[charliebrumfield]

We slapped a giant Clear Path Servo from Technik on the 4th axis of our PCNC770. All was well while under Mach3 control but with the path pilot upgrade I can't seem to make the deceleration work quite right with interpolated moves and it's causing our 4th axis work to come out twisted. See video for details. Anyone have any idea what parameters I need to tweak? I'm not familiar with LinuxCNC or PathPilot yet so this one has me stumped.

https://youtu.be/d_MVdAq-EOk

[PCW_MESA]

Try setting P to 1000 and MAX_ERROR to 0.1

[charliebrumfield]

Try setting P to 1000 and MAX_ERROR to 0.1

ok thanks just tried those numbers, no real change in cutting a test part. but the decell now causes an over shoot of about 5 degrees on a single rotation.

[PCW_MESA]

ok thanks just tried those numbers, no real change in cutting a test part. but the decell now causes an over shoot of about 5 degrees on a single rotation.

Strange, is this with a Tormach PC?

You might try setting MAX_ERROR to 0 ( this is a special case that means no input error bounding on the PID )

Also I noticed you acceleration settings are quite high (0 to full speed in about 16 ms) is this deliberate?

Can you post your HAL and INI files?

Also do you have any backlash compensation enabled?

[charliebrumfield]

Strange, is this with a Tormach PC?

Yes we are using a tormach controller on both out 770 and 1100. (haven't set the 4th axis up on the 1100 with the new path pilot yet)

You might try setting MAX_ERROR to 0 ( this is a special case that means no input error bounding on the PID )

Did that with P = 50 and the decell works flawlessly now. still getting a twisted part though. so now i'm even more stumped.

Also I noticed you acceleration settings are quite high (0 to full speed in about 16 ms) is this deliberate?

Yes, we make very short moves, and need to get the a axis up to speed in a short amount of time. (we spin the a axis for the entire machining run) they are will with in the capabilities of the motor. just to be on the safe side I set the accel rates to stock tormach setting for part of a test run. the die was still coming out twisted. did no let the run finish though as it was going to take over an hour to run completely.

Can you post your HAL and INI files?

yes, just as soon as I figure out where I put those wifi things tormach sent over

Also do you have any backlash compensation enabled?

not to my knowledge. I set the physical backlash adjustment in the 4th axis as per tormach's specs and have not enabled and back lash comp in path pilot.

[PCW_MESA]

Note that Tormachs linear axis accel values will be way too low for a rotary axis

To scale them similarly (same time to full speed) you would set the
rotary axis accel = rotary axis maxvel * linear axis accel / linear axis maxvel

To determine if the problem is still stepgen following related, I would set the ferror limits much smaller
say .01 degree for min_ferror and .1 for ferror to start, if you dont fault with a ferror then you know the
stepgen is within 0.1 degree of its commanded position

BTW I tried a sample config with your previous PID and accel/velocity setting and had no stepgen
tracking issues so I'm still not sure whats going on here

MAX_ACCEL=8000
MAXVEL = 125
P=1000
FF1=0
MAX_ERROR=.0005


https://ibin.co/2jID5Q7uICfc.png

Note that the error is always less than 0.0005 degrees so not really sure what going on here

[tmallard450]

I was looking into those clear path motors the other day and from what i gathered you can plug a usb into the clear path servo and set things like acceleration and deceleration. Their demo video showed the motors doing a self calibration routine too. So maybe you are having to different motor curves fighting against each other. These things look like great upgrade to the Tormach's


Sent from my iPad using Tapatalk

[charliebrumfield]

love the clear path servos. I have 4 other machines running them. they are the ONLY motors to run. they really aren't expensive and they make your control boxes super clean and tiny as the drivers and encoders are built in to the motor itself.

[VegasRhino]

Hi Charle, I have no input for your 4th axis issue, but I am now super excited to hear of somebody bolting some Clearpath Servos onto their Tormach! I have the 770 also and would be pleased to know a little more about what this mod has done for you, experiences, model of Clearpath chosen, ect. Any interest in submitting more info???

[ganderboy]

Hi Charle, I have no input for your 4th axis issue, but I am now super excited to hear of somebody bolting some Clearpath Servos onto their Tormach! I have the 770 also and would be pleased to know a little more about what this mod has done for you, experiences, model of Clearpath chosen, ect. Any interest in submitting more info???

I have a pcnc 1100
I've been having driver issues. 2 dead in the last month
I'm very interested in upgrading to the clear path servos.
Anyone with any more advice about this please speak up.
Thank you

[ralphieboy127]

love the clear path servos. I have 4 other machines running them. they are the ONLY motors to run. they really aren't expensive and they make your control boxes super clean and tiny as the drivers and encoders are built in to the motor itself.

Hi Charlie
I have a Chinese baileigh 2 x 2 plasma table that I Converted to warp9 ess smoothstepper
Using mach3... The machine runs great but if I cut thick materials at low speeds the steppers are NOT happy ...they moan and growl..I have changed the micro steps in the controllers for months with now with NO results will the clear path servos do the trick ??? I put what I need in the cart at their website 3 servos and power supply and all needed cables ...it comes to 1300 or so not a problem....only if it will work...what are you thoughts ???
Thanks Ralph