Thanks Dave,
I've downloaded the manual for the Proma and it seems quite basic. It didn't seem to have any indication of a deadband (like the SPAN in the Candcnc unit). If it hasn't got any deadband I'd see that as a problem too, it would make Mach overshoot way too easily. As a matter of fact a guy on Youtube is using the Proma and can only get his THC Correction speed to 5% (Mach3), any more and he gets instability.
The latest Proma works a little different. It takes over control of the step and direction signals. So normally the step/direction signals come from Mach, etc, but when THC is enabled the Proma sends out the step/direction signals. No doubt the motor position will be lost in the controller. The latest Candcnc system (DTHC IV) seems to do something similar but I think the Z position is continuously monitored in Mach.
Surprising that the Proma unit is slow. Even basic microcontrollers run at megahertz clock speeds so would have thought internal operational speed would have been more than adequate. I believe the Proma uses relay outputs and small relays can energise in about 3 milliseconds. I'm going to also use opto-isolators to send the signals to Mach, so the THC up/down signals will be as fast as possible. From then on it's down to whatever bottleneck Mach3 gives me. Might also put a Lava parallel port card in the PC, they are supposed to be very fast and rock solid.
I was looking at my Z axis the other day and thought to get the most out of Machs basic THC, the Z axis needs to be built as "free" as possible. No stickiness on the linear rails or leadscrew, minimum weight design, and even a weight counteract spring to make the weight of the torch/carrier very little. Might also change the stepper for a servo motor which is supposed to give somewhat better performance with Machs THC.
Will be interesting to see how it goes.
Cheers,
Keith.