Mach3 XML and Stepconfig issues.

Yes we do, looks like pause is a bit shorter.

Is there a way to tell what the output from Mach3 looks like?
Do you see variation in voltage at the converter output?

Are both systems using the same parallel port hardware?

Same hardware
Will have to rig oscilloscope to verify.
Have tried fine-tuning the LM circuit maybe a bit better but still like a bad two-stroke.
Think it is time to rebuild with better BoB that has PWM.

It shouldn't be necessary to change the hardware. If it works with Mach it should work with LinuxCNC.

Is the stepgen.N.frequency pin showing a steady value?

The halscope plot should look something like this. (This shot has 2 increases in frequency from me clicking the increase spindle speed button)

Appears to be so.

Below my settings in Mach3 for the spindle.

We should be able to replicate those settings easily enough. Go back to the original type-0 stepgen config. With these hal line settings.
setp stepgen.2.steplen 5000
setp stepgen.2.stepspace 5000
setp stepgen.2.dirdelay 13000
setp stepgen.2.maxaccel 1875

And if it still is bad, try inverting parallel port pin 6 by adding this line.
setp parport.0.pin-06-out-invert 1

Do you happen to have a real oscilloscope? You could try looking at what the real signal from Mach3 actually looks like. (Mach has little reputation for not actually doing what it says it is.) Then try to replicate it.

I think I have tried that but will try it again, but not the inverted.

Have scope, will try to maybe tomorrow.

I know, it seems like a long shot, but at this point we're flailing in the dark.

Still same with updated hal, but...



Have to comment out that line to get started, but I doubt it will help anyhow.

Will try to rig scope and see what Mach3 spits out.

Cheers
Bengt