Differential line encoder - spindle position encoder

The gear sensors I used for my homemade encoder are open collector, so generate a simple, single-ended signal on each output. However, most controllers require differential inputs at their encoder inputs. So some form of interface circuit is required.

I spotted something that would do the trick on the CNCdrive.com website. The differential line driver module looks spot on. However, I've not been able to get any response from them, despite 2 attempts to place an order and a direct email. Dunno if the guy has croaked or been locked up but if he doesn't look after his business there won't be anything to look after.

Dunno what device they used in their product but there are various parts available out there.

It looks as if an RS-422 driver should suffice, according to the Centroid manual. 

That takes a single ended 5V signal and generates a pair of differential outputs. The TI MC3487NE4 looks like an option. The DIP package will allow me to cobble it together on a piece of matrix board. It won't be pretty but I'm not in the mood for messing about.

The sensors are allegedly capable of operating down to 4.5V. I may drive them with 12V and arrange the pullups from 12V so that the input sees 5V in the high state. Something like this:

One concern is the recommendation for between 8000 and 40000 PPR for the encoder. That sounds like a pretty lazy and shit approach by Centoid. I suppose it makes the software pretty simple - just divide the encoder pulses by the pulses/rev for the servo and also by the ratio of ballscrew pitch : required pitch and send those pulses out to the Z axis servo. If you start with a large enough number (40000???!!), the sums are always simple. But at 6000rpm and 40000PPR, that's a frequency of 1MHz for each channel at 6000rpm. Other controllers manage to screwcut / rigid tap with ONE pulse per rev. I will find out how it performs later.....