Mesa Encoder input-index signal always high, even with grounded input

Hello,
Today I ran into a problem trying to connect my spindle encoder via a Mesa 7i77.
The Pin encoder.05.input-index is always shown as high, even when I hardwire that input to ground. See the Picture of the Halscope.
The Index Signal is OK, I confirmed it with my Oszi. I have also tried quadrature mode and single ended mode on both +I and -I inputs, always with the same result.

Could it be that the Mesa card input is somehow fried? This was my first thought but on the currently free encoder 04 it gives the same behavior, but maybe this one is also toast. I did not test any other encoder channels, as there are the four axis connected. No issues there at all.

How is the encoder input-index connected in software? I looked at the HAL pins to see if maybe something else is overwriting that signal, but this isn’t even connected to a net, as far as I could see.

Would it be possible to use a different input (not an encoder input) as the Z-signal, as it is named in the Encoder documentation?

Since IDX4 and IDX5 don't work, that suggests a problem
with the muxed IDX signal for 4/5 = DB25 pin 13. This could
be a cable, 7I77, or FPGA card issue on DB25 pin 13.

Also to check the index, set the index inputto single ended mode
and use halshow to watch in encoder index pin. The hal index pin
should be low when the input is not connected and high when
the index input is grounded.

(in single ended mode the encoder inputs have  pullup resistor
and are inverting)

You cannot in general see the index pin in halscope since with normal
encoder resolutions, the pulse width is too short to be reliably captured
at the servo thread rate.

Thanks for the quick reply!
I will change the cable tomorrow; I have a second one where a 7i74 is connected. 
Depending on the result I will also try a different encoder.
I was using halshow most of the time, there was no reaction on the "LED“ when grounding the index or letting at float at 5V. I just thought a picture of halscope would provide more context. But nice to know that you cannot use halscope for the index signal, or at least not reliably.

Ok so i tried the following:
-swapping the two DB25 cables
   => sadly no change in behavior (since this would have been so easy to fix)
-changing my C-Axis from encoder 03 to 04 and Spindle from 05 to 03
   => C-Axis now turning endless during homing routine
   => Spindle Index actually visible in halscope on highest sample rate when turning by hand


I also tried (but failed) to rewrite my hal to change between the two DB25 ports on the 5i25 to pin the cause to the 7i77.
Then I created a test machine with PNCCONF to get the 7i77 connected through the internal DB25 port of the 5i25 (P2), but I couldn’t get it to work. After some googeling I now think this is embedded in the firmware of the 5i25, if that is so, is there a firmware where the ports are swapped (7i77 + 7i74)?
See readhmi attachment for current configuration.

What are my options if the 7i77 is damaged. I have a 7i73 laying around waiting to be mounted into a pendant. Can it be used to connect the spindle encoder? Or is there a similar card to the 7i77 without the I/O pins to use for all the encoders and analog signals for the drives?

5i25_7i77x2r.bit is a reversed connector version of the normal 77x2 firmware

If the 7I77 is damaged its likely repairable
 

It works!! Thank you very much, your help is very much appreciated!
I checked my current Bitfile (5i25_7i77_7i74) and flashed a file I found (5i25_7i74_7i77) into the fallback area.

After switching the cables around, I now can see the index signal changing in halshow when I ground it.
In this bitfile the "damaged" Pin 13 of P3 is a SSerial. How will this affect the use of the 7i74?

I have currently connected a 7i84 on Chan 0, a 7i70 on Chan 1 and a 7i71 on Chan 2.
In the future I want to add my 7i73 for a pendant and maybe one other card.
So I would not have a problem if some of those Chanels wont work, I just want to know to avoid headaches in the future.

If pin 13 is damaged on the 5I25 it would prevent the last 7I74
channel from working (though it may still work as it's the TXEN
line and if its stuck low, the sserial channel will still work)

Typically if 5I25 inputs are damaged, it means they have been
exposed to negative currents >50 mA. This will usually short
out the clamp diode on the bus switch, leading to a stuck low pin.

This can happen if you plug in the DB25 cable and this connection
joins two systems with different ground domains (say a machine tool and a PC)
This also why Pin 13 and pin 1 are common victims (they are often the first to connect)

I power all my external cards 5V input through a power supply in the electrical cabinet (the W1 jumpers are set accordingly).
The 24V ground is connected to protective earth (original Hermle installation). I connected my 5V ground to the existing 24V ground.
Would it be better to use a separate 5V ground to protect pin 1 or any other pins in the future?
Or is it the other way around and the ground of the PC internal 5V needs to be connected to the common ground?

Its more that all power should be off on both the machine
side and PC side when connecting the DB25 because of
possible ground differences. The 5I25 is likely fixable by
replacing the bus switch (74CBT16211CDGGR)
 

That explains it, I assembled and dissasembled the setup multiple times in the last month, to get the control electronics from the machine into my living room. 
I will take care in the future. I will also note that part number if I ever would need this pin in the future.