Advanced Search

I'll have to test this out and see if manually zeroing #5210 changes applied G52/92 offsets.

The problem with things like this is that it may work fine for 99% of the configs but cause inexplicable behavior for the other 1%. The simple fact that this gets the python status channel out of sync would be enough for me to reject this solution. It seems strange that '#5210 = 0' was chosen rather than 'G92.2' particularly since these are used in MDI commands and ngc routines.

Looks like I asked this question a couple years ago.  Memory failing.

forum.linuxcnc.org/qtpyvcp/46414-probe-b...eter-question#247625

If #5210 is intended (in LCNC) to be a 'flag' which indicates a G52/92 offset is applied, then writing to #5210 directly seems to be the wrong way to clear offsets.  And may, in fact, not clear out any offsets.

If the PB toolchange intention is to disable G52 or G92 offsets, G92.2 (or 92.1) could be called in the subroutine and a check for #5210 = 0 could be made to ensure the offsets are cleared.

Or a comparison made: if #5210 = 1, then G92.2 (or something like that).

I'll have to test this out and see if manually zeroing #5210 changes applied G52/92 offsets.

Ok, this is a wierd one.   I have a Hurco KM3P mill with automatic speed control.  I'm moving to linuxcnc from factory controls.
The spindle is driven by a standard 3 phase motor and a servo that runs the reeves speed reduction in addition to a standard bridgeport type backgear selection.  I want to abandon the backgear, and just run on the slow speed geartrain, and use the reeves drive for the first part of the speed range, then once it's maxed out, use the VFD drive to get the remainder of the spindle speed up to 4000rpm.  So i'm thinking there must be a way to do this.  Probably write a module?   Anyone done anything like this before?

Trying to avoid screwing with the manual backgear selection depending on spindle desired speed.  It's literally the only manual selection on the whole mill.

Thanks for the answer.
but the issue I see and that makes me suspicious is that I download the image on a PC with W10 and I can't open it. It says it's corrupted, I can't even open it.

bertoldo wrote:

Good morning!

I am using the image from the link that started the forum about ethercat. My raspberry has 8 gigabytes and a 64 gigabyte SD card... and I use the raspberry image to generate the boot. I downloaded this file from the link you sent me now and the error is the same. If I try to open the image, it says that the file is corrupted.
 

 

Hello
the micro SD is very fragile, nothing is enough and it is no longer usable,
the installation with Raspberry Image is successful but the system does not start, I burned three,
with the installation directly on NVMe M2 and adapter
in my case, I solved it, I do not know if the problem is the same
I reinstalled several times without problems
 

machinedude wrote:

it might be that the X axis is under powered compared to the tandem Y axis and that could be why it shows up on contour cutting? that would be pretty easy to do and not redesign things from the ground up in the process It might point me in a particular direction?

• What does the unwanted surface look like on contour cutting?
• Is it a repeating pattern, or random marks/gouges?
• Does it happen at all speeds, or just some?
• When you specify 'contour', do you mean any cutting involving the X axis, or only certain motions?
• Belt profile and pitch?
• Pinion pitch diameter?
• Still using the CL57T drives?
  • Have you modified any settings inside the drive?
  • LCNC output steps per mm or inch?

This should help to narrow down if the problem is primarily due to the coarse motion system you have, or a tuning/settings issue.

Unless the motors are losing steps, I don't think you need to change motor voltage just yet.

for higher power 600V, something like
Onsemi NFVA25012NP2T might make sense
 

Has anyone been able to use HAL to get outputs from the USB port on ClearPath servos?

Teknic uses this usb port to connect their MSP configuration software. The port has a bunch of useful information. This is basically for monitoring purposes, but configuration can also be done through the usb.

Here you can see some from ST: www.st.com/en/power-modules-and-ipm/stpo...lligent-modules.html. The v5 is the middle category in terms of servo motor power.
Look in the HP category.
To which rectifiers are you referring? Ready to use boards or simple diode module? I know there are several architectures for rectifiers.

First the drive must be tuned to the attached motor, with it's own software, follow the procedure as described in the drive manual.
No LinuxCNC at all, until the drive is set up properly, as it is impossible to tune or use.
And a friendly warning, do not change stuff at random and hope it works, usually end up in tears or frustration.
forum.linuxcnc.org/38-general-linuxcnc-q...oder?start=10#309742
forum.linuxcnc.org/10-advanced-configura...to-example-mesa-7i77

That is worth testing, stepper systems at low speeds do not benefit from higher voltage, but at higher speeds the difference is huge.
BTW, low speed for a Nema23 stepper is under 150-200RPM some up to 500, from there the torque drops like a rock, so any added voltage helps a lot.

i think one thing that i could try to see if things improve would be to swap out 36V power supplies for some 48V supplies i think i have 4 of them in another box that i can steal or borrow just to see if it improves and with some extra power?

it might be that the X axis is under powered compared to the tandem Y axis and that could be why it shows up on contour cutting? that would be pretty easy to do and not redesign things from the ground up in the process It might point me in a particular direction?

Most of the time, such issues are caused by dirty pins on the add on card as they get touched.
A good cleaning with alcohol should take care of that.

If the drives are not tuned properly to the motors they are using, no ammount of tuning in LinuxCNC will help.
Normal procedure for tuning analog drives is to tune the torque mode first with motor not attached to machine, then do velocity loop tunning in drive, and if drives have position mode also tune this.
Only after tuning drives to motors you can procees to tune in LinuxCNC, just do not forget to set the drive in velocity mode after testing position mode, if any.

Looks awesome, as usual !
The box seems a bit to heavy to be mounted on the gantry, but if he insisted...