PlasmaCam loses Y axis location

[severedjosh]

Hello, new to Plasma Spider. A friend recommended me posting here as I've exhausted all other avenues I can think of.

I have an old PlasmaCam machine. Model 98-Z, It's been great up until a few months ago. It will initialize great, finds home easily. I've changed belts, inspected teeth, realigned the gantry, used compressed air and blew out the motors and encoders. My problem is as the y axis travels it slowly looses position on the table, the computer shows it in the correct position though. I've attached a 4" circle I've been using a sharpie to check the alignment with not only does the end point not match up, but it will ovalize the circle as well. I'm at the end of what I can think it is other then buying a new motor. I've called PC but they say my machine is to old and the offer no support. Any help would be greatly appreciated. As of now, after every cut I re-initialize and sometimes this makes for a LONG cut time...

circle.jpg

V/R
Josh

[adbuch]

If your table uses closed loop servos, then you may have a bad encoder. Otherwise possible slipping of gear of pulley on motor shaft.
David

[severedjosh]

If your table uses closed loop servos, then you may have a bad encoder. Otherwise possible slipping of gear of pulley on motor shaft.
David

I'm leaning towards the encoder, I have both marked the pulley and shaft as well as tried to put a little counter pressure on the Y axis when moving to try to get some slippage. I haven't been able to see or feel any slippage... My next step was changing the motor orientation and seeing if the problem moved to the X axis.

V/R
Josh

[adbuch]

You may be able to find a replacement encoder.

https://drive.google.com/file/d/152CpOB ... sp=sharing

David

[weldguy]

I would be leaning towards a loose pulley but since you marked it and verified it's ok I suppose the encoder would be the next culprit.

[severedjosh]

I would be leaning towards a loose pulley but since you marked it and verified it's ok I suppose the encoder would be the next culprit.

Probably a dumb question, but how do you get the encoder off the motor? I've turned the slot on top to the unlocked position, tried to disengage the clips and pull the cover up thinking it has mounting bolts on it. No success with either, I wasn't trying to break it as it's still making money....

V/R
Josh

[adbuch]

This if for the DHC/DHC2 motor encoders. Perhaps it will help.
David

removing encoder.jpg

installing encoder.jpg

20241203_154804_resized.jpg

20241203_154819_resized.jpg

20241203_154836_resized.jpg

20241203_154852_resized.jpg

20241203_154905_resized.jpg

20241203_154926_resized.jpg

20241203_154939_resized.jpg

[tcaudle]

The biggest challenge is the 4th picture down abd
1. Finding a small enough allen wrench
2, getting the tiny set screw on the disk to allign with the slot so you can loosen the set screw and you can slide the assembly off the shaft>
3 If you buy a new encoder it will probably come with the allen wrench stuck into the set screw . DO NOT REMOVE IT . getting it back in is a challenge of your patience!

If an encoder is the issue its either because its got debris in body and onto the disk or the disk is loose OR the disk is damaged . The way servos work is they send back 4 pulses per line on the disk and the controls count those and compare them to the steps of moment and decide if the motor shaft has moved all of the steps . There are two channels and two sensors. A and B, By counting the order the pules come in (A then B or B then A) tells the controls which direction the shaft turned and how far.Typically a bad encoder ensds up causing a fault (position fault) I guess if some lines were not getting read the software would keep spinning inot it did get the correct number of counts and that would make it move to fdar . TO get undersized the encoder would have to be sending to many counts and that's hard to do unless there is outside noise involved.

one way you can tell if its mechanical or encoders is to pull the motor and mark the shaft then set the axis to 0 with the mark aligned to something then run some code that moves THAT axis a various lengths then back to zero over and over. When you return to zero at the end, it should always return so the shaft mark aligns perfectly, I don't think that controller uses g-codes so you will have to do it with a drawing. We use a "torture" file to test axis accuracy on testing systems but its g-code .

The important part is marking the shaft and an alignment mark because the encoder is all about where the shaft is .

It the motor never looses any position then its NOT the encoder. Has to be PAST the motor shaft somewhere. If it DOES lose it then suspect the encoder . Never rule out the cable and connection to the encoder . Running a "dry "cut with a marker will let you rule out noise from plasma It it only does it witht he torch on then its noise

[1MORECUT]

Do you have the Instruction Manual? If not P.M.

[adbuch]

https://www.plasmacam.com/archive/man

98-Z encoder.jpg

Plasmacam parts from suppliers.pdf

[url][https://drive.google.com/file/d/1wIAhqe ... haring/url]

[1MORECUT]

I have the Instruction Manual on a pdf.

[adbuch]

I have the Instruction Manual on a pdf.

Here is the Plasmacam DHC Manual.

David

Plasma Cam DHC Manual-1.pdf