Hood,

Why would you want to convert the step/dir signals to differential? Did you have problems with them being single ended?

Daniel

Differential signals are much less prone to noise and as my drives accept them I prefer to use them. I have never tried with single ended inputs as I believe it is better to wire in such a way that the chances of noise are minimsed rather than trying to fix problems at a later stage.
Hood

I have used C23 from cnc4pc and they work great and have all the features you will ever need, that being said they are horrible running on the smoothstepper. If you are uisng two/one parallel ports they are fantastic, the SS is really buggy it dose some strange stuff and I would not recomend it on a production machine, or if you are cutting expensive material or care about safety. If and when the unit comes out of beta which I doubt we will ever see, use the parallel port until it dose, its not ready for prime time as the beta version suggests. If you need really high stepping frequency you may have no choice, I dont know if the kflop is ready or if it has issues but might be a better alternative.

Works fine for me on both my mill and lathe, not mass production but used on a commercial basis. From what I see the main issues are with people that have a noisy environment.

Hood

It must like 50hz better lol.

Maybe but also a lot of people running on 60Hz with no issues

Hood

If and when a new update comes out I will give it a shot again, but too many little bugs at the moment and have seen it do dangerous stuff, I dont have any ground loops and the VFD is outside the case and its only supplied with isolated 0-10v and direction line of course. have not had issues on the parallel port as of yet, I just liked the idea of having it usb thats why I tried them, I was hoping they would be out of beta by now and the minor stuff fixed. I fear alot of people maybe left in the cold but I hope not, its like its just an update away from being very good.

WHat problems did you exactly encounter with the SmoothStepper?

Earlier I payed attention on the GREX, But I did not buy it as it was not ready for what I wanted. SS looked better, but I waited more than a year, till I decided it is good enough to do what I need. Then I bought 2 pieces. Have not regretted so far.
There are no bugs here, But I am not using it on the advanced level. (sometimes when I press buttons repeatedly and rapidly, SS can lockout, bot no other issues discovered)

- Noise isolation has surely to be done carefully.

- High freq Stepping signals have to be handled carefully. E.g. there are not too many optoisolators that can handle MHz ranges...
Don't expect 4MHz performance from a breakout board that was designed to a few kHz range!!! /as with PP/
For this reason I am making my custom BOB, so that it will fill all requirements.

- Tapping, Threading: I do not know how this is, I hope it will be ready by when I need it. I can't imagine however why is it hard to do it, what I imagine: threading/tapping code section should be handled copletely by SS, and it could stream out the step signals for feed directly based on the encoder input from the spindle... why is it so hard? (Mach generates the path, SS just does not give it out based on a standard clock, but rather based on the clock of the spindle input, which better be a quadrature encoder, not just an index signal)

- Expanded I/O: I hope in the near future tis will be available, too.

All in all I am satisfied with the SS, the only thing making me annoyed is that there is no update since the beginning of this year...

"and it could stream out the step signals for feed directly based on the encoder input from the spindle"

When?
Mach / SS - Just a quess, but the "horse and wagon may need to be able to travel together"
                and don't think you will see anything until Mach 4 is released.
                Only Warp9 can speak for itself.

There is a post on that site on requesting slaving the spindle to the Z axis, which could be done rather easily, but was limited to  only one direction and not both. Haven't pursued it since that posting which was months ago.

RICH

Well, that was rather a quick brainstorming, I do not know how exactly communication works, but that method would be quite straightforward, but I'm lacking knowledge about what Mach+SS do exactly, that's why I put the question "why'

Slavign spindle to Z? Hard to imagine, the other way around, slaving Z to spindle would be clearer for me. But it's possible that we are talking about the same thing.

Whatever, as I will have no chance to help the development of SS, I am just hoping that soon all those functions will work. (And the I/O expansion as well)