Print

[Cedric02]

To compare with the A4988, the blanking time is 1µs minimum and the dead time is 30µs so 3.33% of power supply.

[Cedric02]

Hello Doramason.

For your test with the nema14 4V 0.8A 5ohm, the minimum current with the drv8825 is: 1.44V/5ohm= 0.288A.
You have tested with:
 - In 1/16, the second microstep is 10%, so 0.8Ax0.1=0.08A or 0.08<0.288 it's not good.
 - In 1/8, the second microstep is 20%, so 0.8Ax0.2=0.16A or 0.16<0.288 it's not good.
 - In 1/4, the second microstep is 38%, so 0.8x0.38=0.304A or 0.304>0.288  it' good.
Your test confirm what i saw in the datasheet.

For 9V power supply:
 - The minimum voltage with drv8825: 9V*0.12=1.08V
 - The minimum current with your motor (nema 14 4V 0.8A 5ohm): 1.08/5=0.216A
So for :
 - In 1/32, the second microstep is 5%, so 0.8x0.05=0.04A or 0.04<0.216  it's not good.
 - In 1/16, the second microstep is 10%, so 0.8Ax0.1=0.08A or 0.08<0.216  it's not good.
 - In 1/8, the second microstep is 20%, so 0.8Ax0.2=0.16A or 0.16<0.216  it's not good.
 - In 1/4, the second microstep is 38%, so 0.8x0.38=0.304A or 0.304>0.216  it's good.
So no change with 9V power supply.

For the potentiometer it' nothing just be careful when you set the vref...
For the hot nema17, set the vref to something smaller until the motor temperature is good.

[doramason]

Thanks Cedric! I wasn't able to find a 9v supply anyway Most electronic stuff I have at home works with either 5v or 12v.

I will also clean and change the grease in the eq5 maybe change the bearings as well. It might solve my problem in guiding if not I will buy a high voltage stepper, it seems like the only solution, right?

[Cedric02]

Hello Doramason,

Change the motor is the best solution if you want keep the 1/32 microstepping. if I had to change my motors Iwill choose the: http://www.omc-stepperonline.com/nema-14-bipolar-stepper-75v-05a-23ncm326ozin-14hs170504s-p-94.html , but the inductance is very high (26mh)... I'm not electronicien, and i don't know if it is really critical. i think that limit the maximum speed. if an electronicien could confirm this...
For the potentiometer, it jump 0.38V to 2.5V on clockwise or counterclockwise? If it's on clockwise turn it's not good... if it's on counterclockwise it's just a "cleat" is brocken and it's nothing.

[TCWORLD]

@doramanson

If the potentiometer is broken, it will be difficult to repair. It shouldn't matter if you can get the vref to be about right. Otherwise if the reference voltage is too high the driver won't be liming the current and in the long term it may damage either the motor, the driver, or the reverse polarity diode on the input (depending on how much current the motor draws).

In terms of heat, @1.1A, the driver chip will get hot, this is quite normal - in fact they can get hot enough to burn your fingers without damaging themselves. There is a built in thermal protection circuit in the DRV8825 which will disable itself until it's cooled down if it gets too hot.

[the_Geth]

Hi everybody 
My apologies for delay of my reply. Haven't checked the forum in a while.

Attached is a specification sheet of the motor I used for my Arduino Astro EQ.
Their rated voltage is 2.8v. Way below the supplied voltage which is 12v.

I immediately noticed the 1/32 microstep problem when I first set the mount to sidereal tracking using EQMOD and heard the motors doing the microstepping followed by a sudden jump. I later set a really slow custom tracking rate in EQMOD and by touching the RA motor pulley I counted 16 microsteps and a sudden big step.  Days later I took the mount out for testing and I knew it was something really bad when I went to try to focus the telescope and saw through the live view of my camera that the star (Vega) was oscillating up and down. Then I thought that the autoguiding could fix it...wrong. The glitch was so bad that calibration went out of wack and so the guiding . The motors were ignoring most of the pulse commands because the drivers were forcing them to do the described glitch. 16 microsteps then a sudden big step.  The problem is also present in the 1/16 and 1/8 configurations but goes away using 1/4 microstepping so I suspect that the sudden big jump is a 1/4 microstep.
I tried lower voltage power supplies but not luck. Googling here and there for what the problem could be with the DRV8825 I stumbled upon the page Cedric posted and later with the diodes solution. Lacking any of the mentioned diodes, I opted for the fast decay modification. It was really difficult to solder such a small thing! I used resistor leads instead of a wire which are thinner and more comfortable to work with.
The modification was a success. Now the motors run smoother but they make more noise. Not a big deal since my CG-5 was able to pull a 10 mnts exposure with round stars having about 22 pounds of load on it
Haven't tried the diodes trick yet but now I feel tempted.

Cheers

Alex

[TCWORLD]

Nice job with the soldering Alex. I'm glad you got yours going. At least with Astro it is only light you are trying to capture not sound!

I'm tempted to ask Pololu if they would consider making a new revision of the board with the decay pin broken out - even if just to a pad (ideally an on-off-on surface mount switch) in light of this issue. Not sure that they would do it though.

[doramason]

Hi everybody 
My apologies for delay of my reply. Haven't checked the forum in a while.

Attached is a specification sheet of the motor I used for my Arduino Astro EQ.
Their rated voltage is 2.8v. Way below the supplied voltage which is 12v.

I immediately noticed the 1/32 microstep problem when I first set the mount to sidereal tracking using EQMOD and heard the motors doing the microstepping followed by a sudden jump. I later set a really slow custom tracking rate in EQMOD and by touching the RA motor pulley I counted 16 microsteps and a sudden big step.  Days later I took the mount out for testing and I knew it was something really bad when I went to try to focus the telescope and saw through the live view of my camera that the star (Vega) was oscillating up and down. Then I thought that the autoguiding could fix it...wrong. The glitch was so bad that calibration went out of wack and so the guiding . The motors were ignoring most of the pulse commands because the drivers were forcing them to do the described glitch. 16 microsteps then a sudden big step.  The problem is also present in the 1/16 and 1/8 configurations but goes away using 1/4 microstepping so I suspect that the sudden big jump is a 1/4 microstep.
I tried lower voltage power supplies but not luck. Googling here and there for what the problem could be with the DRV8825 I stumbled upon the page Cedric posted and later with the diodes solution. Lacking any of the mentioned diodes, I opted for the fast decay modification. It was really difficult to solder such a small thing! I used resistor leads instead of a wire which are thinner and more comfortable to work with.
The modification was a success. Now the motors run smoother but they make more noise. Not a big deal since my CG-5 was able to pull a 10 mnts exposure with round stars having about 22 pounds of load on it
Haven't tried the diodes trick yet but now I feel tempted.

Cheers

Alex

Wow 10 minutes! My setup is also around 10kgs and I would be more than happy to achieve that. Unfortunately I cannot do soldering that well so this is not an option for me 

Btw I have disassembled the mount following the guides online but most of them are outdated for a new skywatcher eq5 mount. Both ra and dec worms have radial ball bearings and the grease isn't as bad as they've described (it is quite sticky though). After long search I couldn't find the suggested grease in my country(as always) so I ordered some on ebay. So my mount is sitting stripped waiting to be greased. I hope I can assemble it back without any problems, and not make it worse than before. I will post if it improves tracking and reduce periodic error. 

[the_Geth]

Nice job with the soldering Alex. I'm glad you got yours going.

Thanks Tom. Was not easy but doing the second one I found that its easier to solder the wire to the M2 pin first (this will serve you as some kind of support) then make a thin line of solder come out of pin 19 and solder the other end of the wire to it. The picture can give you an idea of what I mean. I strongly recommend using a soldering helping hand tool.

At least with Astro it is only light you are trying to capture not sound!

  True. The noise is not a big deal. Can't hear it unless I'm close to the mount.

I'm tempted to ask Pololu if they would consider making a new revision of the board with the decay pin broken out - even if just to a pad (ideally an on-off-on surface mount switch) in light of this issue. Not sure that they would do it though.

I read of people trying to reach Pololu with the same suggestion but haven't seen any change yet. Hopefully they are considering it and the next version will have an option to facilitate this.

Wow 10 minutes! My setup is also around 10kgs and I would be more than happy to achieve that. Unfortunately I cannot do soldering that well so this is not an option for me 

You could try with the diodes or the TO-220 as Tom suggested. I also recommend using the exact same motor model for both RA and DEC.

Btw I have disassembled the mount following the guides online but most of them are outdated for a new skywatcher eq5 mount. Both ra and dec axis have radial ball bearings and the grease isn't as bad as they've described (it is quite sticky though). After long search I couldn't find the suggested grease in my country(as always) so I ordered some on ebay. So my mount is sitting stripped waiting to be greased. I hope I can assemble it back without any problems, and not make it worse than before. I will post if it improves tracking and reduce periodic error.   

I did the same to my mount about a year ago (disassemble, clean, regrease, assemble and adjust the worm gear mesh). I did see some improvement. Hope yours will run well too

Alex

[Jagho]

Hi,
I have the same problem with 3,6V 1,4A motor.

Change the motor is the best solution if you want keep the 1/32 microstepping.

I HAVE to keep 1/32 microstepping, so what do you think about the solution with the diodes ? You said "diodes are very hot". Why not using TO220 and radiator. Isn't it the best and the simplest solution ? Unless there were something more than hot diodes. Cedric02, you were testing it. What do you think about it.

What diodes you used ?

[Jagho]

I've got the question for Tom. I've found DRV8834 Low-Voltage Stepper Motor Driver Carrier:
https://www.pololu.com/product/2134
Is it possible the simple replacement with DRV8825 ?
... and could it solve the problem low voltage motors (1/32 microstepping) like mine (3,6V 1,4A),
because I can use lower supply voltage.

[TCWORLD]

@Jagho, that looks very interesting. A quick browse through the datasheet shows that it should have a much better performance with low-voltage, high-current motors.

It's not going to be a straight drop in replacement - in terms of hardware it would be fine, plug it in, use a supply of between 7 and 9V. But I will need to make some modifications to the firmware to interface with it correctly.

I will have a look into it, but for the next couple of weeks I am quite snowed under at work, so it will have to wait until after that.

[Jagho]

Thanks Tom, I'll be waiting.

[doramason]

I did the same to my mount about a year ago (disassemble, clean, regrease, assemble and adjust the worm gear mesh). I did see some improvement. Hope yours will run well too

Alex

I have cleaned the old grease (which wasn't that different than white lithium grease that I was going to use) and polished some parts of the dec axis. It didn't improve anything, on the contrary the balancing has become harder. I think it is because of the plastic washers on the both ends of the worm gear, they are preventing the worm gear to turn freely. The one on the bottom of the worm gear especially looked too large for its place and didn't fit easily. Anyway that doesn't affect slewing so I didn't bother with it that much. So my experience if you buy a recently manufactured eq-5 mount there is no need to disassemble and change the grease. It looks like a waste of time. 

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. There is no visible backlash to me when I turn the slow motion knob so I am not sure what is causing it. In guiding I still have similar problems of overcorrection on both sides. At this moment I am not sure what is the cause; is it the backlash or the missing microsteps or both. I will try to guide only one direction next time to see if that helps. 

I was wondering what would happen if I use a 5v power supply for the steppers instead of 12v? Would this improve the microstepping? or the diode solution is my only option? If the 5v power supply would improve I can buy one with the fitting male connector.

[Jagho]

I was wondering what would happen if I use a 5v power supply for the steppers instead of 12v? Would this improve the microstepping? or the diode solution is my only option? If the 5v power supply would improve I can buy one with the fitting male connector.

The mimimal power supply for DRV8825 stepper motor controller is 8.2V, so it will not work for 5V, I suppose.
DRV8834 could be the solution - power supply 2.5 - 10.8V - but we have to wait for Tom's firmware update.