Mesa 7i92 Installation, considering old ATX psu for I/O voltage source...

Good points Rod. The stepper options listed on moon's webpage did include several useful models.

The form factor on the DIN mounted Meanwell supplies are certainly more efficient, and I hadn't realized how
affordable the MDR series was...

I'll go ahead and order a MDR-20-5 on monday from Mouser, they have them listed for about $25us, so
at that minimal cost it will be easy to isolate my logic level power from any other general relays and such.
Should be money well spent, and drop some btu's inside that hardware enclosure.

My ATX psu will still be useful for testing/running other components but that MDR is a good fit with the Mesa cards.

*EDIT: I did some more research and answered several of the questions I had when I made this post, took out the
less relevant text to keep this post more concise and updated.

Long weekend of Linux Limbo wrapping my head around more bash commands and then reading thru the
Mesa Manuals several times more...



Stepper drivers and the other Power supplies have been ordered, I ended up going with;

Gecko G214 stepper drivers

Mean Well MSP-600-48, SDR-120-24, and MDR-60-5 power supplies

The 24v and 5v supplies will be specifically for the Mesa cards.


With regard to the Mesa hardware, I have a 7i92TF, 7i85S rev.B, and a 7i84 rev.A, all
of which have been mounted down in that order with aluminum risers to ground out to
the frame. I do not intend to connect the 5v common or 24v common to that frame ground.

(My current focus is now on the Mesa hardware install, but I am not powering any of it
until I have verified the following information...)


Mesa hardware configuration tree for my system as I currently understand it:

7i92TF:
    J1 <-- RJ45 Jack on Computer Motherboard
    
    W1 - UP, W2 - UP (Is this correct for DB25 connection to 7i85S ? )
    W3-DOWN, W4-DOWN (Pins 22-25 grounded on P1,P2)(Local 5v from P4)
    W5-DOWN, W6-DOWN (Default 192.168.1.121 address, may need to set W6-UP
              to ping 10.10.10.10 from EEPROM ? )
    
    P1 --> not currently used (may expand I/O later)
    P2 --> 7i85S (J1)
    
7i85S rev.B:
    W3-DOWN (J1 Pins 22-25 grounded)(Local 5V from P4)
    
    TB3 --> Differential Encoders 0,1,2(not connected yet)
    TB2 --> Differential Encoder 3, Stepper Drivers 0,1(not connected)
    TB1 --> Stepper Drivers 2,3(not connected)
    
    (TB1 Pins 17-24 --> 7i84 J1)(Smart Serial/RS422 ? )

7i84 rev. A:
    W1-LEFT (VIN connected to VFIELDB)(Can this introduce interference to internal logic?)        

    TB1 <-- (VFIELDA) 5V PSU, (VIN/VFIELDB) 24V PSU
                (should VIN have a dedicated psu?, all TB1 grounds are common?)
    
    TB2 --> 5V I/O (VFIELDA PSU)
    TB3 --> 24V I/O(VFIELDB PSU)
        


Additional questions/notes:

Should I do some form of power-on testing of the psu's for the Mesa cards to verify
there is not an excessive spike at power-up, or should the Capacitors in those
units take care of that?

What sort of input resistors/capacitors/diodes/fuses may be needed to keep these cards safe
with common I/O connections?

There will be various typical inputs/outputs to setup after these Mesa cards are connected, and
I do intend to source some form of differential encoders for a closed-loop feedback system, but
these items shouldn't have too much impact on the initial config. files for the Mesa cards?

When I do finally power up the Mesa cards, what should I watch for from the Led outputs?

Should this Mesa hardware combination be configured/powered on together or sequentially?




Once I have the information above verified, my focus will shift to handling Mesaflash, card
drivers, and Hostmot2.

Meanwell supplies will be fine, been using them for years.

I power mine together, actually I think mine are powered directly from the 7i92.

The Mesa I/O are pretty hardy and have no issues using them directly.

No need to worry about card drivers these, pretty much everything is handled via the hostmot2 driver & the hm2_eth driver, which are loaded via hal. Config is done when loading the hostmot2 driver.

Once the hardware side is setup, load the required firmware via mesaflash and then you can almost forget about mesaflash. The rest is getting the ini & hal files done. This is a mix of using the config program of your choice and editing by hand.

 
This video of mine will help.


Its for the 7i76e but its all similar.
 

Still waiting on various parts to arrive in the mail, but I did receive the 5v and 24v Mean Wells today.

I've watched Rod's video a few times and got some good info there, but also ended up looking thru a catalog
full of din rail terminal block parts afterward...


Currently, my work has been on how exactly to configure my 7i92TF/7i85s combination.
It looks like the simplest option might be:


"7i92t_7i85sx2d"


As that should provide the native card functions for a 7i85s, and therefore also have the serial channel for a
7i84d to communicate thru?

Just as a functional check, is that a valid choice for a new 7i92T with the

Trion T20F256 FPGA chip?

Using this configuration, would the encoder pins start from 0 on the P1 header by default, thereby designating
P1 as the connection point for my 7i85s?

If this checks out, I will try to connect the Mesas to my pc today with the network setup for something like a
10.10.10.11 pc address, and then, ping the eeprom 10.10.10.10 address on the 7i92T.

With the new 5v and 24v meanwell psu's wired up and my network settings updated, I pinged the 7i92T:


--- 10.10.10.10 ping statistics ---
48 packets transmitted, 48 received, 0% packet loss, time 48106ms
rtt min/avg/max/mdev = 0.071/0.099/0.158/0.013 ms


The 0.158ms was the first packet, and watching the output, most were very near 0.100ms which verifies
the 0.099ms avg. Does this sound ok for what I have running on this system?

All three cards were connected so that may have had some effect on the numbers...



There was also a solid red LED lit on the 7i84d, but I am guessing that was probably from a watchdog fault, I've
only just connected them so far and haven't done any other configuration yet. Supply voltage was good.

Those ping numbers look OK
(the first is typically longer because it includes an ARP response)

A red LED is expected on the 7I84D since it will have
(as you guessed) at least a watchdog fault when idle

Thanks for the Function check PCW.

Will I need to do anything special code wise during configuration on these, or should it mostly be the normal setup
and then just pin associations / various timing variables?
 

Working on understanding the exact mesaflash/pncconf steps for setting up the firmware/linuxcnc associations for the
7i92T right now...

I got a copy of the "7i92t_7i85sx2d" zip file from mesanet on hand

Also, already went into pncconf to get an idea how THAT interface is handling this setup. It appears the discovery option
does identify the 7i92T (so that should mean the TRION fpga 7i92 is properly handled?), but once I get to the 7i85S an
"unsupported device" error is generated. Perhaps some sort of ini help file or other redirection method would be the solution
at this step. Maybe a one card at a time approach is also part of the method, still figuring out the specifics for this.

Considering the firmware may be handled at this stage, I am cautious of editing anything yet, but I suspect this shouldn't
be too difficult to accomplish, its just a matter of identifying what exactly needs modification and where...

Alright, after multiple queries through the search feature on this forum I eventually found enough material to
reflash my 7i92t, then make a basic pncconf configuration for linuxcnc with these three mesa cards.

I will do further work with the configuration files today and update this post later to better document the code
 and input selections used in pncconf...