[kalawa]

We have a Mitsui Seiki HR5B horizontal machine center with fanuc AC spindle motor A06B-1012-B-100, drive A06B-6044-H112 and fanuc 11M control.In the last few months it has been intermittently functional and very sporadic.

All the time though it would stop with error 38 - spindle speed abnormal. At times it would run perfect at 200 rpm but would alarm out at 1000 rpm and sometimes it would do the opposite. At the same time the loadmeter would stay at maximum for a longer period than normal at start up before it alarms out. Turning the control off the loadmeter would stay at around 50% and motor turning a good half a minute unless the maiin switch is turned off.

Having minimal understanding of cnc I would assume though that the parameters must be ok if it function ok at times. One of the encoder cable wire was found frettaged and had only a few strands left so we install a new encoder with new cable with no effect to the problem. By the way the encoder compartment behind the motor was half filled with oiil. The motor was mega tested ok though.

When it is running ok, it would then sometimes make noises different to normal running and coincide with fluctuations of the loadmeter. The odd sound appears to be coming from the motor and at the back of the drive assy behind the pcb.

I would appreciate all the help given to resolve my machine problem.

[CHAD LAWSON]

If belt drive is there an external speed sensor mounted as part of the spindle. It would be a magnetic pic running of a gear with a lot of small teeth. If so i would check it out should be gapped at about .003" and be very smooth and clean on the front. It also must be mounted square to the gear or the amplitudes will not pick up properly. Usually would use oscilloscope and check for +-5vdc amplitude on your feedback. Also some of the older fanuc controls required external inline amplifiers for the feedback signal i have seen many times where connections to these became corroded. Sorry if i am of no help not specifically familar with machine you are working on.

[kalawa]

Thanks for the suggestions Chad. I have to find an oscilloscope to carry out the signal checks. We have tried re-positioning the encoder to 0.003" and as right angle as possible to the teethed gear with no success.
But behind the pcb, where the fan and switches are, we have found the terminals caked up in dust particles. We will clean it up before carrying out further tests.

Cheers

[kalawa]

After hours of cleaning out the spindle drive, there was a 103m capacitor which is attached to the mag contactor with one leg broken and 2 500ohms resistors also broken. Repaired the capaitor and replaced resistors but upon running the machine, the same problem persists.

Have checked Pa and Pb with an oscilloscope and would get 2.5v and around 0.5v wave amplitude when machine is running at 0% load. The motor sound would suddenly change and beome noisy iwth an increase to loadmeter up to 50% (motor running without the gearbox). The Pa, Pb waves would also change becoming fuzzy and rattling on screen.

Is it possible that the motor itself is playing up?

All help would be appreciated.

[fanucman]

uh sensors have nothing to do with this.. alarm 12 is overcurrent in dc bus..

remove motor power cables from spindle drive.. give mo3 s200 command.

if alarm remains you have a bad drive meaning you need to check all the transistors inside (the large black modules) if you find one bad and they are in parallel you need to replace all 3..

the top spindle board might need to be replaced also if any of the fuses are blown..

[kalawa]

Thanks fanucman.

Removed motor cables,gave m3s200 and error 39-abnormal spindle speed appeared. Checked all fuses including FA-FG all ok. Checked the resistances of the transistor modules with the following results(there are 4 in this machine)

C1-E1 infinite C2-E2 infinite
E1-C1 330 ohms E2-C2 330 ohms

C1-B1 infinite C2-B2 infinite
B1-C1 485 ohms B2-C2 480 ohms

B1-E1 560 ohms B2-E2 540 ohms
E1-B1 560 ohms E2-B2 540 ohms

The above readings are consistent to all 4 modules. Info I was given is that positive probe to C,negative probe to E should give several 100s ohms and infinite when probes are reversed. If that is the case then all the transistor modules are faulty. Would appreciate your proffessional opinion.

Thanks in advance

[fanucman]

ok so meg the motor.. it needs to be higher than 100 megaohs phase to ground, do all phases

check the motor with a regular ohm meter between phases and make sure you get like .3 ohms or something close..

this might be a bad motor..

also check all the clear fuses on the spindle board..

[kalawa]

Hi fanucman

All phases hover around 100-150 Megaohms whilst cranking and goes over inifinity when stopped and 0.4 ohms between phases.The spindle pcb fuses FA-FG were all tested ok.

The only other thing worth mentioning is that the leads out from the motor windings are a bit oily which suggest that oil leakages from the valves above may have somehow got into the windings. I am just not sure if it is worth pulling the motor apart.

Thanks

[hcm1500]

To kalawa

This sounds like a faulty drive to me. The speed sensor was a good start and it seems you have that set correctly, as long as your waveforms are staying the same amplitude (the waveforms changing amplitude indicates the gap between the gear and the sensor is getting bigger and smaller, if thats the case you could have run out or some other problem). The motor checks sound good. If the drive is faulty it can cause the motor to speed up and slow down erratically.

Two things I thought of

1. Where did you meg the motor from?
It sounds like you did it at the motor (which is best) but if you did, hook the wires back to the motor and disconnect them from the drive and meg them to see if you get the same readings. (I had a similar problem and the condiut from the control cabinet to the motor was filled with coolant, I have also seen cables deteriorate internally and short)

2.You stated that you checked the transistors but I wasnt clear which meter lead was where during your tests. Use this

The instructions are given primarily for an NPN transistor as these are the most common types in use.

1.Set the meter to its ohms range - any range should do, but the middle ohms range if several are available is probably best.


2.Connect the base terminal of the transistor to the terminal marked positive (usually coloured red) on the multimeter


3.Connect the terminal marked negative or common (usually coloured black) to the collector and measure the resistance. It should read open circuit (there should be a deflection for a PNP transistor).


4.With the terminal marked positive still connected to the base, repeat the measurement with the positive terminal connected to the emitter. The reading should again read open circuit (the multimeter should deflect for a PNP transistor).


5.Now reverse the connection to the base of the transistor, this time connecting the negative or common (black) terminal of the analogue test meter to the base of the transistor.


6.Connect the terminal marked positive, first to the collector and measure the resistance. Then take it to the emitter. In both cases the meter should deflect (indicate open circuit for a PNP transistor).


7.It is next necessary to connect the meter negative or common to the collector and meter positive to the emitter. Check that the meter reads open circuit. (The meter should read open circuit for both NPN and PNP types.


8.Now reverse the connections so that the meter negative or common is connected to the emitter and meter positive to the collector. Check again that the meter reads open circuit.


9.If the transistor passes all the tests then it is basically functional and all the junctions are intact.


The final checks from collector to emitter ensure that the base has not been "blown through". It is sometimes possible that there is still a diode present between collector and base and the emitter and the base, but the collector and emitter are shorted together.

Even if all the transistors are good you could still have a problem with the drive. Let me know what you find

[kalawa]

Thank you for the instructions tcm1500 and we have carried them out with following results.

1. Motors re-megged from pcb end and remained same.

2. Transistors were re-tested as per your instructions with same results as previous. In brackets are the norms as per your instructions.I also removed all connections to do measurements on transistor modules with same results.

+C1-E1 open(open)
+E1-C1 330 ohms(open)
+C1-B1 open(deflection)
+B1-C1 485 ohms(open)

+B1-E1 560 ohms(open)
+E1-B1 560 ohms(deflection)

From the above results we can deduce that the transistors are faulty. The readings are consistant to all the 8 transistors in this machine. Is is it possible then that all these trasistors can deteriorate so bad that they malfunction all at the same time?


Looks like I'll have to start looking for new transistors. Thanks so much for the advise and has deepen my understanding of cnc machines.

Thanks

[kalawa]

To tcm1500,

I've ust recieved a manual for the spindle drive. There are a few discrepencies between manual limits to your instructions which are in the second brackets.

+C1-E1 open(open)(deflection of several 100ohms)
+E1-C1 330 ohms(open)(open)
+C1-B1 open(deflection)(deflection of several 100 ohms)
+B1-C1 485 ohms(open)(open)

+B1-E1 560 ohms(open)(deflection of several 100 ohms)
+E1-B1 560 ohms(deflection)(deflection of several 100 ohms)
The actual readings taken would still ender my transistors faulty from both the manual and your limits, I would appreciate if you let me know the reason in +C1-E1 and +B1-E1 being different in limits.

From the manual, after removing fuse 7, I have also measured the pcb base-emitter voltages on all 8 transistors from from connector CN6 & CN7. With normal limits at -0.8 V to -1.3 V, all readings were zero except CN7, pins 12-13 which measured -0.2 V. From the above I am satisfied that I have faulty transistors.

Cheers