[sbaer]

I have a perfectly functional Brother TC229 with a conversational control that I would like to replace with LinuxCNC, I plan to use this machine as tool to make money and am prepared to spend several thousand to get it to work, including making the toolchanger and rigid tapping functions operate.

The servodrives seem to be some customized variant of a Sankyo Denki PZ series drive. Unfortunately a manual or even a pinout of the drives in this machine does not seem to be available, Sankyo Denki is not too interested in providing me with any documentation.

My plan is to build a special wiring harness (15 pins) to place between the controller and the drives to try to determine what signals are being sent between the two. Is this totally mission impossible? Has anyone else accomplished such a thing? what I can I expect the command signal from the controller to the individual drives to look like? The X, Y and Z amplifiers all have the same part number, where do the servo tuning parameters (which must be different for each axis) reside?

If interfacing with the existing drives is too ambitious, can I at least keep the motors and replace the drives with ones that have full documentation?

[sbaer]

Spent a few hours making a special cable to allow me to probe the wires linking the controller to the amplifier with an oscilloscope and have come to the conclusion that the there is some form of serial communication between the two. There is also the A, B and index encoder pulses. I could find any evidence of analogue signals or step and direction signals.

Stan

[Tkamsker]

Hi I did/Do similar of an Bridgeport 412 interact we ended up only using Old servos and use Generalmechatronics and granitedevices New encoder and you Are done i write an carousel Type atc component right now Thomas

Gesendet von meinem SM-N9005 mit Tapatalk

[Al_The_Man]

I believe these are controlled under the serial method called Sercos.

SERCOS (Serial Real-time Communication System).
Al.

[ihavenofish]

I believe these are controlled under the serial method called Sercos.

SERCOS (Serial Real-time Communication System).
Al.

is there a source for this notion?

i ask because i have a tc229 on its way and can find zero info on it and sercos doesnt seem to have been adopted widely until after my machine was built though i guess brother could have been an early adopter. my tc225 is analogue, and every other brother model of the era (mid 90's) seems to also be analogue.

[ihavenofish]

testing 123 - odd, this thread didnt get flagged as updated.

[Al_The_Man]

I have no info on SERCOS just remember it vaguely coming across it from a few decades ago.
Al.

[ihavenofish]

I have no info on SERCOS just remember it vaguely coming across it from a few decades ago.
Al.

ah, ok.

[ihavenofish]

edit: i didnt read well enough. :P

so, the chips on the motherboard are rs422, proprietary sanyo.

sanyo ga1022 chips.

heres the paper that mentions them:
https://www.sanyodenki.com/archive/d...TR28e_j_en.pdf
https://www.sanyodenki.com/archive/d...TR10e_d_en.pdf
https://www.sanyodenki.com/archive/d...TR16e_h_en.pdf

ga1045 and ga1060 are similar newer chips.

so, how are we going to get this to talk to linuxcnc? hmmm.

[mactec54]

Spent a few hours making a special cable to allow me to probe the wires linking the controller to the amplifier with an oscilloscope and have come to the conclusion that the there is some form of serial communication between the two. There is also the A, B and index encoder pulses. I could find any evidence of analogue signals or step and direction signals.

Stan

This may be of some help if you don't already have it

[ihavenofish]

page 84
http://www.sanyodenki.biz/products/I...7910-GA-en.pdf

talks about the ga1060 protocol. i suspect its very similar to the ga1022, just faster and more payload.

[ihavenofish]

yes and no. yes, its of help to see the signals in action. i can make my own though. no, its of no help in that it doesnt actually tell us the commands that linuxcnc would have to send it. otherwise stan would have done it instead of replacing his drives

it does seem that the encoder feedback is separate from the serial connection. so the serial is likely sending a speed command, and getting back some extra info like alarms, torque load, settings, etc.

[ihavenofish]

http://www.sanyodenki.biz/products/I...B_Manual_E.pdf

so, this may be enough info to begin some tests. i think the 1060 and 1045 are basically the same, but faster/more data per transaction. the 1022 is different in that it is not synchronous, but otherwise, it seems to be similar, but slower / less data.

might be good to try to splice in something like an arduino to "capture" a bit of the serial input/outputs as the machine turns on, then moves around. then we can see if we can get the framing and try to figure out the signals based on the ga1045 and ga1060 specs.

and if that does work, i hear yaskawa makes nice servos :P