[jeck]

Hello friends,

Here I am on the machine with Fanuc spindle unit A06B-6052-0004 and it gives ALARM 12 (8+4) (High current DC Link) when starting M3 (M4) S300 and above. If I start S100 it begins to rotate jerking, pulsing and if I began to increase S by 100 it can go turns to S1000 and then gives Alarm 12. But if I start S300 it gives directly Alarm12.
I measured the resistance of the motor- it’s OK , to Earth also by megaohmmeter there’s not problem. I measured by Ohmmeter Transistor power module – A50L-0001-0209 : between B1-B2-U0, B3-B4-V0, B5-B6-W0 the measurings are the same. (But actually I don’t know how to measure that Transistor module.)

I found a post in another theme:"FANUC OT B Spindle rpm fluctuating and causing 1008 Spindle ...": of : tmbruno28 "ok, first thing, you have to check sensor signal by an oscilloscope turning spindle by hand . you have check points at the bottom of the pcb, just near CN2 connector .they are called Pa,Pb, Ra, Rb. You must found a sinusoid of about 3.5 volts on the a, and continuous 3 volts on the two others. "

So I started S100 M3 and measured by multimeter between Pa and Ra – 0.03V. Between Pb and Rb - 0V.
Unfortunately I don’t have with me oscilloscope and couldn’t find oscilloscope here , but if you can give me some advice – is it possible that Pulscoder is not OK ?

[jeck]

Some more information from my investigation on our AC spindle unit.
I measured all fusses (there are 7 fusses 0,3A on the firing plate) – no problem fuse.
Also, I measured Transistor module : each of U,V,W to P, N it the right and reverse direction. It shows me 1,2Mohm.
Another thing, I found some maintenance for Transistor module (GFZ 65015E) http://support.ge-ip.com/support/res...en_US/4240.pdf - my measures are the same, except B1,B3,B5 to U,V,W and B2,B4,B6 to N : in right and reverse directions – 160 Ohms, but may be the Transmod is Allright (may be the manual is for another type Transmod).

Is it indicative for serious problem in the AC spindle unit that in idle at low speed - up to100rpm. the spindle shivers (interrupts for an instant) at every turning? At 200rpm. and higher this shivering becomes indetectable, unseen.

[mike_Kilroy]

1x per rev 'shiver' at low speed is often indicative of an OPEN output transistor; so ur ohm measurements will not show it. can u give M3 S10 command? slower than 100rpm?? often if told to move slow enough an open transistor will cause the rotation to stop rather than shiver if it does not have enough inertia to roll thru the open spot. this too can cause current cspike on next phase output to try to make up for missing pulse so at high speed give overcurrent fault.

can u measure output of each transistor while running slow speed to see it turn on/off compare all 3?

[jeck]

Thank you for contribution Mr.Kilroy,
Yes, I have tried to order slower speed than 100rpm. (S50, S30) and system receives the command, but there wasn’t obvious difference (I mean the number of rotations per minute stays the same - about 100).
I fully understand yours considerations and I think they are enough reasonable.
I will measure the phases at the possible lower speed (in our case – about 100). I think you mean by oscilloscope (you can correct me if you mean another measuring). It must be U, V, W – towards Earth – am I right?

[mike_Kilroy]

tnx FOR THE im - I will reply here and there

actually I mean from dc bus PLUS and dc bus MINUS; reading to ground doesn thave much meaning. If u go slow enough then u can see the outputs switch on and off on multimeter. sometimes u can just power up, enable drive, and rotate motor by hand slow (so encoder feedback goes to drive and makes the outputs switch thru their sequence too...

[jeck]

Thanks for the prompt about rotating by hand, it's realy tricky, Mr. Kilroy,

I’m not arrived to the machine yet, I’m away for a couple of days. Meanwhile something appeared in my mind: I remind that the AC spindle unit gives that kind of alarm (12) in two directions of rotation (M3 and M4). It remind me that eventual problem “OPEN transistor” couldn’t be only one..
Or I’m not right?

[mike_Kilroy]

not sure what u r saying but yes, open transistor tick/stick would happen in BOTH directions.

[jeck]

I wanted to say that I supposed that if there’s only one open transistor, it couldn't influence in two directions (M3 and M4). I supposed that only one open transistor would influence only in one direction.
But you said that “open transistor tick/stick would happen in BOTH directions”.
If I’m not trouble you and waste your time, could you say how it happens. I thought that there are 6 transistors for M3 and 6 for M4 directions, and the M3 transistors don’t affect to M4 transistors and vice versa.

[mike_Kilroy]

no, there are 6 total sistors - to go other direction drive just switches which phase leads which - like swapping 2 motor leads. all 6 used for both directions.

[jeck]

Thank you Mr. Kilroy
I think I manage to consider why I have brought myself.
Last case we handle with more powerful AC unit “FRENIC” of FUDJI electric where they have 12 power transistors. (But it’s was not DC link powered but 3 phase powered.).
whilst that FANUC Transistor Module is DC link powered.
As we see on http://support.ge-ip.com/support/res...en_US/4240.pdf
I think it should have has 6 transistors (thera are 6 Bases B1....B6) – so obviously they use all of them in both directions.

[mike_Kilroy]

http://www.ctiautomation.net/PDF/FUJ...Mega-Guide.pdf

this is from the Frenic manual above: 6 transitors total. perhaps this manual does not cover the really large models and maybe they are line regen so have 6 on the input side also?

[jeck]

"they are line regen so have 6 on the input side also" Yes I think it was such case: Frenic 5000 V2. (15kw). Insteadn of the diodes in your image there are 6 trans. called TS, and other called TM (I don't know what of abbreviations are these.) I obviously made big mistake concluding that TM are for the right direction and TS for reverse.

[mike_Kilroy]

Dear Mr. Kilroy
I would share with you something about this another case – Frenic5000V2. Your clarifying inspired again an idea in my mind. In that drive we have replaced for the 5th time the same transistor from that REGEN group. Obviously there is some problem in the control board, but nobody is going to repair such old drive in our and the nearest countries.
I discuss with some colleagues what if I just disconnect that transistor and some of them said that it will destroy some of others transistors. Now, thanks to your explanation, I realized that this group of transistors activates only when stopping the spindle (REGEN group).
I would ask you about your opinion: What do you think about the idea of disconnecting the problem transistor?


hmmmm... never heard of more than ONE xsistr on DB circuit.... even our 300amp drive has just 1 big one. no need for more than 1. db regen just shorts dc bus momentariy so there are not 6. so I do not understand why u see 6 extra xisstors.

[jeck]

Dear Mr. Kilroy,
May be I'm not correct when calling "REGEN" about that group of 6 more other transistors on the 3-phase suply side. I will send you the diagramm of that unit.

[mike_Kilroy]

no diagram rcvd. post here instead maybe

[mike_Kilroy]

yep, saw ur schematic. simple set of 6 igbt for motor output and 6 igbt on ac input to make the dc bus. that extra 6 igbt front end was done for 1 of 2 reasons depending on the drive requirement it was designed for:

1) for full line regen: if motor decels and puts power back into drive, those 6 igbts will push it back on the ac line

-and/or-

2) to make a variable dc bus: this was called PAM for pulse amplitude modulation to reduce the harmonics on the motor output current: for instance, if motor only needs 30v the input makes the bus 30v rather than chopping 700v down with very skinny pwm pulses to get 30v rms equiv, which adds extra heating to the motor.

I know of only 1 drive that did both; and another that did PAM for reduced motor heating, all others were for line regen only.

Not sure how this is going to help your problem of having a humpy motor at slow speed in both directions that smooths out when inertia carries it thru a missing output igbt pulse.....