Drive control question?

The auxiliary encoder inputs on the servo are differential. The 'Tie' terminal should be jumpered to the A- and B- inputs when you're connected to a single-ended signal. You should also tie the servo's DC common (pin 14) to the AD's DC common even if you don't use the servo's 5VDC supply to power the AD's pulse output.

The CW/CCW mode means that a pulse train at the A terminal will command motion in one direction and a pulse train at the B terminal will command motion in the opposite direction. I repeat, I know nothing about the AD module, so in my ignorance I'm going to suggest that you make sure you're not outputting an A quad B signal from the AD module.

To make sure your configuration of the servo is correct, try connecting/disconnecting the servo's +5VDC to the A input to see if you can generate any motion. You may have to set a fairly high master/slave ratio to be able to observe much motion from a single pulse.

I do not know the hardware you're talking about, but on other setups, I have seen simular problems a number of times.

As a couple of people here suggests, the signal level(s) are not read correctly. Either its due to a tie problem as suggested or a signal level problem.

With an oscilloscope you can monitor the signal and verify that the duty cycle for the signals are 50/50. Most puls inputs will accept some variation, but not too much (gets more critical as frequency increases). Furthermore, with the oscilloscope you can monitor that the logic "1" input has a level that is read as logical "1" by the drive. As with the logic "0".
If levels are not reached, which is a very common problem, you need to tie up (logic "1") and/or tie down (logic "0").

e.g. Single ended inputs, problem with logic "0" signal level:
Tie down logic "0" by inserting a 1k ohm (5V signal level) or a 2k ohm (12V signal level) resistor between signal A+ and A- (which then is connected to common, which in turn is connected to ground). Do the same for B channel.