Strange behavior

[tafcnc]

Hello everyone,

Changed my system from G100 to smoothstepper. Old system was working ok except for known issues with the G100 networking.
Installation of the smoothstepper has been "smooth" and it looks very reliable.
I'm struggling with some strange behavior:
Axis move correctly in both directions when the movements are one axis at a time, both jogging and from gcode.
X, Y axis move ok simultaneously in all directions, except when they both run in the positive direction, in that case, motors make awful noise and
loose steps (lots of them). No matter what setting I use in the plugin config panel, behavior is consistent, everything moves fine, except
when x,y go in the positive direction. Any help will be really welcome.

Mach version is R3.041
Smooth stepper firmware Beta2 013c
Drives are geckos 203v
No breakout board as drives inputs are optoisolated, cable is shielded and nothing else is connected to the smooth stepper I/Os.
I've read somewhere that pulse width setting in motor tuning panel is not used by SS, but I have not found if there is any other
place it can be modified.
I have a two smoothsteppers and both do the same thing. I have swapped the Gecko drives, still same thing. Alternate grounding scheme,
same thing.
First I thougt about noise, but individual and simultaneos moves are ok, except the said combination.
Now I'm guessing about some math or overflow error somewhere...

Thank you

jose

[Chaoticone]

Jose, Did you set up a new xml from scratch for the SS or just redo your one for the G100? I would start by creating a brand spankin new xml. In plug-in control you have a few settings.

Brett

[tafcnc]

Brett,
Thanks for the reply. Yes, I've uninstalled mach, remove directories, install new version and create specific profile for the SmothStepper, in fact two
profiles, as I've tested with two different ss, (different serial number) in the same machine. I've tried the settings in the plugin config window, both
controller and max step frequency, the don't make any difference. Even I have tried with all the max and min values for each param, same thing.
I've have really run out of ideas about what could be going on. 

jose

[Hood]

Do you have more than one plugin in the plugins folder for the SS? Some people reported that if they had two or more they could have issues, now when I install a new plugin I rmake a habit of renaming the old one from DLL to M3P, just a precaution on my part as I have never seen this problem but why take risks LOL

Hood

[tafcnc]

Thanks Hood,

No, just one profile for ss, tested in clean installation of Mach.
Axes move always ok even with with different motor tuning profiles, the problem only shows, in an orderly manner, only when moving x,y in the positive direction simultaneously. Any other simultaneous movement works fine.

jose

[Chaoticone]

Jose, you using the default mach screen set? Checked none of your outputs are used twice?

Brett

[XLR84x4]

Hey there,

I have done a very similar swap over as you have. I have my SS connected directly to the G203 also and have installed the Grex plugin to work as I/O only on Port 3. I am running SS 13C plugin and just upgraded to the latest beta test Mach but was running 3.041 with no problems.

My only querks were random Estops however a few resistors/capacitors and the new plugin took care of the gremlins. I have steppers on the Y and A and a servo on the X axis and can run my little machine up to 10000mm/min as a three axis machine with X-Y gcodes as you suggested with no issues.

It was very easy to setup the new mach for the SS by creating a new profile etc I did have a few minor querks getting the Grex loaded with the new plugin but it was a matter of re-reading the manual.

All I can suggest is re-visiting your motor setup ensuring you have the step pulses correct. My vel is 20000 with Accell of 1250 on the X and Y and vel25000 with acell of 20000 on the A Axis stepper. No stepper losses with these high settings. My only quirk is the tangentail control not running CV hence the high A axis, just waiting on the new Mach release to hopefully fix this.

If you need any other setup info from my machine, I would be happy to assist as the guys on this forum have passed on the same help when I started out. Geez it made life easy asking all the dumb questions and getting good help

Jason

[tafcnc]

Many thanks,

@Brett:

Yes, default mach screen. The only outputs in use are the ones corresponding to x and y step and dir, all other set to zero.
In fact I don't know if this is the problem In this machine I only use x and y axes. But I have added another motor in the Z axis to test different combinations of movement. Guess what? Problem disappears. All the combined movements of the three axis
(including the one that caused the error) work ok.
Surprisingly when I remove the z motor and test again with x and y, works ok also, the only difference the cable connected to the driver of the z axis remains in place.

@Jason:

What nice rapids should you have with those speeds. I'm driving a small vertical mill (optimum BF20 L) that I use to make
the parts for the big one. I'm using the same tuning as with the G100 for the motors. In fact, axes moves ok if jogged individually or if x,y movements are not in the positive direction for both of them simultaneously.

I have replicated the setup in another machine, same motors, drives, firmware and mach versions. everything works ok.
So I would say that this is some noise related problem. Noise filtering is activated, but none of the inputs is used, so I'm
not sure this makes any difference.

One question I will like to know is which is the recommended grounding scheme for the ss.
It has a jumper to connect one of the mounting holes to ground (I have tried both possibilities, but problem persists).
I don't know if the USB connector connect the ground  from the PC to the ss, if that is the case and the jumper is set to
connect mounting hole to chasis earth, I will not be surprised if ground loops do occur.

The manual of the ss, mentions that the ss is intended to be used with breakout boards, but in this case only the motor outputs
of the ss are connected to the optoisolated inputs of the G203V, so I don't see the benefit.

Jason, I would really appreciate if you can provide some info about how you have grounded your smoothstepper and how you
used the resistors/capacitors to eliminate random trigger of Estops? Are you using any breakout board with the ss?

I am getting a warning "ESTOP not assigned to a SmoothStepper Input!" every time the firmware is reflashed (during startup of mach, and when changing configs) is that ok?

jose

[tafcnc]

Hello again,

Some of you may wonder, that if the only difference is the cable connected to the driver of the z axis
what happens if I remove it and revert to the initial configuration. Strange behavior shows again! Even if in Mach Z axis is
now disabled.
So, to recapitulate,
- x (port 1, pins 2,3) and y (port 1, pins 4, 5) connected alone -> motors loose steps and make awful noise if axes go in the positive direction simultaneously.
- Connect also z (port1, pins 6, 7), nice, clean, and smooth movement in all directions
All the checks mentioned before been taken into account.

I can't help but wondering what's going on
Problem seems fixed now, but It would be great to understand what is happening.

jose

[Hood]

When you say you have the filtering enabled, what have you got them set to? Also filtering shouldnt really be needed unless you are getting E-Stop problems but putting some in certainly wont hurt.
 I dont have any of my SS grounded (3 of them) and dont have any problems. The lathe has only two axis so its not a case you definitely need three axis connected.
 Sorry I am not being much help

Hood