Mitsubishi SSCNET

Do to circumstances I am no longer able to work on this project. I still have access to my machine but it is essentially in storage for the moment.
dm17ry was very close to having something for me to try out but I did not have time.
I don't know if he is still working on it.
If you are in the states, I have one of the SSCNET cards that you will need.
If you choose to move ahead I can sell it to you.
If you are anywhere near Milwuakee WI. I would be glad to help move this forward.

Hi. Thanks for the reply. Unfortunately im quite far from WI. Im in South Africa.

I am interested in taking this forward but i would like to understand a bit more about what it involves before i decide.

Can you PM me some info about the card you have and any other relevant documentaton?

Thanks
Henk

if dm17ry does not see this. I will send him a note it a few days to rattle his cage.
He is the one with all of the information on the Mitsubishi Drives.
He designed a PC card to communicate with these drives that you will need.
Since you are not near me, you can probably get it from him easier.

The last thing that I needed to do was to set up a computer and see if his card would initialize the motors and move them around.

I had one of these machines that I could put my logic analyzer on and capture the communications between the drives and the controller. I would send these captures to him and he did the magic.

I am still on the lookout for a smaller version of the same machine. Mine was just too big from my new shop.

hi!
i suspended the project for a while due to lack of new hardware to test on, until recently. a couple of guys found me - one with MDS-A-V1 drives, another with MDS-B-SVJ2. and neither of those liked my protocol implementation but the guys helped me a lot with debugging and i think i've nailed the problem finally. i will confirm it in a few days.

i have rewritten the firmware and working on the driver now with multiple servo protocols in mind. there'll be some results soon, i hope

Great news! Looking forward to hear from you.

Henk

I hope we can use MR-J4W2 series servo soon. This kind of servo make easier to handle double y-axies.

i also hope... it's been a long time

meanwhile, i've got the basic spindle support finished. verified on MDS-B-SPH: initial parameters upload, velocity control, velocity feedback and spindle-at-speed signal. position control is there too, but have to test it yet. MDS-A-SVJ and MDS-B-SVJ2 do work. no absolute encoder/linear scales yet.

still couldn't get the only MDS-B-V1 drive i have a (limited) access to working. keep getting an AL36 after initialization. have to find a working setup to snoop at the protocol...

also implemented the SSCNET-I protocol. this should allow to control up to 8 (probably 15) MR-J2-B/MR-J2S-B drives in position/velocity mode. for now it uses a 3.55ms cycle, but i will try to get it down to 0.88ms later. at least for J2S drives...

Anything I can help just tell me. The hardware is always cheaper in China than in Russia. MR-J3/MR-J4 is a huge progress.

thanks! i've got everything i need for sscnet-iii for a while now. just have to find some spare time to take on it. hope to do it soon.

btw, how're your multi-axis machines?

i finally done multiple cards support...

a bit of update for whom it may concern. implemented the spindle control, both velocity and position. position control may require some external HAL glue logic to switch gains in the amp, the pins to do it are present in the driver. so MDS-A-SVJ, MDS-B-SVJ2, MDS-A/B-SP should be functional. will begin more extensive testing next week.

and also i finally got the SSCNET-III interface card mostly working. need to finish the research on the protocol, try to fit everything still missing into 0.44ms (or better yet 0.22ms) cycle and implement required changes to the linuxcnc driver. few other things here and there and then all those J3/J4 amps off the ebay should be usable