Help asked for connecting 7i76e + 7i85 + 3 sserial daughter cards

Dear Linuxcnc forum members,
Since this is my first use of the forum, I would like to start with thanking the community in providing this very nice software and all the support around it. I am trying to equip an old (1980) MAHO 300 NC machine with complete new servo's, wiring and of course Linuxcnc. For this purpose I procured a MESA 7i76e card and several MESA sserial daughter cards (1x 7i85 2x7i84, 1x 7i87).

I installed Linuxcnc a while ago (version 2.9.4), established the ethernet connection with the 7i76e and installed 7i76e_7i76x1_7i851D.bit file using the mesaflash tool on the 7i76e card itself. I think that this went well, see attached response on on a --readhmid command. I also made the 3 RJ45 connection cables which fit between the 7i85 and the individual sserial daughter cards (1x 7i84, 1x 7i84U & 1x 7i87). I also conclude that these cables are correctly assembled, as I do see the name of the daughter card and pins in a first machine configuration I made for the machine using pncconf. However.... They only work on one of the 7i85 sserial channels, namely the one identified in the MESA manual with Serial 4. None of the cards are visible when I plug these in in the other Serial connections on the 7i85. In pncconf I also encountered the "issue" (not sure if it is an issue or intended functionality...) that I could only select 7i76e_7i76x1_Dpl, where I had expected to also find something similar to the name of the 7i76e_7i76x1_7i851D.bit file. When I run --serial in mesaflash I also see that there is only 1 sserial  channel available, namely channel 6 (see second attachment). This while I had expected to find 6 channels in total (1 on the 7i76e and 5 on the mesa 7i85). During my own attempts to solve this I also changed the hal file line which refers to hm2_eth  to:loadrt hm2_eth board_ip="10.10.10.10" config="num_encoders=5 num_pwmgens=0 num_stepgens=5 sserial_port_0=00000000"Unfortunately this still doesn't allow me to connect a second sserial daughter card to any of the other sserial positions on the 7i85 other than Serial 4. Since I am stuck on this already for a while I decided to write this post in the hope that somebody is willing to help me out with some advice on how to proceed. Some specific questions:-What do need to do on the linuxcnc PC to correctly install the bit-file “7i76e_7i76x1_7i851D.bit”? I do see some info on this topic on: www.linuxcnc.org/docs/html/drivers/hostmot2.html but I lack experience to perform the right steps to do this install using the Synaptic Package Manager. I only have the zip file from MESA with the *.bin and *.bit files in them but don’t see this in the Synaptic Package Manager.
I also browsed to the "-Do I need to occupy the sserial on the 7i76e before I can put daughter cards on the sserial on the 7i85?
Many thanks in advance for your support.

A couple things:

1. SSerial ports will only show up in LinuxCNC  if a SSerial remote is connected and detected.
(so you would not see 6 SSerial channels in LinuxCNC unless 6 devices were connected and detected)
Only the low level hardware listings like readhmid will show all possible channels

2. With pncconf for unsupported firmware, you can use the "Discover" option

3. Do the other functions like the encoders work on the 7I85?

4. I can verify that firmware + 7I85 later today

Dear PCW,
1.) Ok, understood. Since I connected 3 boards (2x 7i84, 1x 7i87) I should thus see only 3 SSerial. I was confused as the pins from one of the 7i84 connected to SSerial position 4 on the 7i85 is visible under this name: "sserial device at channel 6: 7I84 (unit 0x1800C0A7)" (string taken from --sserial command using mesaflash).
2) Yes I read this also somewhere else on the forum and made a new machine configuration in pncconf using this option. If I look in the hal file made afterward I indeed see a modified config string in loadrt hm2_eth:  'sserial_port_0=0000000x' and the number of encoders increased. However, unfortunately the other 7i84 and the 7i87 are stil not visible (these are connected via other SSerial channels on the 7i85 other than the one numbered 4 in the manual.
3) This evening I wired an encoder and checked it on all the 4 available enc positions (0, 1, 2, 3). I do see the steps increase/decrease in the watch window of halconfig. The signal pin names I look at are: hm2_7i76e.0.encoder.00.count up to hm2_7i76e.0.encoder.05.count. The enc. 0 (identifier from MESA manual image) corresponds to ....00.count and enc 4 to ....03.count. All seem to work nicely (cool to see this work!)
4) Off course I would appreciate this a lot. I also looked at the naming of a digital input from one of the 7i84 when connected via SSerial 4 (MESA manual image) on the 7i85 and this shows as : hm2_7i76e.0.7i84.0.6.input-00 and so on. These digital pins also work in this configuration Again I don't understand where the 6 comes from. In case I connect the same board to any of the other SSerial positions on the 7i85 than they don't show up at all.

OK tested a 7i76E with 7i76e_7i76x1_7i85x1D.bit firmware:

md5sum 7i76e_7i76x1_7i85x1D.bit
0d0922e5c4306ceba4e077fff3c02097  7i76e_7i76x1_7i85x1D.bit

And it works as expected, that is sserial devices 0..6 work

SSerial channels are organized as follows:

0 Local 7I76E field I/O
1 7I76E Serial expansion slot TB3 15..20
2 7I85 Serial 0 TB2 9..16
3 7I85 Serial 1 TB2 17..24
4 7I85 Serial 2 TB1 1..8
5 7I85 Serial 3 TB1 9..16
6 7I85 Serial 4 TB1 17..24

 

Now is it possible you have a 7I85S instead of a 7I85?
(look at the paper label on the back of the card)

That would show these symptoms ( only the last sserial channel works
as TB1 pin 17..24 is the only serial port on a 7I85S )

 

Dear PCW,
Again many thanks for helping out. This morning (before work...) I quickly cross checked the label and you are right: I unintentionally bought/received a 7i85S rather than a "normal" 7i85....  . This evening I will order a correct 7i85 replacement. My excuses for bothering the forum with such a beginners mistake, yet I am very glad I received this hint as it would probably had cost me many more days (in frustration) to find it out by myself.
Cheers,
Henk