Gecko 540 0-10V output is voltage sinking which means it has to have a external power source... in this case the speed controller's power supply. Its a long shot but the power supply for the controller may not be capable of providing enough current for the Gecko. Could test the Gecko by using a 9v battery (Terminal 9+ and 8-) in series with a multimeter and see if the voltage drops. This will eliminate the speed controller's 0-5v.   

Good idea, I'll try that when I get home!

I'm still going to try it, but now I'm skeptical it will work.  Anything that fixed it would need to explain why restarting Mach3 fixes the problem, and is the ONLY thing that fixes the problem, plus the acceleration issue.

Any problems in the the spindle, spindle controller, or G540 should be able to be fixed by restarting those, and unaffected by restarting Mach3.  That's obviously the opposite of what's happening.  That really only leaves the SmoothStepper USB, the SmoothStepper USB plugin for Mach3, and Mach3 itself.  Which of course is why I came to post on these forums in the first place.

Can't you just follow the voltage level feeding the controller. Check the voltage level feeding the Controller. IF the voltage level is not constant follow it all the way back to the source.  What is feeding the logic to control the speed in teh G540 is it step/dir or PMW.  Check those levels coming out of Mach3 to the G540. I can't see that being able to droop to zero. (But never say never) The G540 just takes those values and converts to 0-+10 which is feed from the controller. The controller seems to be the culprit Check for a steady level voltage that feeds the G540 . IF the voltage level at that point is steady then it points to the G540 ?

Just a thought,

(;-) TP

I can see the PWM feeding the G540, and it seems ok, but I don't have an oscilloscope to see what the PWM signal is doing.  The way I understand it, the G540 is looking at the duration of the PWM pulses, whatever percentage of the pulse cycle is high (say, 50%) it then makes an output voltage that is also that percentage of the input voltage (such as 2.5V is 50% of 5V input).

OK then Check the output voltage from the g540 to the controller while running IS it drooping as time passes or does it remain steady ?? IF drooping then check the out voltage from the controller to the G540. It is steady or drooping. If steady then the problem maybe G540 IF drooping it maybe the controller Power supply drooping under load. IF needed install an outside PS to power up the G540 VFD voltage input.  Wallwarts are cheap. Just make sure it has enough capacity so it does not droop.

Follow the voltage trail,

( TP

I just checked, and the spindle controller supplies a solid 5.01V the whole time, it doesn't drop.  The G540 output voltage to control spindle speed drops slowly (e.g. 5.00V, 4.98V, 4.95V, 4.94V...).

So it looks like the spindle and spindle controller can be excluded as the cause.

Can you check the PMW with a volt meter ??  The higher the % the higher the voltage ? 

(;-) TP

Not really, I'd need an oscilloscope.

Have you tried it ?? Do you have another PC to test with or another  lpt port ?

Just a thought, (;-) TP