Does the 7i92-7i76-7i74 firmware support SSI ?

1-3 - It's clear

4. Drives with absolute encoders usually have a way to read this absolute position

What is this way?

Some can be SSI slaves, others have a RS-232 or RS-485 interface that is used at startup, and of course drives
with real-time bidirectional interfaces (Ethercat, Sercos, Ethernet Powerlink, etc) can provide real time absolute position
to the host as part of the process data interchange

Some can be SSI slaves - can you recommend some models?

Other interfaces - how can I read data from this interface to LinuxCNC into start position parameters?
I want use 7i92

Thanks!

I dont recall which specific drives do this, just that I have encountered them over the years
helping people doing retrofits.

Sorry for the late answer, can you mak 5i25_7i77_7i74 + ssi as well.

Rght now I'm working on the system with 5i25 as main Anything IO board.

freeby.mesanet.com/5i25_7i77x1_7i74x1D_ssi.bit

Hi,

Finally we have tested the firmwares.

On 7i92 it works good.
But on the 5i25 0-3 channels a lot have of noise. Especially when we power up the servo. See attached halscope screenshot - sharp spikes on the signal. Channels 4-7 work good. May be there's a bug in the firmware?


By the way we've found that 8xRJ45 connector's shield is not grounded, so until we connected it to the ground we had a lot of noize on the ssi channels. I've not found anything about that in the manuals.

Another thing - using pair of 7i92+7i77 we have some notches on the 4th analog out (other channels work fine). Notches appears mostly near 0 output value from one to ten seconds per notch. We have tried to change 3 different 7i77, changed the cables with no effect, but if we connect the same 7i77 board to 5i25 everything works smooth.

Like Andy, i suspect the noise in the the SSI channels is bit errors, probably due to a ground loop putting current through the 5I25 --> 7I74 cable and crosstalk in the cable (especially if you are using a flat cable)

I suspect its my fault as the SSI input has no digital filter so may capture short
noise pulses even though the SSI data rate is typically fairly slow (400 KHz being a common rate)

I'll make up a test bitfile with a digital filter on input as a test, what is your current clock rate?

The 7I77 / 7I92 error is very odd, I suspect its a driver bug of some kind since the sserial firmware
is identical on 5I25s and 7I77s and only the driver knows what a 0 values is

What LinuxCNC version is this? By notch do you mean the output does not change while the
analog voltage in in hal changes?

freeby.mesanet.com/ssid200.zip

7i92 and 5i25 7i77_7I74 SSI configs plus changed sources

These have a 200 ns fixed digital filter on the SSI data

On the analog output problem, this may be a solved issue, are you using the latest version of linuxCNC?, there was a sserial race bug found in the driver a couple of months ago (Doit Issued before all transfer data registers written) this would likely show up as corrupted data on a particular data register and would explain the 5I25/7I92 difference (different register write timing)