Fanuc absolute pulse coder pinout?

[jsheerin]

Hi all,
I bought a Fanuc 30F motor, A06B-0353-B831 that has what I think is an absolute encoder. The number on the encoder is A860-0320-T011. It says it's an 'ABS Pulse Coder, 2000P'. I have some pictures of it here: http://www.cnczone.com/forums/1054004-post261.html

Does anyone know anything about these, like the pinout or what the communication protocol is? If I can figure this out I might be able to use it with Linuxcnc, but if not I can just replace it with an incremental encoder.

[Al_The_Man]

AFAIK you cannot use these with Linux or most controls, if you replace with a standard quadrature encoder it has to have commutation pulses on it and it has to be aligned to the rotor position.
Digging a bit deeper, I discovered that the absolute encoders appear to be quadrature differential but have battery backed position retention?
But I still cannot confirm the nature of the transmission output, I suspect it may be serial?
Al.

[jsheerin]

Thanks Al.

You can use normal incremental quadrature encoders without commutation pulses with linuxcnc (previously EMC2 - now changed for legal reasons) and the bldc component. Andy, the author of bldc, tells me you can use absolute encoders as well but without knowing what the output is he can't say if the one I have will work or not.

Is the 6VA line for the battery? I have an alpha series manual that references that on an absolute encoder. Or is the battery backup in the drive itself?

If the output is the same as a normal quadrature encoder, I could just use it that way... I'd just have to figure out where to apply power.

[Al_The_Man]

The battery is usually in the drive, Some of the ones I found have serial communication, oddly they appear to be similar to the Tamagawa encoders used on Mitsubishi's, the pin terminology is even the same?
Al.

[tc429]

the older fanuc absolutes are very simple- not serial, I played around with one a while back, trying to get my led box to work on it- if you strobe pin 20 at powerup, the quadrature channels on pins16/17 and 18/19 'blink' up to the current position- looks like its rotating via quadrature... kinda a neat approach

[jsheerin]

Well, I'd say this counts as an old one. I think it's from 1987.

What are the pin numbers you reference?

Also, what do you mean by 'strobe pin 20 at power up'? When you say the quadrature channels blink up to the current position, do you mean the electronics in the encoder essentially output all the counts to get to where ever it last stopped? Do you have any more info on this? Sounds interesting...

[tc429]

Im thinking of the pinouts on the honda plug end...we only had these on a couple machines. Ive got a sketch somewhere on the amphenol plug, sure it should be in the 6050 servo drive book too, I'll try to find it today.
yes the quadrature just 'counts up' to its last position. the strobe to synch up is like a 12v 20 msec pulse IIRC, but its been a long time, I dont think it was 5v/ttl though... all the incremental absolutes worked this way, even the red ones. the newer serial absolutes were just serial A/C/alpha with a battery added to the plug, they synch thru the serial link of course, all are pretty high resolution.

[Al_The_Man]

Originally Posted by jsheerin Well, I'd say this counts as an old one. I think it's from 1987. What are the pin numbers you reference? .

J, what is the MS number of the connector?
Al.

[jsheerin]

Al,
It's the 19 pin connector, the 22-14.

Tc, if you can find more info that would be great.

I think the pinout might be the same as on the newer absolute encoders (alpha series). Looking at that manual, the C1-C8 commutation signals are on the same pins, and the other pins seem to match up as well (connected 0V pins). I say this so far only from checking resistances between pins and comparing it to an incremental Fanuc encoder that I know the pinout for, but the commutation pins have exactly the same resistances to ground. So if this turns out to be the case (hopefully I can check tonight - I'll have to wire up a plug and degrease the motor before putting it on the bench...), then I would have A,B,Z and the differential pairs, the 4 commutation channels, +5V, 0V, shield, REQ, +6VA, 0VA. So I assume the +6VA and 0VA are for the battery, and the REQ line is what tc is talking about putting a pulse into? If anyone can confirm that would be great. Otherwise I'll play around and post whatever I find.

[jsheerin]

Thinking about this some more, if it works as a quadrature encoder then I could really just skip the absolute part... All my other axes will have incremental encoders so having an absolute encoder on one axis wouldn't be that big of an advantage.

[tc429]

didnt have a scanner, but heres the signal/pinouts for the ms3102a22-14p

A/A
*A/B
B/C
*B/D
Z/E
*Z/F
C1/G
C2/H
C4/J
C8/K
+5V/L
0V/M
SHIELD/N
OH1/P
OH2/R
REQ/S
+6VA/T
0VA/U

the 'REQ' is the one thats strobed at bootup- didnt get a chance to look for my notes, maybe someone with a ABS equipped machine can toss a scope across pin 20-0v on the honda plug side and see what it is- seems like it *should* be a 5v signal, or maybe just a ground pulse, but I for some reason still think it was like +12v for 20msec to initialize the encoder- in reality guess the strobe length probably dont matter, and just thats what the control puts out... IIRC you wont get any quadrature until the REQ is strobed though.

brainfart- hey just thought of something- the 6050 amps have a jumper (S31
or 32?) its the only one on the right/lower part of the pcb- its set high for ABS encoders- we never use it that I can think of, but I know its there... I bet the amp generates the strobe when ENB goes high, immediately before sending VRDY back...the 5v to the encoders is generated at the VCU supply, its just commoned to the motherboards 0v... might make sense that the initialization would occur synch'd with VRDY so the drive holds counts as it 'counts up'

on the 6057/6058 digitals, the abs bit is set in the parameters at bits APC/APZ or something like that, but those take encoder plugs direct to mother board, not passing thru the drive to leave the C1~C8 commutation signals like on the 6050s...

[tc429]

just looked over your project thread- cool

hey, if you could get this working, why not do the same for your other axes? weve got a boatload of the old 30f/20f motors stored away off scrapped 420 robots, and probably 500~1000 of the black 2k abs/taper shaft mounted encoders (with resolvers and weird plugs - can be cut off) that we pulled from 380 robots we scrapped... heck we might have several thousand of them- literal hoppers full. a buddy of mine saved me about 50 of the 30Fs for nothing (but he pulled the encoders- likely mixed red and black style) I was going to use for windmill/generators- he might still have them too...

we also have a boatload of the 6050 h005 drives too- but they all have the goofy -R00 topboard, the 3-7 jumper for reverse rotation config wont work, and from what Ive seen gain linearity is a big tuning mess compared to the last E/F revs...they will drive the big motors, but linearity would likely cause distortion when doing any circular interpolation... they are all 185 volt input, +/- 10 analog, the encoders all pass thru the drive- they both leave the commutation signals at the drive AND theres a onboard freq/volt converter so it generates it own tach off the quadrature slew rate...they were awesome drives- but the robot R00 version wasnt anywhere as linear as the later ones...I ran some in a mill once with 0M-A control, ran into the issue of not being able to reverse the electrical rotation and some linearity, so never tried again... I have one in a 'jogbox' for moving analog fanuc ac servos using just 120 volt cord input, its handy for folding unpowered robots or jogging stored machines, its used pretty often, been in there at least 10 years...
the old 6050 will run at reduced power on single phase- just need minimum of 170 volts on terminal A and 1, then jumper 1-2 and it fools the phase detection... I run the 0M-A bridgeport in my garage with (6047 DCs) the same way- can rapid all 3 axes at 100 IPM without tripping a 10 amp 120 volt single pole breaker...