Print

[Cedric02]

Hello.
 Thank you for your project! I made an Astroeq 2560 with drv8825. My setup is 1.8 stepper angle, 1:3 motor gear ration, 1:102 worm gear ratio, 32 micro step. In goto mode it work fine. But in tracking mode, the motor make a short pause every second. What can do this?
 Sorry for my poor English... And thank you!

[TCWORLD]

It is probably still stepping, but you may not have enough Torque (or the current limit not properly set). With your configuration you can expect about 22 microsteps per second. With 1/32 microstepping this means that each half step will be a little under a second, so if the Torque is too low the motor may not be moving the microsteps and instead jumping to each half step as it comes.

[Cedric02]

Hello, thank you for the rapid response. My motor is an 4.1V, 1.1A. I have set the vRef to 0.5V, the problem is the same and the motor is very hot about 5min... My drv88225 have not a vRef "via hole", I set the vRef with the center pin of the potentiometer.

[Cedric02]

Hello, my problem is solved by switching the 8825 on fast decay mode. I applied the method on this site: http://reprap.harleystudio.co.za/tag/pololu-drv8825/ . Thank you for all!

[the_Geth]

Hi,
First, I'd like to thank Tom for this wonderful project.
 
I also had this problem when I completed my Arduino Astro EQ and solved it the same way. When tracking using 1/32 microstep mode, the motor rotates about 16 microsteps then it seems to stop or jump a full step. Looking here and there I found that it looks like the DRV8825 has a problem microstepping from 1/8 to 1/32 in its default mix decay mode. The mentioned fast decay modification seems to solve it but the motors get a little noisy (high pitch noise). 

Another possible solution (which I haven't tried yet) can be found here: http://cabristor.blogspot.com/2015/02/drv8825-missing-steps.html . The solution still uses the drivers mix decay mode but the 1/32 microstepping is smooth and the motors run more silent. In one of the paragraphs it cites " in mixed decay mode, the driver cannot do any voltage lower than 12% of the supply voltage. So the small currents needed for the microstepping will not be possible ". This seems to be the culprit of the missing microsteps when using the 8825 with motors rated with a lower voltage than the supply. At least is true for 1.8° steppers.

Like I said before, I haven't try this fix yet. I'd like to know if its practical or better than the fast decay setup or if there's a more simple solution to the missing microsteps problem.

Regards

Alex

[Cedric02]

Hello The_Geth.

  With my old test motor the fast decay mode solve the problem. With my final motors (nema14 3.6V 1.2A) the fast decay mode don't solve the problem... I try with the soltion with the diodes ( in link of The_Geth post), it work fine, but the diodes are very hot... i want to try with the a4988 if the problem persists. when i receive it i post the result.
  Has anyone have a config work well with the drv8825 in 1/32 microstep, can you tell what motor you use? i think the problem appears when the voltage motor is under 5V...

Cordially, Cedric.
 

[TCWORLD]

Missed some of these posts when I was on holiday.

That block on the DRV8825 makes for interesting reading. I hadn't read about that behaviour before, but it makes sense. I will make a page in the FAQ section about fast decay mode - will experiment with my mount and motors when I get back home at Xmas.

Cedric, could you let me know where you got your driver boards from. If they don't have the via, then they are probably clones which may not have the same design (current sense resistors may have different values for example). I know of a couple of people who used clone driver boards in their build and the quality of the clones isn't great.

[Cedric02]

Hello,

My drivers are the clones of drv8825. I don't think if the quality is at issue, because the problem explanation lies is in the drv8825 datasheet. I will anyway check the curent sens resistor value because when i set the vref at 0.6V for 1.2A the motors are a little hot.

[doramason]

Hi guys,

First I want to thank Tom for his help with my DIY Astroeq project.

I have been trying to do some astrophotography for the last month with pulse guiding using PHD 2. Everything is fine with the RA axis but I have been having issues with the DEC. There is erratic movement which I have thought was the cause of backlash in the worm gear and the bad/sticky lubricant that is used in Skywatcher EQ5 mount. So I was planning to disassemble the mount grease it with better lubricant and change the bearings. I will probably do it anyway but after reading this post now I am also suspicious about the missing microsteps as well.

I have used a Pololu Nema 17 4V 1.2A stepper for RA and Pololu Nema 14 2.7V 1A for DEC. I have no troubles with the Nema 17 other than it is heating too much even when it is under rated current, so I have plastered a heatsink to its back. But with the Nema 14 I have also noticed the missing/jumping of steps while I was testing it at 1/32 microsteps. It moves smoothly for maybe half a step then jumps forward like a clock ticking (I’m exaggerating it a little bit). And when combined with sound it is like a pulsation. Is it like what you observe as missing steps? As Tom has suggested I have changed the clone drivers with original Pololu drivers but it hasn’t improved. I have also changed the stepper with a 4.6 V 0.8A no brand Chinese one but the pulsation still exists. When I use full step or half step instead of 1/32 microstepping it vanishes and everything moves smoothly.

I want to have the highest resolution possible so I want to be able to use microstepping. Which method do you suggest putting diodes or doing the fast decay mode? Or is this problem not caused by missing microsteps at all? By the way I have noticed the pulsation at low speeds while stepper was sitting on a table, I haven't noticed anything while it is connected to the mount. And in high speeds I don't notice anything.

[Cedric02]

Hello

For my lost microstep. My motors are 3.6V 1.2A 3ohm.
On DRV8825 the PWM frequency (for current control) is 30Khz and the blank time is 4µs (time to the driver "measure" the current).
1 period of 30Khz is: 1/30000hz=33µs and 4µs is 12% of 33µs.
I powered my motors with 12V. 12% of 12V is 1.44V.
So with my motor the current with 1.44V and 3ohm coil resistance that is 1.44/3=0.48A.
In 1/32 microstep at the second microstep the driver limit the current at 5% of the motor current so for me 1.2A*0.05=0.06A, it's not possible for the DRV8825 with my motor.
At ninth microstep the driver limit the current at 38% of the motor current, for my motor that is 0.456A;
So i lose 9 microstep/32 microstep...
I do not know if it is fully correct, what do you think?

[Cedric02]

hello

Today i received my A4988. They work fine with my motors!!!

Cordially, Cedric.

[Cedric02]

Hello Doramason, i missed your post. For the " jumps forward like a clock ticking",  yes it's the same for me. Are you sure that the Ra motor work fine? Following a star without autoguidage, it's ok? If yes, I ask myself questions about my problem with DRV8825... For your problem I would advise you to change the DRV8825 by A4988. You lose the 1/32 microstepping, but it's cheaper than changing the motors...

Cordially, Cedric.

[TCWORLD]

I've never had any issues with the DRV8825's, but then I didn't know about the 12% voltage limit until this thread.

I'm going to look into the different solutions. Until then you can try any of the approaches, and see what helps. The diode trick is probably the easiest to wire up and test, but the diodes can get quite hot - probably worth looking at TO220 packages as they are easier to heat-sink. Fast decay mode may help if you can get a wire soldered on to the pin - but that can be very tricky without a fine tip soldering iron and a steady hand. A4988 should work fine, but as has been pointed out you loose the 1/32 micro-step mode.

Out of curiosity, does the issue go away if you run the DRV8825 in 1/16th mode?

Another thing to try come to think of it is to use a lower supply voltage. If you have say a 9V supply handy (or something >8.5V), you could try AstroEQ with that. This will mean that 12% of the supply voltage is a smaller amount, so you may find it works better. Worth trying if you have one to hand.

[doramason]

Cedric I have looked more carefully and yes the RA motor (Nema 17 4V 1.2A) has the same pulsation. I can hear it but the visual ticking is much less than the nema 14 motors. With good drift alignment I can expose up to 60 seconds or maybe even more and the RA graph is good in PHD. 

I think the nema 14 4v 0.8 has less ticking than the nema 14 2.7v 1A but they are almost the same.

I have tried the 1/16, 1/8 microsteppings, they have ticking as well but it becomes less and less evident. And in 1/4 I couldn't notice anything. I will try to find a 9v supply and try again.

Btw, I think I broke the potentiometer on one of the drivers, the vref jumps from 0.380v to 2.5v directly. Is it something fixeable or should I change the driver? I'm afraid it might not hold the current and damage the stepper. Or am I worrying too much and should I just leave it be?
Also the nema 17 4v 1.2 A stepper is getting too hot I can bearly touch it and I can't even touch the heatsink on the driver. I set the vref at 0.55v so the current is 1.1A. Is this normal and there is nothing to worry about or should I set the vref to something smaller?

   

[Cedric02]

Hello,

The 12% is in the datasheet, but not clearly. The minimum blanking time is set at 4µs and the dead time is 33µs(30Khz). So 12% of the power supply (4µs on and 29µs off).
I have a 1/16 microsteps it's a little better. I have tried with the 9.2V power supply, it's little better.
On the web the problem is know, try a search on google with: DRV8825 low motor voltage.
But i don't departs the possibility of my DRV8825 isn't great quality... especially after the post of Doramason...