Fanuc servo drivers with fiber optic

Hallo friends,
I would like to connect Fanuc servodrivers with FSSB interface (optical) to Mesa and LinuxCNC. I am afraid it is not easy First I checked if analog interface is still present in these drivers, but probably not. I still thinking to try to hack this optical link, but first I would like to ask if somebody has tried this before and would like to help me or cooperate. Or tell me it is mission impossible
Thanx for answer.

did that for mitsubishi sscnet iii which uses similar optical interface. it required some effort to get the data. packets are XOR'd with a pseudo-random sequence and then transmitted using a slightly modified 8b/10b encoding. took some time to understand how they generate that sequence to implement the RNG in hardware.

if fanuc didn't bother with some kind of real encryption then it is probably possible to hack given some dedication and yes, you'll definitely need a working controller to log the link...

i have an FPGA PCI card with 125Mbit POF transceiver which should be compatible with fanuc's

yurtaev.com/yssc3p.html

andypugh wrote:

Do you have a working optical controller and a way to log the transactions?

Yes, I have. It is in a new, cca. 1 month old machine - my friend's mistake We have a machine with LinuxCNC, but he needed next one and ordered machine with Fanuc. Now he see how stupid Fanuc controller is (compared to LinuxCNC) and he wants to rebuild it. I do not want to change servos, it is a lot of work and money, I preffer to change controller only. Problem is with optic interface.
I have no idea about data protocol and speed used (should be 50Mbit?!?), so my first idea was to connect oscilloscope to optical transceiver in controller and want to see at least something, measure the speed etc. and then decide next step. Unfortunatelly, I didn't found any testpoints and connect to optical transceiver terminals is a bit tricky, it means to diassemble the controller too much. Well, if there is no other way, I will do this. But first I am looking for way to connect some interface to optical bus and start there.

dm17ry wrote:

did that for mitsubishi sscnet iii which uses similar optical interface. it required some effort to get the data. packets are XOR'd with a pseudo-random sequence and then transmitted using a slightly modified 8b/10b encoding. took some time to understand how they generate that sequence to implement the RNG in hardware.

Thanx a lot for answer! Very good information that there is a way to solve it.

if fanuc didn't bother with some kind of real encryption then it is probably possible to hack given some dedication and yes, you'll definitely need a working controller to log the link...

No problem, I have full working system to log communication and many experiences with reverse enginnering, but I am not sure it will be enough

i have an FPGA PCI card with 125Mbit POF transceiver which should be compatible with fanuc's
yurtaev.com/yssc3p.html

Very nice! I thought FPGA is right way. Is it possible to get such card?

I have no idea how to do this, but Google seems to suggest it is possible

i used a splitter like this:
www.amazon.com/Digital-Fiber-Optical-Tos...litter/dp/B000I97G4Q
and regular toslink audio cables which are compatible with those duplex F07PN connectors

Is it possible to get such card?

yes, but it won't offer much tools for datalogging... though it will be easy to simply route RX signal to the header and hook up a scope/logic analyzer. are you familiar with xilinx tools?

dm17ry wrote:

I have no idea how to do this, but Google seems to suggest it is possible

yes, but it won't offer much tools for datalogging... though it will be easy to simply route RX signal to the header and hook up a scope/logic analyzer. are you familiar with xilinx tools?

OK, thanx a lot for suggestions. Yes, I can connect to electrical signals, but what about to use optical splitter like you and make a simple tool with optical transceiver for logging? Well, I don't have too much experiences with FPGA, I am working with MCUs long time and want to try FPGA. So it seems, the time is now What Xilinx tools do you mean?

what about to use optical splitter like you and make a simple tool with optical transceiver for logging

if you find a compatible receiver - why not? those japanese ones are pretty hard to source, but maybe it is possible to take a discrete photodiode with a bit of circuitry to get a readable signal. e.g. avago sfh250v

e.g. www.digikey.com/product-detail/en/broadc.../516-2283-ND/2391591

i have no experience with it though. or maybe you can just solder a couple of wires to the controller/amp PCB

if you get some logs to analyze - i will jump in

dm17ry wrote:

what about to use optical splitter like you and make a simple tool with optical transceiver for logging

if you find a compatible receiver - why not? those japanese ones are pretty hard to source, but maybe it is possible to take a discrete photodiode with a bit of circuitry to get a readable signal. e.g. avago sfh250v

Well, I know, transceiver in controller is Toshiba TODX2910, I cannot find datasheet, only similar ones. Byt similar transceivers should be possibe to buy.

e.g. www.digikey.com/product-detail/en/broadc.../516-2283-ND/2391591
i have no experience with it though. or maybe you can just solder a couple of wires to the controller/amp PCB

I know that soldering wires to the transceiver can seems to be easiest way, but not. You need to disassemble controller to the last screw (This I made one time) and anyway, I am afraid that connected wires can jam communication by some parasite capacity. Unfortunatelly, machine is running in production and I cannot block it for too long time.

if you get some logs to analyze - i will jump in

Great! Thanx for your offer. I will try to find a best way to log something and I will inform you.