I can't reverse the spindle rotation.

PCW wrote:

I would check the output of the direction control PC817 with a voltmeter/ohmmeter

I checked for continuity between the V output and GND using a multimeter, and even though the LED is lit, there is no continuity. 


Tomorrow I’ll check the resistance at the output. It’s all very strange because I have two brand-new modules, and they behave the same way. 

They always worked perfectly on a parallel port, but I no longer have the PC with a parallel port to test them again...

The Isolator is doing 5V to 24V translation so unless the BOB has a relay/24V output
it cannot go directly to the VFD

Sounds like maybe  grounding issue (or damaged OPTO)

Note that a parallel port ground is nearly always frame ground
but the 7I92 ground may be floating
 

The servomotor uses the ECAC XP200 driver, and according to the manual, the digital inputs can only be controlled with 24V.

I also have the ground for the 7i92 tied to the common ground.

Alright, bit of progress. I did some testing and managed to get the PC817 module working — at least for now.

The module has a 302Ω resistor on both the input and output side. When the LED turned on, the output voltage only dropped to about 21V, so the servo driver wasn't seeing it as a proper low signal.

I tried putting another resistor in parallel with the one already on the output side, and bingo — reverse works. With a 3k resistor in parallel, the output drops to around 14V. With a 1.5k, it drops to about 10V. Both values make the servo reverse properly.

So my question now is, what's a safe resistor value to leave in there permanently? I don't want to damage anything, but I need it to work reliably. What would you guys recommend?

Honestly it sounds like you should be using a Darlington OPTO in place
of the low CTR PC817, say a LTV-815 or similar.

( PC817 mas a minimum CTR of 50% vs the LTV-815's 600% minimum )