Motors spin too fast


W Maxwell
 

Motors spin, but tracking way too fast. I would estimate perhaps as much as 16x-32x based visually on what I see from the stock Losmandy 492 motor movement .

Slewing stalls out at 1/2x max.

I have verified that I do have 400 step/.9 degree motors. Other settings in attachment.

Setup: MiniPCB 2 , TMC 2130, Nema 17 .9 motors, SHC, Losmandy GM8, direct to axis (no pulleys/belts)


W Maxwell
 

also...I have been choosing TMC2130 for the configuration generator. What is 2130 Quiet/VQuiet? Is that a different setting for the same hardware or a different driver altogether?

If it matters, I have .9amp motors.


Khalid Baheyeldin
 

On Fri, Nov 13, 2020 at 11:32 AM, W Maxwell wrote:
also...I have been choosing TMC2130 for the configuration generator. What is 2130 Quiet/VQuiet? Is that a different setting for the same hardware or a different driver altogether?
They are options for the same driver.

VQuiet does not exist anymore. Do not use it.

Start with the regular TMC2130 for now until everything else works, then, and only then, try out Quiet.


W Maxwell
 

and..if it matters: 12v 5 amp power supply.


W Maxwell
 

Config file


Mike Ahner
 

On Fri, Nov 13, 2020 at 10:18 AM, W Maxwell wrote:
Motors spin, but tracking way too fast. I would estimate perhaps as much as 16x-32x based visually on what I see from the stock Losmandy 492 motor movement .

Slewing stalls out at 1/2x max.

I have verified that I do have 400 step/.9 degree motors. Other settings in attachment.

Setup: MiniPCB 2 , TMC 2130, Nema 17 .9 motors, SHC, Losmandy GM8, direct to axis (no pulleys/belts)
This seems like the TMC2130s are not operating in SPI mode, so the stepper motor doesn't the same distance OnStep believes it has moved. Who did you buy the TMC2130s from and what version are they?


John Petterson
 

Take the guard off of the worm on the RA axis, then start it tracking.  Time how long it takes for one revolution of the motor, you should be able to see whatever adapter you are using to attach the motor to the worm.  Put a mark on it if you have to.  It should take 8 minutes to make one revolution.


John Petterson
 

Howard and Khalid,

This may be a problem in some versions.  I loaded the latest master I have downloaded tonight (4.21d) and installed on my GM8.  The RA worm ran one rev in 80 seconds, 6x too fast.  Went back to version 4.16j and the problem went away.

Was there an issue that you know about?

John


Howard Dutton
 

On Tue, Nov 17, 2020 at 11:05 PM, John Petterson wrote:

This may be a problem in some versions.  I loaded the latest master I have downloaded tonight (4.21d) and installed on my GM8.  The RA worm ran one rev in 80 seconds, 6x too fast.  Went back to version 4.16j and the problem went away.

Was there an issue that you know about?

Version 4.21d was broken.  Apparently fixed more or less in 4.21f since about two days ago.  Fixed completely AFAIK in 4.21g as of yesterday.

The Config.h file below is not from version 4.x, so I assume he's using release-3.16 and this is likely not relevant here.


Howard Dutton
 
Edited

On Fri, Nov 13, 2020 at 10:27 AM, Mike Ahner wrote:
This seems like the TMC2130s are not operating in SPI mode, so the stepper motor doesn't the same distance OnStep believes it has moved. Who did you buy the TMC2130s from and what version are they?
Probably you are correct Mike.  I didn't comment until now since you pointed out the most likely cause already.  Just in-case someone is thinking this is a long-shot I want to elaborate:

Non-SPI TMC drivers seem to end up in 4x micro-step mode (when you attempt to use them in SPI mode) from the several other users we've seen with this issue.  He's asking for 64x (Config.h) and getting 4x so it'd be 16x too fast and he estimates it's running in the 16 to 32x too fast range.


W Maxwell
 

On Wed, Nov 18, 2020 at 05:56 AM, Howard Dutton wrote:
Non-SPI TMC drivers seem to end up in 4x micro-step mode (when you attempt to use them in SPI mode
Is there a way to get something more than 4x microsteps bypassing spi mode?

I realize the best thing to do would be to swap out the drivers, but for a variety of reason that isn't going to happen.


Howard Dutton
 

On Thu, Nov 26, 2020 at 07:44 AM, W Maxwell wrote:
Is there a way to get something more than 4x microsteps bypassing spi mode?
In that mode they should behave like TMC2100's and the best microstep mode is probably 16x spreadCycle.  For that CFG2 needs to be OPEN and CFG1 need to be grounded.
The pin positions on the TMC2130 are the same as TMC2100 and the labels can be seen here.

Note that OnStep can't set the open state on a pin so the only way is to electrically disconnect it.  This is my best guess but don't recall anyone doing it.

If you're up for the challenge there are usually solder jumpers on the drivers that can be set to change to SPI mode but it's difficult and some succeed and some fail.


W Maxwell
 

Since it appears I have the wrong drivers and it now appears to me the drivers can be pulled easily and hopefully also be easily be swapped out, it might just be easier to swap them. In current operation the pulses are so broad I can see it jerking in the eyepiece.

That being the case..

Could someone please tell me exactly what drivers I should get?  I have  TMC2130s  in it now that apparently can not operate in SPI mode. I don't really care about performance, more something that will be a direct swap out and will bullet proof work.


W Maxwell
 

On Thu, Nov 26, 2020 at 08:02 AM, Howard Dutton wrote:
For that CFG2 needs to be OPEN and CFG1 need to be grounded.
In case I choose to mod the drivers I have...  This would mean cutting the CFG2 pin and connecting CFG1 to the grd pin?

This being the case, would it then run in up to 16x microsteps adjustable via the configh file?


Howard Dutton
 

On Thu, Nov 26, 2020 at 08:33 AM, W Maxwell wrote:
On Thu, Nov 26, 2020 at 08:02 AM, Howard Dutton wrote:
For that CFG2 needs to be OPEN and CFG1 need to be grounded.
In case I choose to mod the drivers I have...  This would mean cutting the CFG2 pin and connecting CFG1 to the grd pin?
As far as I know.  As I said, I don't recall ever seeing it done so if this destroys the drivers, that's the way it goes.

In theory you would set OnStep for TMC2100 drivers in 16x mode (no micro-step mode switching) and cut off CFG2.

This being the case, would it then run in up to 16x microsteps adjustable via the configh file?
No.  Fixed in 16x mode, not adjustable.


Khalid Baheyeldin
 

On Thu, Nov 26, 2020 at 11:26 AM, W Maxwell wrote:
Since it appears I have the wrong drivers and it now appears to me the drivers can be pulled easily and hopefully also be easily be swapped out, it might just be easier to swap them. In current operation the pulses are so broad I can see it jerking in the eyepiece.

That being the case..

Could someone please tell me exactly what drivers I should get?  I have  TMC2130s  in it now that apparently can not operate in SPI mode. I don't really care about performance, more something that will be a direct swap out and will bullet proof work.
Before I answer this, did you check the video that I sent you a week or so ago to verify
if the drivers are indeed not in SPI mode? It would be best to verify this with certainty
before concluding that it is non-SPI.

As for what drivers to use:
Remind us again what board are you using, and what version (if applicable).
Also, what are the motors that you are using (how many amps they are rated for).

Here is why I am asking the above: the board can impose certain restrictions on certain
type of drivers.
And the type of driver can also have a maximum current.
Don't worry too much about the minutia.


W Maxwell
 

Thanks for the reply. Maybe I'll give it s shot since plan B is to get different drivers anyway. Plan C is to put it in a box in the closet since my tribulations with the project have been rather well documented. Its been there before..

As a note to your theory that it is in non spi mode and operating at 4x microsteps-  I tried two different configs, one at the original 64 microsteps and a number 1/16th the GR2, the second 4 microsteps with GR2 normal. Same outcome.

In other words by two routes I ended up with the same thing. It appears you were correct in that it was defaulting to 4x microsteps and running 16x too fast set at the 64 microstep config.


W Maxwell
 

On Thu, Nov 26, 2020 at 08:47 AM, Khalid Baheyeldin wrote:
Before I answer this, did you check the video that I sent you a week or so ago to verify
if the drivers are indeed not in SPI mode? It would be best to verify this with certainty
before concluding that it is non-SPI.

Yes. I also ran experiments based on Howard's premise and the results also seem to confirm his premise that its the non SPI drivers.
As for what drivers to use:
Remind us again what board are you using, and what version (if applicable).
Also, what are the motors that you are using (how many amps they are rated for).

MiniPcb v2 
Nema 17 400/.9 ,  0.9 amp motors


Khalid Baheyeldin
 

On Thu, Nov 26, 2020 at 12:01 PM, W Maxwell wrote:
MiniPcb v2 
Nema 17 400/.9 ,  0.9 amp motors
Great.
That board does not have any restrictions, and supports all recommended drivers.
Also, the motor's maximum power makes it usable from any driver.

Next: are you in the USA? If so, do you prefer ordering from Amazon, or eBay/Aliexpress
with a longer wait?


W Maxwell
 

On Thu, Nov 26, 2020 at 08:47 AM, Khalid Baheyeldin wrote:
Before I answer this, did you check the video that I sent you a week or so ago to verify
if the drivers are indeed not in SPI mode?
In reviewing your previous responses ,  you have mentioned these drivers:

https://www.amazon.com/BIQU-Printer-Stepstick-TMC2130-Stepper/dp/B07RSX2W3V/

I don't mean to be cheap, but with unemployment running out and the pandemic raging here in the US I'm not sure I can justify another $70 on an experiment. Are there other options that are lower performance, plug and play hardware-wise that are a bit less expensive?