DRV8825 or A4988 for extruder control? Does it matter?
Edit: I guess I should have mentioned that I am planning 1.75mm filament feed.
The DRV8825 offers more microstepping at higher amperage max but the microstepping only decreases possible speed and reduces torque. Now having the option between either of the two stepsticks, which one would you use, at what microstepping factor and what are the reasons for using one over the other.
Thanks.
More microsteps can make movement more silent. To realy improve precision use a stepper motor with 800 steps instead of 400
800 steps? 400 steps (0.9deg) have barely started to take off and be adopted by the reprap community I haven’t heard of 800 steps motors, that is 0.45deg… the cost must be crazy.
i use a drv8825 because it can handle higher amperage, which i sometimes use for faster prints.
you can configure microstepping using the jumpers on your ramps board, so in case you’re not happy with the drv8825’s performance with it’s maximum microstepping you can always go back a setting to the a4988’s microstepping and get higher torque.
Don’t use 8825 on 24v, it will produce rippling/pulses in the motion because it can’t hit the really low current rising microsteps. It usually works ok on 12v unless your motor is very low inductance. You can plug your specs into my motor/driver simulator to check for yourself: https://github.com/rcarlyle/StepperSim
Microstepping does NOT decrease torque. Two 1/32 microsteps is exactly the same as one 1/16 step. Finer microstepping applies the torque in smaller, smoother chunks, but the torque response of the motor is identical. It will always produce full torque at ~1 full step of position error (load angle), regardless of microstep level.
@Tom_Keidar
thanks for answering my next question 
… Right now the jumper is at the middle, which means something I guess. So getting it down on the microstepping scale means moving it to one side or the other… is there something I can visually follow to know which jumper setting is the lower factor?
@Ryan_Carlyle
I just read one of nophead’s forum posts explaining this and barely understood the lingo. He said he would increase the voltage as he increased microstepping and as a result torque will stay the same (down from here: http://forums.reprap.org/read.php?1,79959,79985#msg-79985).
yeah, probably should have mentioned @Ryan_Carlyle ’s simulator. i have consulted with it twice now… first time while deciding if to try the drv8825 and the second while deciding on a motor model for the 2nd printer i’m working on. it saved me from buying the wrong motors for what i wanted.
it really should be more well known!
Makes my head spin … will give it a go though.
@Florian_Ford Nophead is referring to motors listed with RMS rated current, which is the correct limit for both-coils-on-full-stepping. Those are pretty uncommon these days – almost everything in the NEMA 17 range is microstepped now. If you have an RMS current rating for your motor, you would need to set the microstepping drivers to 1.41x the RMS rated current to get full output torque. But usually the motor rating these days is given with peak coil current and you just set the driver current to the motor’s rated current.
He was also referring to the original Makerbot Cupcake motor selection, which sucked because they used low-speed full-stepping motors with very high inductance. They weren’t suitable for microstepping and had bad ripple issues.
None of that is relevant to microstepping vs torque in general. Torque is an approximate sine function of position error between the rotor and coil energization, with period equal to four full steps (or one rotor tooth). So peak forward torque is produced with 1 full step of lag between the rotor angle and coil energization angle. (And it reverses at 2 full steps angle error, which causes skipped steps.)
For loads <50% of max torque or angle errors less than a half step, the motor’s torque response is essentially linear with error angle. Which means microsteps add up linearly. When the driver advances the coil energization by two 1/32 steps, that is the same total rotor/coil position error as one 1/16 step. Likewise four 1/16 steps is equal to one 1/4 step.
If you just compare one 1/32 step to one 1/16 step, then yes, the 1/32 step provides less torque, but that’s a stupid comparison, because the controller will command twice as many 1/32 steps in the same amount of time. So it completely averages out. (Except that finer steps produce less vibration and noise.)
@Florian_Ford - the microstepping jumper settings for the ramps 1.4 board are listed on the reprap wiki page here - http://reprap.org/wiki/RAMPS_1.4#Stepper_Driver_Boards
tldr; it’s :
jumper Yes/No step size
1 2 3
no no no full step
yes no no half step
no yes no 1/4 step
yes yes no 1/8 step
no no yes 1/16 step
yes no yes 1/32 step
no yes yes 1/64 step
yes yes yes 1/128 step
so you are currently (if i understood your setting correctly) on the 1/4 step setting. to get full micro-stepping for the drv8825 board you can either put the 2 extreme jumpers in, or just put them all, since that driver doesn’t support 1/128 stepping it will go to it’s highest which is 1/32.
@Ryan_Carlyle - thanks for a very informative post.
if you were ever to write an article about practical stepper usage and facts, maybe a driver/stepper choosing guide i’d definitely spend the night with some good scotch reading the hell out of that
a bit of an off-topic, but notwithstanding the price and the specific motors would the silentstepsticks (tmc2100’s) be a good recommendation than?
Thanks guys for the informative responses, sometimes I feel so dumb, I consider myself quite capable IT engineer but with electronics I’m handicapped, can barely understand anything (having no previous training) but I am reading your posts and searching online for the missing pieces to have some reads.
@Nathan_Walkner I know about the reverse positioning of the DRV8825, one can actually be certain he has it right looking for the GND text on both the stepstick and the board. The GND pin on the stepstick should go to the GND receiving “black connector” on the board.
@Ryan_Carlyle so basically for 12V PSU DRV8825 with 1/32 is advisable as far as I understand.
@Tom_Keidar I’m halfway through writing a book on 3dp design, and I have the 20,000 word chapter on steppers written already… going to try really hard to get it out in time for MRRF 2017, but we’ll see.
@Florian_Ford 8825 on 12v is generally fine, but I recommend plugging your specific motor specs into the simulator to be sure. 1/32 stepping is quieter than 1/16 stepping but otherwise doesn’t change much. Depends on the controller you have… if your step pulse rate goes too high for the processor to keep up with (eg >10khz on Marlin) the controller will fire multiple step pulses at a time to save clock cycles, which effectively drops your microstepping to a coarser setting. In which case the higher nominal setting doesn’t actually help. Where this would be an issue, for example, is if your extruder retractions require so many step pulses per second that it has to drop the microstepping precision for your XY motion. Which, again, is stuff that the simulator can help with a little bit (although that specific limit is not too hard to calculate on your own via steps/mm and mm/s rates).