HBS57 and oval circle issues.

Do the steppers have the same torque?
Have you tried lowering the accelerations?
And what is the error you allow LinuxCNC/ do you use g64?

I'm fairly positive that the hardware are not the problem since I can swap out the hbs57 closed loop driver with my DQ542ma open loop driver - without changing anything else, the machine functions perfect. So I think the problem has to be between the driver and Linux cnc.
I don't know much about the software side, could it be possible that there is a step pulse timing issue between Linux cnc and the driver? 

​​I haven’t tried lowering accelerations - will try this afternoon

When I swap out the drivers (hbs57 to dq542ma) I leave the same motor attached - I just leave encoder unattached  with the dq542ma driver.
I haven't used g64 codes. 

To me to looks like one axis is slower in accelerating or braking.
And this results in the strange corners and the oval.
Have the drivers the same on both axis the same settings?
And is the closed loop tuned? Are the position Singhalesen looped Back to LinuxCNC?

Loos like something might have gotten lost in translation.

"position Singhalesen"

LOL

= position signals

Ein hoch auf Wurstfinger...

Yea should be something like position or encoder signals.

This behaviour could be explained by poor tuning in the HBS57 drives

OP-

Those drives - as well as similar drives - usually ship with a 'following error alarm' of something like 4000 counts, as well as pretty loose PID tuning.

Given the 1000ppr encoder, this means the motor can be a full turn off during a move. Turning off the closed-loop feature does not change this, it just means there's no attempt to correct any error.

Go watch some Youtube videos of folks testing out thier HBS57 (or derivatives) drives.  You'll see them putting a wrench or something on the shaft and watching them 'servo' by moving the wrench.  Witness how cool it is when the wrench swings a bit and springs back: Joy.

But... a grown-up servo (or closed loop stepper) will resist that torque and not move until the current exceeds the settings and then it will fault out and go slack.  Obviously this is dependent on the following error settings, but following error tolerance tends to be set just enough to stop dithering at rest.

The problem is that the tuning software needed to adjust the stepper following error tolerance is pretty miserable.  Depending on the drive mfgr, you'll need a USB to RS232 converter to connect to the drive.  Other than genuine Leadshine (and re-badges, like Automation Technologies and the high-end StepperOnline), the software has zero instructions.  And it usually defaults to Chinese characters, which is fun.

Before chasing anything else in LinuxCNC, mechanical issues, or similar... buy/beg/steal a tuning cable and find the tuning software for the drives.

Then adjust the following error (or whatever oddball term is used in the software), and re-test.  Below is screenshot of the StepperOnline software settings - note the "4000" counts as default and the term there is 'Position deviation alarm', but you get the idea.

If you can't find the right software right away, keep poking around until you find one that will connect to the drive.

Good Luck
-R

 

Hi, and have a good day.

I have this cable, and HBS57 driver (from penny-G at aliexpress) seems is somehow responding on it.
While not connected motor coils it stop blinking red-green leds at driver when RJ45/USB convertor is connected to PC (ch340 seems inside connector, COM port is visible under putty - with no response).

But i can't find software for this HBS57 clone anywhere. Leadshine's one is not working - not connecting at all.

Can anybody help with this software?

have tried some software too.... but without luck