not get the 401, it most likely points to the motor. In most cases you can slowly move the motor/ballscrew shaft in the direction of motion you would normally see so the encoder roughly keeps up with the HW, if you do this carefully, you can actually do this without the following error. Al.

Hi Al,

Thanks for your reply.
I was hesitating to disconnect the motor as I was not sure which pins to jumper on the pcb. I have one manual that says to short S23, but that is for an earlier pcb than the one I have. Anyway I took my chances and put the jumper on S23, disconnected the 4 motor cables and could move X with out error. I did not get any following error. Next I cleaned the brushes and blew out carbon dust with air. I also did a diode check between motor chassi and motor cables. Hooked everything back up, but same problem. The maintenance manual only describes cleaning the commutor area for motors with tachometers, encoders etc. It says that motors with only pulse coder (like my model 10m) there is not much to do. The commutor surface looks ok, but it is hard to really see looking through the brush holders. I dont really want to try to rip the motor apart as I dont know anything about this unless I can get some directions.

One thing is weird: The manual talks about checking for short between the 4 motor cables. They are all shorted, both on the amp terminal and on the motor Cannon connector. This is also true for Y axis which works fine.

Andy

[Al_The_Man]

The armature cables are typically in 2 pairs, there is no need to jumper anything, just disconnect the A1 and A2 terminals on the output of the drive or at the motor.
There is very low resistance between each of the pairs due to low armature resistance which may appear to be a short.
It sounds like a motor problem you have.
Al.

[Andy Fritz]

Originally Posted by Al_The_Man The armature cables are typically in 2 pairs, there is no need to jumper anything, just disconnect the A1 and A2 terminals on the output of the drive or at the motor. There is very low resistance between each of the pairs due to low armature resistance which may appear to be a short. It sounds like a motor problem you have. Al.

Al,

I am not sure it matters, but the motor terminals are labeled 5,6,7 and 8. I have now found a maintenance manual which have the same VCU pcb I have.
It does say to jump S23. Anyway, here is some news. I tried to use the replacement pcb I got of ebay and soldered back the op-amp. Then right of a sudden it worked! I could jog, zero return etc. Only noticable thing was that the X didnt feel as smooth in acceleration and decceleration on slow jog as before. Gain adjustment needed maybe? Anyway I was so happy it worked so I sat down and wrote down everything from phase rotation, incoming wall voltage, voltage to amps etc. Half way through my scribling the damned HCAL on X came on again when the machine was in idle so now I am back to square one again. Only now when I disconnect the armature cables and jump s23 I can not "jog" the X like before. It only takes one pulse to get a 410 X servo error. But it can be reset and the HcAL does not come on.

I should also ad that half way through all trials I had a few HCAL on Y which is a first. They typically come on if I dont get them on X.

Does anyone know how to get to the commutor area on a 10M DC servo with pulse coder back?

Andy

[Al_The_Man]

You have to remove the encoder and strip the motor, usually 4 bolts after the brushes are removed.
Remove the armature carefully as the magnets are easily chipped or cracked.
Al.

[Andy Fritz]

Originally Posted by Al_The_Man You have to remove the encoder and strip the motor, usually 4 bolts after the brushes are removed. Remove the armature carefully as the magnets are easily chipped or cracked. Al.

Got the 4 nuts off the bolt shafts, but it doesnt come off. I cut the paint with a razor, but still no go. It seems that either I have to use some force which doesnt seem compatible with "carefully", or some thing else is looking the armature to the main body from the other side. Al, are you familiar with the 10m pulse coder motor? I worry that this might be one of those " strip it and its ruined" motors!

Cheers,

Andy

[Al_The_Man]

It depends on how the encoder is coupled, some are more fragile than others, you will have to remove the encoder before the brush assy.
Normally the stator comes off easily, you could try a dead blow hammer to tap on the rods to drive the end cap off, it is only the bearing that hold it on, the magnet force also tends to resist the attempts to pull the rotor off.
Al.

[Andy Fritz]

Hi Al,

If I tap the bolts it would force the whole motor forward toward the ball screw coupling. The only protruding object to tap the brush holder assy backwards would be the cannon connectors or four platic brush holder caps. Oh well I am sure it will work somehow. meanwhile I played a bit more with disconnecting the four armature cables for both X and Y axis. I also jumped pin S23 on both to prevent the TGLS alarm. When I power up there are no alarms. Moving X only one pulse throws Alarm 410 X servo. Doing the same with Y does not cause any alarms. As both motors are disconnected, wouldnt this suggest something else than the motor is wrong?

Andy

[Al_The_Man]

IIRC 410 indicates position error, which you may expect if the motor should be moving.
Do you have access to a Insulation Megger?
With the motor leads disconnected, you could megger between one lead and ground, if the megger output is adjustable, use the lowest setting, e.g. 500v .
Al.

[Andy Fritz]

Originally Posted by Al_The_Man IIRC 410 indicates position error, which you may expect if the motor should be moving. Do you have access to a Insulation Megger? With the motor leads disconnected, you could megger between one lead and ground, if the megger output is adjustable, use the lowest setting, e.g. 500v . Al.

The thing I dont get is why I only get the 410 on X and not Y. The conditons are the same. No motor is moving, only the numbers on the display. It is logical to get a position error as the control expects the pulse coder to confirm the display position, but then why do I not get the error on Y?

I dont have a megger, but I did check with the fluke meter between motor case and all 4 pins on the motor armature connector. resistance was around 4,5 M ohm. Next I checked signal to ground on the amp which gave about 2,7 M ohms and about the same between the two signals. I tested both with and with out the motor cables connected and it came out the same. the maintenance manual does not say what the resistance should be so I dont know, but there is defenately no direct short. The only thing I can not test is the above test with the X motor energized as I get the 401 or power up if it connected.

I will order a megger and hope it pays off because they cost more than a motor!

One thing that keeps popping up in the manual is that 220VAC can not be used with 50hz. Our wall power here is 50 hz and there is a little pin on the spindle drive to select between 50 and 60 hz. The manual also goes on about another frequency selector on the servo amp pcb, but that has an older pn than my pcb which does not have any hz selector. I have wired the servo transformer for 200VAC in, but the wall power is 205 - 220VAC depending on time of the day. All the DC voltages seems in range. The VCU pcbs are the same for X,Y and Z and as I can move Y and Z it should not matter unless there is something magical about the X.

Will have a go checking the feedback signal from the CNC to the drive...

Andy

[Al_The_Man]

An ordinary meter may not tell you if there is carbon tracking, it often show up with high voltage.
There is another way if you are careful to roughly substitute for a megger.
Use an ISOLATED supply at least 240v if possible and connect the motor frame to one side of the supply and the other supply end to around a 100w lamp, the other side of the lamp would go to one armature conductor, if the lamp lights brightly then you have tracking.
If no light, your motor could be OK.
Al.