7i80HD+7i48+7i42TA+7i84 Mesa firmware request (morbidelli author 503 retrofit)

Thanks PCW, this got us our first axis movements on the Morbidelli since starting the project. Yes, I did want 2x modbus ports and 2x smartserial on this if possible. Could you confirm that this should work (with the RS422/485 adapters) through 7i42TA pins also?

The 7i48 doesn't seem to be reading the encoders. Tried with the servo's built in ones; oscilloscope sees a/b/z pulses; tried with a cheap lbd3806 A/B enc both at 5v and at 24v - again, the scope can see a/b pulses, but nothing from the 7i48. (differential/single-ended set appropriately for each different encoder).

mesaflash readhmid shows there are encoder modules but is it weird that 'Registers:' is 0 for MuxedQCountSel?:

Module: MuxedQCount
There are 12 of MuxedQCount in configuration
Version: 4
Registers: 5
BaseAddress: 3600
ClockFrequency: 100.000 MHz
Register Stride: 256 bytes
Instance Stride: 4 bytes

Module: MuxedQCountSel
There are 1 of MuxedQCountSel in configuration
Version: 0
Registers: 0
BaseAddress: 0000
ClockFrequency: 100.000 MHz
Register Stride: 256 bytes
Instance Stride: 4 bytes

PWM is working on the 7i48 and making servos move, just no feedback.

What have I missed?

24V ?

If the encoder has 24V outputs, that would likely damage the 7I48

Are these single ended encoders (5 wire) or differential (8 wire)?

Is the 7I48 jumpered to match the encoder type?

One way to check the encoder inputs is to disconnect them from the encoders,
set the input modes to single ended and the selectively ground the A,B and index
inputs while watching the hal a,b,and index pins with halshow. These
pins would be named something like:

hm2_7i80.0.encoder.00.input-a
hm2_7i80.0.encoder.00.input-b
hm2_7i80.0.encoder.00.input-index

 

Only the temporary test encoder was given 24V (directly, not via the 7i48) and then only briefly, so fingers crossed I haven't killed it.
The 3 encoders I need to use are the split output from 3x Yaskawa servopack, so physical encoder on back of servo is connected to the servopack, which provides a, /a, b, /b, c, /c outputs ("Divided encoder signals are output"). Unplugged from the 7i48, these are visible on an oscilloscope.
The single ended/differential jumpers were correctly set. 
I watched encoder.01.position, encoder.01.rawcounts, and encoder.01.counts (iirc) while the Y axis was moving and Y encoder was wired in to the 7i48; all encoder counters remain at 0.

The one thing that stands out to me on the wiring side is that there's no route for the +5V encoder supply from the 7i48, to go into the servopack. However per the attached diagram (screenshot from the servopack manual), no +5V line is shown. 

The 5V connection is not required (this is just for powering stand-alone encoders)

You might try the input test I suggested (this does require single ended jumpering)

will do