just another plasma build and then some :)

he has all kinds of video's and was looking a little this morning and it looked like he was using a 7i76e with the 7i85 s so if this is the case it should be no problem to add the THC cards to it down the road if this sorts the lag issues out on positioning. Linux cnc is one of the few systems that can do this from what i have seen.

glass scales are out for me but magnetic scales look like a perfect fit for my application. my traverse speeds are not crazy fast so the reader heads should pulse fast enough for my machine easily.

i like my coffee so i got that part covered that's the easy part

found his project over here and this is the link to it.it looks like he had the hybrid steppers with encoder feedback already and made a dual loop feedback system? i seen mention of PID so this a research project from the looks of things.

forum.linuxcnc.org/49-basic-configuratio...le-closed-loop-setup

was reading a white paper on closing the loop on steppers and seen the real magic happens in the drive's. this paper made reference to running them as a 2 phase brushless servo ? i might check with gecko to see if something like this can be done with their 320x drives? they are about the same cost as the stepper drives honestly. basically you need a drive that knows what to do with the feedback from the encoder. but if something like that would be something that would work changing the drives out would be money well spent to have a machine that can position to the same resolution as the encoder your using. by using a linear encoder you bypass the the backlash and screw lead variations all together which is what i am after.

Make sure you get some feedback from Stefan about the gecko Servos. I don't think they are a good choice.

the 201x drives i have are pre set to 10 micro steps and they have no adjustment in that respect so it takes 2000 pules to get a rev out of the motors. right now i am running off the parallel port but that's just temporary for now so i can sort things out somewhat. once you go to a motion control board i think you get better performance out of the machine as a result.

if they would miss to many steps and alarm out that could be a good thing to a degree in the sense the machine stops so you can fix the problem before it's to late to save a part. lost steps are just a short coming of stepper motors in general.

one other thing that got me thinking pretty hard on the topic was getting that long screw to hit a size on a part 5 foot long. i used a set of 24" digital calipers and worked my way through 2 foot of travel 2 inches at a time and the lead inaccuracies can stack up fast. they actually would rise and fall over an average so looking at a longer section of travel is what it took to get larger parts to come out right. getting things tuned on the long axis was a real pain in the you know what that's what sparked the idea of linear encoders to track actual position rather than a rotational drive calculation .

i thought i seen something about fault signals in the manual for the 320x? i will have to go back and look at it some more i just skimmed through it real quick. for a servo dive it is not fancy by any means but it would work with everything else i am using with no problems on the DC power supply and amps on the motors i have now. i just don't know if it would work with a bipolar stepper motor or not? it works off of just step and direction signals from what i seen at first glance. this is a mixing of stepper and servo parts and becoming a real frankin mess of sorts