[eriksalo]

Hi;

I have a Fanuc Series 3 AC Servo Amp driving my spindle.

The Amp/Spindle works great but after prolonged (>20Minutes) of high speed operation, the spindle alarms and shuts off showing an error #2. For ~10 seconds prior to failure, the motor sounds different and it appears to hunt for speed. After the failure occurs, the spindle easily and freely rotates by hand.

"Speed is deviated from the command value due to overload and others."

The Amp is a model A06B-6044-H007 (Model 3) that drives a Fanuc AC Spindle Motor Model 3 (A06B-1003-B100, 5/7.5Hp). I have a 2 speed gearbox.

The systems works perfectly in normal / short operation. I make power with a rotary phase converter and can't run the motor to its max RPM. The max RPM is 6,000 in high gear but I only run at max 4,000 RPM due to (I think) current limitations. If I try and run fast, I get speed deviations and then the same error #2.

I'm assuming this is a heat related issue. Any advice on where to start looking?

Thanks; Erik

[emexcee380]

Erik,

I just solved a problem with Alarm 2 (Spindle speed deviation) on a Fanuc 0T control spindle drive unit by replacing the one IGBT ("transistor module" in the spindle drive base unit. I guess it had gotten weak over time even though it was not shorted.

First, check your spindle belts (unless it's an integral spindle). If they are slipping slightly that could account for the alarm.

If the spindle belts are in good condition and properly tensioned, you might want to replace the IGBT(s) in the spindle drive unit.

Be aware that older model IGBTs are getting harder to find. If you don't want to attempt this yourself, PM me and I can make a recommendation for a good repair source.

[eriksalo]

Thanks for the advice. Great idea to check the belts... they are ~30 years old and I don't think I've ever looked at them.

I see you can get the transistors and also the electrolytic caps new from China, haven't looked much past that.

[emexcee380]

Erik,

One bit of advice......avoid Chinese electronics at all costs. (Most are the epitome of "crap", IMO.) Contact AllCom Electronics - Wholesale Electronics Distributor for IGBTs.

[cjfisher]

Erik,

I'll second and third EME's advice on capacitors. I posted this a while back on this thread: https://www.cnczone.com/forums/fanuc...0061-b002.html .

"Avoid ebay caps - more to the point Chinese made caps! I would look at digikey.com. Personally, I prefer Panasonic or Nichicon. In addition to the Voltage and Capacitance, pay attention to the temperature (yours are 105 C rated), lead spacing (the wrong spacing won't kill you but the caps won't fit flush), and lifetime @ temperature rating. I generally buy the the longest lifetime @ temperature rated cap I can find from my preferred manufacturers. Digikey has a great filter to help narrow down the selection. I don't know where you're at, but if you live farther away (Digikey is in MN) I have found the fastest and cheapest way to ship stuff is USPS. Most electrolytic caps at the higher voltage/capacitances are rated at 20%".

Cj

[eriksalo]

I bought four new IGBT's. Fuji from Japan. Will see if they work!

For other debug, I checked all the fuses the manual recommend (F5, F6, FA, FB, FC, FD, FE, FF, FG), all good. The belts look good and reasonably tight. The spindle turns freely by hand when the system in in failure mode.

I experimented a bit more and after it fails once, it fail more quickly the second time. After a few fails, the spindle won't perform any motion at all without setting an alarm #2. Won't orient or turn at a slow speed (M3 S100 for example). When I command spindle motion, there's a bit of noise from the motor and then it throws the code. The spindle turns freely by hand when this mode is present. I let it sit for a while and everything's fine again for ~20 minutes of operation.

I'll check the feedback signals that are adjusted by RV18 and RV19 and will also take apart the spindle speed sensor and inspect. However, this really seems like a heat related issue. Hence the new IGBT's.

[eriksalo]

Further debug that points further to the IGBT's.

The manual shows six test points to look at:

CH3 and CH4 to 0 Volts:

Spec: Should be a sine wave with average amplitude ~2.5V and Peak-Peak of 0.36~0.5V
Results: See attached scope trace. Looks good.

Attachment 413710 Attachment 413712


CH5 and CH6 to 0 Volts:


Spec: Should ve 2.5VDC +/- 0.2V
Results: Measured ~2.4 and ~2.5V

CH7 and CH8 to 0 Volts:

Spec: Should be a square wave with low voltage of 0.4V and a max of 4.5V, waves should be of equal width.
Results: One of the traces was symmetrical but the other was not. See before and after scope traces.

Attachment 413726

I adjusted the CH7 to be symmetrical but still had a failure after some time running.

Attachment 413708

After the system failed once, it would not run at all and always alarmed #2. Will be fine tomorrow if previous behavior holds

Will wait for the new IGBT's and report back when they are installed.

Any other advice?

[harrimansat]

Check current signals, CH22, CH24, CH26,
this driver is know for bad current isolated amplifiers
CRU and CRV signals in diagram produces CH22 CH24 and CH26, ISA1 and ISA2 are isolated amplifiers connected to shunt resistors
Regards!

[eriksalo]

Well, still failure.

I replaced all four IGBT's. It was a bit fiddly getting the wires off the old and on the new. Whomever mentioned the strong need to label everything beforehand was right on. Would have been impossible without carefully marking all the (many) wires, bus bars and components.

Sadly, I get just exactly the same failure mode with the new transistors installed.

After 10-20 minutes of the spindle running it starts to "hunt" for speed, or that what it sounds like. There's a modulation in the sound the motor is making. After ~5 seconds of that, I get an error #2.

Happens with the rest of the machine moving or not. I thought it might have been related to my big air compressor coming on but further investgation shows that to be not the case. Also thought maybe when the screen saver came on was related but that also isn't it. Thought the IGBT's might be getting hot but it's pretty cool in my shop this time of year and the fan is working, blowing cold shop air across the cooling fins. Exhaust air feels cool blowing on my hand.

I'll check out the isolation amplifiers next.

Any other suggestions?

[harrimansat]

I had a nightmare with a similar driver and a intermitent failure
https://www.youtube.com/watch?v=goXOIAL01F0
https://www.youtube.com/watch?v=KKU_D26XOUk

[eriksalo]

"]I had a nightmare with a similar driver and a intermitent failure "

What was your result? Did you ever find root cause?

[harrimansat]

Yes, the symptoms were the same, motor is running and it makes a noise. When a current sensor is faulting it measure less current and the feedback increases current in one phase motor. Now is working. Check current waveforms. There is a way to put driver in test mode and check current waveforms when driver is failing, but you will need make a test connector o make some jumpers, disconnect dc fuse and connect a external power supply.
Regards
https://www.youtube.com/watch?v=hZHpamk8EFo

[eriksalo]

@harrimansat, thanks for your reply.

I had already ordered new replacement isolation amplifiers (A76L-0300-0035) based on your previous post.

Will get a scope on it before and after, see how it works.

Fingers crossed!

[harrimansat]

When you replace it clean well PCB, the area that is under the amplifier, to ensure isolation.
If you want to measure the fault, first connect the scope in current channels, CH22 CH24, if you have more channels CH26, but this one is calculated from the other two. Wait until the motor makes the characteristic noise and measure again. If you have a current clamp will be very useful to measure simultaneously real and measured current. In series 3 driver is 16.7 amp/volt in CH22 and CH24.
It measure DC current too, so if you remove DC fuse, and you connect ARS terminal to 0, applying a DC current to shunt resistors you can measure if amplifier is working properly, 16Ampers in one shunt resistor is 1 volt in respective channel, 1 ampere is 0.06 volts. If you have a regulated power supply you can make this test.

?
Good luck!
?
Regards
Miguel?

[eriksalo]

I replaced the isolation amplifiers and it didn’t help the issue.

Thinking I might have a hardware issue, I removed the spindle and 2 speed drive system. All of the bearings seem to be in good condition as well as the spindle assembly., Ran the spindle motor with no load. It ran for about twice as long, about 45 minutes before failing with an error 02. Same as before.

It seems something is getting hot on the board and failing.

Any ideas? The spindle controller is the only piece of electrionics that I kept from the original system, getting ready to replace it with a modern VFD and somehow figuring out how to do spindle orient without the Fanuc system

[harrimansat]

Your driver have 4 leds for error indicator?

- - - Updated - - -

Your driver has 4 leds for error indicator?

[eriksalo]

Yes, LED #2 comes on during failure.

[regina]

Following with interest. I have one of these drives I will need to use in the intimidate future. What manual and schematic are you guys using to troubleshoot?

[mactec54]

Hi;

I have a Fanuc Series 3 AC Servo Amp driving my spindle.

The Amp/Spindle works great but after prolonged (>20Minutes) of high speed operation, the spindle alarms and shuts off showing an error #2. For ~10 seconds prior to failure, the motor sounds different and it appears to hunt for speed. After the failure occurs, the spindle easily and freely rotates by hand.

"Speed is deviated from the command value due to overload and others."

The Amp is a model A06B-6044-H007 (Model 3) that drives a Fanuc AC Spindle Motor Model 3 (A06B-1003-B100, 5/7.5Hp). I have a 2 speed gearbox.

The systems works perfectly in normal / short operation. I make power with a rotary phase converter and can't run the motor to its max RPM. The max RPM is 6,000 in high gear but I only run at max 4,000 RPM due to (I think) current limitations. If I try and run fast, I get speed deviations and then the same error #2.

I'm assuming this is a heat related issue. Any advice on where to start looking?

Thanks; Erik

You main problem is the RPC they will eventually damage you cnc machine even with a CNC rated RPC, the best is a Phase Perfect which cost more than a RPC a RPC does a number on most of the electronic component's in Drives this is what you have to expect when using RPC for your power supply, they are also very inefficient

[harrimansat]

Check CH28 with oscilloscope with motor running. 0 to 10 votls is desviation from comand value. If error signal is error feedback singal, if is high for long time you will have error 2. Other option is set ARS to 0, short ARS with 0, and you will disable alarms and monitor what is wrong.