Servo Encoder set up issue

You cannot reverse the analog output unless the 10V supply is floating
(it probably has a common ground with the encoder signals from the drive)

I'm not sure I understand what you mean. I'm simply swapping the polarity of the 10v via a relay to swap the motor direction.
This has always worked fine, even with my homemade encoder setup.
Its become an issue since I swapped the old encoder setup for the built in servo encoder. As soon as I pull the LPT plug the spindle rotates CCW as required. CW direction is unaffected. The issue appears to be something I've done recently when connecting the servo encoder.

Should I collect all the pins I've wired on the servo side and list them here?
The way I've used GND may be the issue here.

I thought it may be easier to explain if I attached examples of my current wiring.
Note, at the bottom of the schematic page I have included how my previous encoder was wired. This method worked fine with both CW & CCW directions. Because it was a little untidy and was picking up some noise I decided to make good use of the built in servo-encoder arrangement.

Your reversing relay connects VREF to GND and applies a positive voltage to AGND
when the spindle is reversed.

This is fine if the analog input and and PWM to analog output are floating,
but it looks like both the drives analog input circuit and the encoder output share a common
ground, which will short out the PWM-analog output signal when in the reverse direction.

What would be my best solution here? Would connecting the ABZ signals via an opto coupler taking 5v from the PC's USB work around this?

Also, please accept my ignorance here, but what is meant by the term 'floating' ?

Thanks,

"Floating" in this context means no connections to common
terminals (like GND)

5V from a PC is not floating because it GND is frame GND

If the drives AGND and encoder GND are connected on the drive
you should still be able to make this work if the PWM-->analog
circuit is fully floating (no GNDconnections on the analog side)

I've just noticed an error in my previous wiring diagram and attached the correct version. Previously It didn't show the supply from the PSU to the PWM-Analogue module.
I think I'm getting a grip of what you are telling me, but I keep staring at the wiring diagram and wonder where the actual issue/connection resides, and how to correct this.