Fanuc Red Cap AC Motor Model 0 hot

[Andy Fritz]

I have this motor driving the C axis on a Wasno Lathe and it the chassi gets hot even when the motor is idle. It also makes a high pitched noise and some times it starts to occillate giving a humming sound which increase in intesity over time. I tried playing with the axis drive RV1 gain and RV2 but even with Rv2 set to give exactly 0V between CH1 and CH3 pins on the drive and gain lower than the manual recommends at least the high pithed noise does not stop. The occillation can be stopped with some tweaking, but comes back at random depending on the position. The current at idle is 1,2 amp on two phases and 2,3 on the third leg mesured with a clamp meter. The other motors which are larger (model 5 and 10) only pulls 0,2 amp at idle and their chassis are cool even after hours at idle. I meggered between chassi and each phase and got about 210 Mohm on two phases and 174 Mohm on the third. I also checked between the phases and got only 1,6 ohm? Are they meant to be shorted when not under power? Next I meggered the power cable and got 1200 Mohm and even higher between phases.

There also seems to be some mechanical problem as the vibrations and noise changes depending on the position of the axis. For example; between 0 and 150 degrees there is mostly high pitched noise and between 160 - 300 degree there is lots of vibration when jogging. Maybe some shaft is bent or could this be a problem with the pulse coder?

When I pulled the red cap off I noticed some moisture on the inside of the cover.

What is the best way to pull the rest of the motor apart with out getting into problems?

Any ideas why the motor pulls so much current at idle and gets hot?

Andy

[Al_The_Man]

It sounds like the symptoms are the motor is overloaded, at rest it has a problem holding position, the resistance check is normal.
A test for this is to run the motor mechanically disconnected from the C axis to see if the symptoms persist.
Al.

[Andy Fritz]

Originally Posted by Al_The_Man It sounds like the symptoms are the motor is overloaded, at rest it has a problem holding position, the resistance check is normal. A test for this is to run the motor mechanically disconnected from the C axis to see if the symptoms persist. Al.

Al, thanks for the advise. I just did this and found the following;

The motor does not get hot at all and the noise is gone!
I checked current also which is much lower;
Black wire; 0,18 amp just after power up and 0,5 amp after jogging 2 minutes.
White wire; 0,04 amp just after power up and 0,1 amp after jogging 2 minutes.
Red wire; 0,04 amp just after power up and 0,1 amp after jogging 2 minutes.

So there seems to be something with the black wire phase.

Eventhough the noise is gone, just holding the shaft I can feel some oscillation. Basically 0,7 seconds vibration, 0,4 seconds no vibration, 0,7 seconds vibration and repeat 9 cycles then 1 second vibration 0,2 seconds no vibration and repeat the whole cycle. I guess this vibration gets amplified through the mechanical contact with the axis. So is there an adjustment to be done or is the drive bad?
Could it be the pulse coder is not in phase with the motor and some drift compensation is struggling to keep them aligned?

Andy

[Al_The_Man]

The encoders are keyed on the shaft so there is only a slight adjustment by slackening off the the 3 screw on the outer ring.
It sounds like it could be a partial short on one of the phases.
If you spin the shaft by hand as fast as possible with the connector off, and it shows resistance at some point it could indicate a partial short due to the braking action effect of the short.
Al.

[Andy Fritz]

Originally Posted by Al_The_Man The encoders are keyed on the shaft so there is only a slight adjustment by slackening off the the 3 screw on the outer ring. It sounds like it could be a partial short on one of the phases. If you spin the shaft by hand as fast as possible with the connector off, and it shows resistance at some point it could indicate a partial short due to the braking action effect of the short. Al.

I disconnected the power cable and if I give it a good crank by hand it only free spins between one to one and a half turns measured from the point I release it. It does make a normal bearing sound and nothing seems to grab.
If it should be possible to crank it so it free wheels more, then I guess something is not right.

When I measure resistance I get only 1,5 - 1,7 ohm between the legs on the motor connector.
Is there another way to check if there is a partial short between phases?

Yesterday I removed the pulse coder and motor back flange to have a look inside. There was some orange sticky stuff on the windings and a thin film of this stuff on the inside chassi wall. I dont know if this is old coolant or some insulating material to prevent a short? Anway I marked the position of the pulse coder to get it back in the same place, but exactly how accurate a pen marking is, is another story. What would be the expected effect if the pulse coder is a little out of position? I guess it is 180 degrees off I would get some alarm.

Also about meggering; No manual says what an acceptable resistance between power and ground is. I had a look in some generic manual called "A stitch in time" by Megger company. There is a 1 Mohm/1000V rule. This is a 85V motor. Maybe they mean 1 Mohm per 1000V DC testing voltage? Anyway if my megger is actually working it seems my motor is ok???

Andy

[Al_The_Man]

If you have a 'scope you can connect a 1k resistor to each phase and make a star point from the other ends of the resistors and use this as a reference for the scope ground lead.
By backfeeding the motor at a constant rpm, 100~200 rpm is enough, the scope should show all three phases of equal generated amplitude.
There is a method of ensuring the commutation alignment is ok by locking the rotor in place with a DC voltage and looking at the level of the four comm. pulses.
The encoder can be off 180deg and still be OK if it is an 8 pole motor.
If you need it I can post it, I need to know the size of the motor, i.e. 0, 5, 10 etc.
A megger should ideally show at least 1meg to ground.
Al.
OK, I see it is a 0.

[Andy Fritz]

Originally Posted by Al_The_Man If you have a 'scope you can connect a 1k resistor to each phase and make a star point from the other ends of the resistors and use this as a reference for the scope ground lead. By backfeeding the motor at a constant rpm, 100~200 rpm is enough, the scope should show all three phases of equal generated amplitude. There is a method of ensuring the commutation alignment is ok by locking the rotor in place with a DC voltage and looking at the level of the four comm. pulses. The encoder can be off 180deg and still be OK if it is an 8 pole motor. If you need it I can post it, I need to know the size of the motor, i.e. 0, 5, 10 etc. A megger should ideally show at least 1meg to ground. Al. OK, I see it is a 0.

Hi Al,

Thanks for the advise. As I get near 200 megaohms between leg and ground, maybe I should buy a more expensive megger?

I do have a scope, but I have never used it. It is a 25 mhz digital with two channels. I could go through some on-line tutorials quite fast to get a rough idea how to use it.

The motor is a Fanuc Red cap Model 0. It is 85V, 2000 rpm, permanent magnet, 8 pole, 4,6 amp stall, 2,9 Nm torque (stall). The pulse coder is 2000P. The drive is a Fanuc A06B-6050-H102 with a A20B-1000-0560/11F top PCB. The X axis has the same drive, but the top pcb has a different revision number and some of the jumpers are not the same. As I dont know the difference I have not tried to feed the problem motor with the X axis drive.

What I would need to know is what you mean by "star point", how the trim pots should be set on the drive top pcb when doing this test and which pots to turn if the result on the scope does not look like it should.

If you can walk me through this I will owe you one, big time!

Andy

[Al_The_Man]

You do not use the drive for this test, or any power on the machine.
With the motor off the machine, I drive mine with a DC motor on the bench to back feed it.
The star point is the junction of the 3 resistors connected to each phase.
You can monitor two phases at a time with a double beam scope.
The test for aligning the commutation is to connect the U & V phases across a DC supply that will supply ~ 6amp for the 0S motor.
The connection as follows U +vdc & V-vdc, the W phase also goes to -v
I use an automotive battery and resistor to limit the current.
This aligns the armature under the U phase for alignment.
You then supply 5vdc to the encoder and monitor the status of the four comm lines,
When locked they should show all high (=1) or all 1 and C8 = 0 or 1.
YOu see the pattern change as the encoder is moved.
Al.

[Andy Fritz]

Originally Posted by Al_The_Man You do not use the drive for this test, or any power on the machine. With the motor off the machine, I drive mine with a DC motor on the bench to back feed it. The star point is the junction of the 3 resistors connected to each phase. You can monitor two phases at a time with a double beam scope. The test for aligning the commutation is to connect the U & V phases across a DC supply that will supply ~ 6amp for the 0S motor. The connection as follows U +vdc & V-vdc, the W phase also goes to -v I use an automotive battery and resistor to limit the current. This aligns the armature under the U phase for alignment. You then supply 5vdc to the encoder and monitor the status of the four comm lines, When locked they should show all high (=1) or all 1 and C8 = 0 or 1. YOu see the pattern change as the encoder is moved. Al.

Sounds tricky, but I will give it a go.

BTW if you would disconnect a healthy AC motor from a machine but keep it connected would it be normal for it to vibrate or should it be rock steady?

After some tweaking I now have it slightly vibrating, but now it is constant and not on/ off like before. Unfortunately when connected to the axis it is noisy. There is a little play on the motor mount and I have tried to move the motor a tiny bit both towards and away from the C axis gear, but if I get it to a position where the noise and vibration stops, it starts again after jogging the axis a bit.

Andy

[Al_The_Man]

Either the gain may be off or the encoder need a slight tweeking in alignment.
Did you try marking the encoder and then rock it slightly either side?
Al.

[Andy Fritz]

Yes I tried that while the motor was off the machine and it did not do much difference. I tried all possible positions of the pulse coder, but still vibrations.
As it only vibrates when off the machine but does not get hot, and when on the machine it makes the high pitched noise and gets hot as it does not have any play to vibrate freely so it seems stopping the vibration all together is the first step.
I mustered some guts and connected the C motor to the X drive and vice versa.
The X motor seemed like normal and the C motor vibrated, so that rules out the drive.

As I dont have a DC motor with rpm control available to do your test I will try to follow the manual for the AC drive. It suggests using a DC power supply to feed the star connection up to the motors rated DC current which 9Amp. The manual does not say anything about resistors. Should I use 1Kohm per leg like you suggested? It says if the phase relationship is good I should see high on C1, C2, C4 and C8 indicated as a "1". Is this in diagnostics or what does it mean?

Andy

[Al_The_Man]

To test the commutation alignment you need the high current DC source, automotive battery etc, and a resistor to limit the current, the 1k were for the rotation test.
This has to be a high wattage one as the resistance of the stator windings are very low.
As mentioned, the battery DC+ is connected through the resistor to U and V & W connected to DC-.
You can usually get away with much less than 6amps to lock the rotor in place.
If you do the test using the encoder plug connected and power on to obtain the 5vdc for the encoder, then you need access to the C1 to C8 connections to measure the level from encoder common to each commutation pin with a volt meter.
The level should be 1 on each, C8 can be a 1 or 0 and still work.
Al.