Need Help!- Fanuc encoder lines C1, C2, C4, C8...

[Jonne]

Fanuc encoders have resolver lines C1, C2, C3, C4. Is that data in
normal bit format or gray code? I need to change that resolve lines
compatible for Omron servodrives some TTL-McGyverning...

[Al_The_Man]

The C1 - C4 lines are a binary pattern than is used by Fanuc to produce a Pseudo Sine wave for AC commutation.
I had no success in translating the pattern in order to convert to BLDC to simulated hall effect as the pattern does not allow coincidence with the BLDC timing.
I ended up replacing the encoder with one with BLDC 8 pole timing pulses.
Renco have some on eBay right now for $20.00.
Al.

[Jonne]

Let's think. Copy C8 straight to A-channel.
B-channel = C4 XOR C8

CCCC Channel
8421 A B

0000 0 0
0001 0 0
0010 0 0
0011 0 0
0100 0 1
0101 0 1
0110 0 1
0111 0 1
1000 1 1
1001 1 1
1010 1 1
1011 1 1
1100 1 0
1101 1 0
1110 1 0
1111 1 0

0000...1111 = one full rotation

So we have now get quadrature pulse in A & B ( or R & S ) -channel.
Why that can work? Is there something what I miss?


PS. My servos are AC

[Al_The_Man]

It depends what you are trying to achieve, I wanted to use the Fanuc Red Cap (AC) motors with a BLDC analogue input drive.
I could not get the binary pattern to coincide with the required armature angle to provide commutation at the right time.
The Fanuc encoder is keyed to the shaft so apart from a few degrees adjustment, the encoder position is fixed.
If you can come up with a BLDC solution, that would be great.
Al.

[Jonne]

Oh, it is not normal bit code. I have to think that more.
Do you have more datasheet about Fanuc encoders? I'd
like know where Z-pulse is located conserning C1...C8.

[Al_The_Man]

The motors/encoders I have played with so far have a standard quadrature encoder with marker pulse, the four C1-C8 are extra tracks on the disk used for Fanuc commutation.
So I have the encoder output, I just need the BLDC comm. signals.
Ala http://users.tinyworld.co.uk/flecc/4...otor031102.swf
Al.

[jmelson]

Originally Posted by Al_The_Man The C1 - C4 lines are a binary pattern than is used by Fanuc to produce a Pseudo Sine wave for AC commutation. I had no success in translating the pattern in order to convert to BLDC to simulated hall effect as the pattern does not allow coincidence with the BLDC timing. I ended up replacing the encoder with one with BLDC 8 pole timing pulses. Renco have some on eBay right now for $20.00. Al.

I just developed a converter for these motor/encoders. The way I have it set up, it uses the 1-2-4-8 signals at first, until it sees the encoder index pulse, then switches over to counting the quadrature signals from the encoder. I have a simple 6-step drive that uses standard Hall signals, and this converter allows it to work with the Fanuc motors. This converter should work with any other drive than expects the Hall signals, as well.

I will be demoing a prototype at the EMC Fest in Wichita May 21-27, and should have a finished product out by the end of June or so.

Jon

[jmelson]

Originally Posted by Jonne Let's think. Copy C8 straight to A-channel. B-channel = C4 XOR C8 CCCC Channel 8421 A B 0000 0 0 0001 0 0 0010 0 0 0011 0 0 0100 0 1 0101 0 1 0110 0 1 0111 0 1 1000 1 1 1001 1 1 1010 1 1 1011 1 1 1100 1 0 1101 1 0 1110 1 0 1111 1 0 0000...1111 = one full rotation So we have now get quadrature pulse in A & B ( or R & S ) -channel. Why that can work? Is there something what I miss? PS. My servos are AC

The codes are not a straight binary counting sequence as you have written above.
It starts out like this :
0011
0010
0000
0001
etc. It is a form of gray code, where only one bit ever changes state at a time.

Jon

[Jonne]

Wait a minute... When I use AC servo do I need any HAL-outputs?

[Al_The_Man]

With DCBL just the three commutation pulses are needed.
For AC sinusoidal, there appears to be at least two methods, one is to use the three BLDC commutation pulses initially to detect the position of the armature, after which the differential encoder is used for the subsequent commutation.
The other method uses a two phase resolver only.
Al.

[Al_The_Man]

Originally Posted by jmelson I just developed a converter for these motor/encoders. The way I have it set up, it uses the 1-2-4-8 signals at first, until it sees the encoder index pulse, then switches over to counting the quadrature signals from the encoder. I have a simple 6-step drive that uses standard Hall signals, and this converter allows it to work with the Fanuc motors. This converter should work with any other drive than expects the Hall signals, as well. I will be demoing a prototype at the EMC Fest in Wichita May 21-27, and should have a finished product out by the end of June or so. Jon

I would be interested in knowing more about the converter, is this built into a drive or is this a circuit module that goes between the motor/encoder and any BLDC drive for example?
Al.

[jmelson]

Originally Posted by Al_The_Man I would be interested in knowing more about the converter, is this built into a drive or is this a circuit module that goes between the motor/encoder and any BLDC drive for example? Al.

Right now, it is a separate module. I may eventually build it into a version of my drives. But, what I have so far would be a separate module that would go between the Fanuc encoder and a standard drive. I make my own brushless drives, and they take the standard Hall signals, so I need a converter for any motor/encoder that doesn't produce those standard commutation signals.

Jon