MECHATROLINK-II interface

i never tried M-I, only M-II. i haven't found any manuals on the SGDM-DN either.
with MP2300 the 1ms comm cycle looks like this:
0x02ca in the sync packet is a 16-bit sequence counter. no garbage packets, everything is mostly by the book... edited init sequence log attached.

do i understand correctly that you check the CRC and all those random data packets are ok?

a while ago when i was messing with mitsubishi QD75M it also has random data in packets. it was just uninitialized RAM contents changing with each power-on. only a few meaningful fields were set in each init step. CRC was ok to. maybe the same is here?

Yes, I do check the CRC, it's CRC16-X.25. They are okay.
I'm also planning to check watchdog, since i've found the field layout.
Probably it will help to solve some minor puzzles. For example, sometimes servopacks respond to 30h (SMON) as if they were replying to NOP (00h command and data according to NOP layout) and vice-versa. I wonder if there's some kind of "delayed answer" possibility, i.e. to return response next during next transmission cycle. Though it doesn't sound real-time, which is the whole point.

RAM initialisation... Hm, gotta check whether those bytes change after a powercycle.

Anyway, I believe all Sigma-II drives support M-II, so this won't be a problem. How hard is it to modify your firmware to support some more commands (parameter read/write)? Our OEM didn't bother much about tuning and we experience some nasty vibrations (and of course their hell-of-proprietary software doesn't have tuning options).

I thought there would be some kind of PM on this board, but it turned out there is none. So I'll send you (AFAIU, sscnet@[your-personal-site]) another email.

answered via email, duplicating here for the reference
yes, amps do answer with a 2 cycle delay. picture from the manual attached
my card firmware supports parameter reading/writing while servo is on, i just omitted it for simplicity.
but i doubt SGDM drives support M-II. but i could take on implementing it if there's a reason to do it. i have the hardware on hand for debugging.

Oh, hi, i finally figure out this was you, haha.
I'm looking at a biesse with sgdb mechatrolink drives. its got 6 of em, so it seems better to try and keep them than replace, since the performance is adequate.

Whats the status of this card? if i buy this biesse, can i buy a board and run with no issues?

Thanks

i still haven't got an opportunity to test my M-I implementation with an SGDB drive.

so far it works ok with SGDS and SGDV set to an M-I 4Mbps/17byte mode. just yesterday tried with SGDM-xxDN's - there's an issue: drives showing a warning 95 right after controller connects, not every time and not sure why yet. i hope i will sort it out in a few days...

also tested with Omron R88D-KNxxx-ML2 (i believe it's made by panasonic) - the ADJ command is implemented differently and the card firmware can't query connected motor's native encoder resolution. added a module param to manually specify it as a workaround for now. otherwise seems to be functional. M-II implementation details are not in the manual unlike yaskawas. purchased myself such a drive, we'll see what i can find out about it.

as for SGDBs - pretty sure they will work. perhaps it will require some software tweaking during initial testing, but apart from that i don't expect any show-stoppers.

thanks

still not sure if i'm getting this machine, transport is turning out to cost more than the machine so far, oof.