That seems a bit short

I have a device that measures voltage by doing a voltage to frequency conversion then counting the pulses on the parallel port pin.

I have found that about 40 kHz is the frequency limit that can be reliably counted (essentially captured in a DO loop).

A 5 microsec pulse is equivalent to 200 kHz I think you are missing some