Print

[TCWORLD]

Even ignoring the DRV8825, the minimum voltage the AstroEQ requires is 7V (there is a regulator inside which requires that much to ensure 5V to the microcontroller). With the lower voltage driver a 7.5V supply would work (once there is firmware support).

In terms of guiding, you may want to check the settings in EQMOD - on the EQMOD window if you open the advanced settings (where all of the alignment and guiding stuff is), there is a section for "ASCOM PulseGuide Settings". These settings will need to be adjusted to suit your mount/motors. If the mount doesn't move south when doing calibration, you should try increasing the "DEC Backlash" setting a bit and also the "DEC Rate" setting - these can be tuned to get EQMOD to drive the mount through any backlash in the gears so that the calibration can be successful. If you find that the guiding is oscillating in RA - lots of pulses in one direction, then lots in the other, you can also increase the RA Rate setting to make the corrections faster (means each correction will be a slightly larger amount which may help).

[the_Geth]

Hi,

For research sake and the will to help others with the jumping/missing steps problem, I decided to try the diode trick.
The good news is that it really worked!
As you can see in the pictures the circuit I made for the diodes was to simply solder their leads together in a loop, leaving
an overlap of about 7mm on each side for connection. Sorta looks like rabbit ears when connected to each phase of the motor.
Something much more elegant can be made using a project circuit board or somehow integrate the diodes in the AEQ circuit.
(Please ignore the resistors on the left side of the breadboard. They are from another project)
Although I used an Aduino Uno, a different motor and a different code for this test, the missing steps were still there at 1/32 microstepping.
After connecting the diodes the problem dissapeared but the sound coming from the stepper motor was similar. It ran smoothly but I didn't notice any difference
from the fast decay PIN 19 to PIN M2 soldering modification.
I was using 1N5408 diodes which have maximun continuous forward current rating of 3A. After having the motor running for about 40 minutes they never got hot not
even warm. The Supplied power to the DRV8825 was 12v - 2A and the Vref for the 2.6v and 1.2A rated Nema 17 stepper motor was 0.59v.

Although I have adjusted the worm and worm gear to be quite tight I have problems with phd calibration due to backlash. Star goes north but doesn't come back to its place.

I wouldn't worry so much about the mount not going back south during the PHD calibration. Mine also does that and the guiding is excellent anyways. This glitch
also used to happen when my CG5 was stock. Maybe is a software problem.

In guiding I still have similar problems of overcorrection on both sides.

Like Tom mentioned, guiding problems like overcorrection can be atenuated in EQMODs "ASCOM PulseGuide Settings" as well as "ASCOM PulseGuide Monitor".
Further information on how to use these settings can be found at the EQMODs website. 

If you still haven't set the DRV8825s into fast decay mode or done the diode trick, I'm afraid that your guiding problems are related to the mix decay problems these drivers have.

Hope all these helps

Alex
 

[TCWORLD]

Given the success that you've had with the diodes, I've designed a quick PCB (see pic) which would go inline with both motor cables to add the diodes.

I'm going to send the board off for manufacture and will test out how well it works while I'm back over Christmas/New Year period.

[the_Geth]

That's awesome news Tom!

Looking forward to knowing how things go with it.

Glad to bring some help to the forum. 

Alex

[doramason]

I have found someone with better soldering skills and now the drivers are in fast decay mode. It has improved things but still not running perfectly smooth. It seems like it is going faster and then it slows down and then runs faster again. So the pulsation still exists but much less, at least it is not jumping like before. If the weather permits I will try guiding this week.

Meanwhile I was thinking of changing the motors with higher voltage 0.9 degree ones. Like this one http://www.omc-stepperonline.com/download/pdf/17HM15-0806S.pdf
It would also give me better resolution or maybe I can use 1/16th step mode so things will be smoother. Or am I wrong and this will cause me more trouble because a 0.9 degree stepper is more prone to missing steps? And what do you think about the specs of this motor?

[TCWORLD]

A 0.9* motor with 1/16 stepping will likely be better than 1.8* with 1/32 stepping because more of the resolution comes from the mechanical side of things and less from microstepping. Those motors seem fine, though you'd have to verify how to wire them up - it looks like a Unipolar motor (in which case either skip the centre taps, or one end tap on each coil), though the datasheet says it is 4 phase.

[doramason]

A 0.9* motor with 1/16 stepping will likely be better than 1.8* with 1/32 stepping because more of the resolution comes from the mechanical side of things and less from microstepping. Those motors seem fine, though you'd have to verify how to wire them up - it looks like a Unipolar motor (in which case either skip the centre taps, or one end tap on each coil), though the datasheet says it is 4 phase.

I have the same wiring with my current nema17. I've left center leads unconnected and it works bipolar.

[doramason]

Hi,

For research sake and the will to help others with the jumping/missing steps problem, I decided to try the diode trick.
The good news is that it really worked!
As you can see in the pictures the circuit I made for the diodes was to simply solder their leads together in a loop, leaving
an overlap of about 7mm on each side for connection. Sorta looks like rabbit ears when connected to each phase of the motor.
Something much more elegant can be made using a project circuit board or somehow integrate the diodes in the AEQ circuit.
(Please ignore the resistors on the left side of the breadboard. They are from another project)
Although I used an Aduino Uno, a different motor and a different code for this test, the missing steps were still there at 1/32 microstepping.
After connecting the diodes the problem dissapeared but the sound coming from the stepper motor was similar. It ran smoothly but I didn't notice any difference
from the fast decay PIN 19 to PIN M2 soldering modification.
I was using 1N5408 diodes which have maximun continuous forward current rating of 3A. After having the motor running for about 40 minutes they never got hot not
even warm. The Supplied power to the DRV8825 was 12v - 2A and the Vref for the 2.6v and 1.2A rated Nema 17 stepper motor was 0.59v.

Although I have adjusted the worm and worm gear to be quite tight I have problems with phd calibration due to backlash. Star goes north but doesn't come back to its place.

I wouldn't worry so much about the mount not going back south during the PHD calibration. Mine also does that and the guiding is excellent anyways. This glitch
also used to happen when my CG5 was stock. Maybe is a software problem.

In guiding I still have similar problems of overcorrection on both sides.

Like Tom mentioned, guiding problems like overcorrection can be atenuated in EQMODs "ASCOM PulseGuide Settings" as well as "ASCOM PulseGuide Monitor".
Further information on how to use these settings can be found at the EQMODs website. 

If you still haven't set the DRV8825s into fast decay mode or done the diode trick, I'm afraid that your guiding problems are related to the mix decay problems these drivers have.

Hope all these helps

Alex

Hey Alex, which solution do you think is better diodes or fast decay? Have you had the chance to test your guiding?

[the_Geth]

Hey Alex, which solution do you think is better diodes or fast decay? Have you had the chance to test your guiding?

Hi,
I haven't checked the guiding using the diode configuration but to tell the truth, I didn't noticed any substantial difference from the fast decay mode during the test on the bench. I'm pretty sure both will work fine but if you or a friend can do the soldering to accomplish fast decay on the driver, then stay with that. You will get something more compact.

I would also recommend to use pulleys and belts and get rid of the gears. You will always have tracking/guiding problems if you keep using gears.   

Regards

Alex

[doramason]

Hey Alex, which solution do you think is better diodes or fast decay? Have you had the chance to test your guiding?

Hi,
I haven't checked the guiding using the diode configuration but to tell the truth, I didn't noticed any substantial difference from the fast decay mode during the test on the bench. I'm pretty sure both will work fine but if you or a friend can do the soldering to accomplish fast decay on the driver, then stay with that. You will get something more compact.

I would also recommend to use pulleys and belts and get rid of the gears. You will always have tracking/guiding problems if you keep using gears.   

Regards

Alex

I have pulleys and belts the only gear is the worm and I can't get rid of that . diodes are a better option I think because if the driver breaks down it would be easier the replace it with a new one without fast decay mod.

[igorb]

Given the success that you've had with the diodes, I've designed a quick PCB (see pic) which would go inline with both motor cables to add the diodes.

I'm going to send the board off for manufacture and will test out how well it works while I'm back over Christmas/New Year period.

Hi!
Is there copper Layer PDF Printout for download anywhere? My motors are 1.3A 2.5ohm/phase so I think I'll need to add diodes to prevent skipping microsteps.

[watchgear]

Hi Everybody,

I also had this problem with DRV8825 drivers and NEMA17 steppers. First I tried the diode trick and it worked. But the diodes got hot and even if I was not afraid they could be damaged, I did not want to drain my batteries, heating something I do not need...

Then I tried to do the trick with that bridge to M2 - but I decided to brigde the pin to constant 5V. Even if the problem should only occur in 1/32 step mode - there where some people that wrote, the problem also shows in other modes - only not that bad.

When I tried to solder a piece of wire to that pin, I made a bridge to the neighbour pin - by accident. Than I looked to the DRV8825 schema and saw, that this pin is on constant 5V - it is the FAULT pin, which is not needed with the DRV8825, but is on 5V in the AstroEQ design for compatibility to the A.... driver.



So - if someone wants to use the fast decay mode and is afraid of the soldering (like I was) - it is quite easy just to bridge these two pins. And it works...

Best regards
Klaus

[TCWORLD]

@watchgear

Interesting way of doing it. That should indeed work so long as the motor driver never overheats and goes into a fault condition. Though even if a fault condition happens, all it would do is put the driver into a different decay mode until the fault is cleared, which is probably not actually an issue.

Nice job!

[Washy]

Hehehehe I bridged to the pin the other side mistakenly - it was direction so I only ran in one.  Have fixed it now and my motors apart from noisy now do run smoothly at 32nd stepping, I guess I'll have to have another go at current limiting but I just think part of fast decay is noise.

http://s15.postimg.org/l30tlwdbf/20160324_122026.jpg

I saw some various sites talking about diodes, but I think that's only going to fix an issue if the steppers can't take the full VMOT voltage which mine can because they are 24v rated?  Is my understanding correct?

[doramason]

I have got some IN5408 diodes to test my new motors if they perform better with them rather than fast decay mod. The motors I have bought are unipolar/bipolar(6 leads) 0.9 degree 6V 0.8A nema17s.

When I tried them with diodes they have performed even worse than having nothing(no fast decay mod or diodes). And with fast decay mod there is a slight improvement than original.

Everyone else says there is huge difference with diodes so I was a bit shocked that they have bad performance.

Then I have tried the diodes with my worst motor in terms of jumping steps; bipolar(4 leads) 1.8 degree 2.7V 1A nema14. And it worked flawlessly! Even better than fast decay mod.

I'm an industrial engineer so electronics are a foreign language to me, I cannot come up with a reason why this is happening. My best guess is the unipolar/bipolar difference.

I would be very happy if someone can describe me the possible causes of this so that maybe I can find a way to use the diodes with my new motors. Because even with fast decay mod I can sense the motors are moving irregularly, though not so bad.