Stepper Motor Driver Spec Comparison

13 thoughts on “Stepper Motor Driver Spec Comparison”

1. Ah, your getting caught in the trap of labeling. So most if not all the drivers you listed (sorry I did skim a bit) use DIP switches to set current. So when they say thing like minimum or maximum, that’s just a very poor indicator of the min and max switch combos for current. So what is now of huge importance is the exact motor you intend to drive. You have to look at each manual and figure out what the nearest setting that driver can do via the DIP switches compared to your motor. So for example, you may have a NEMA17 and require 1.2 Amps but you don’t want to buy the smallest drive in case you think you might upgrade and drive a bigger motor one day. Well, the issue is that that 4.5 Amp driver has very corse jumps via the dip switches. You might only have a low setting of say 0.85Amp and then the next jump is 1.3 Amps. So you either clip or underdrive the stepper.
   Sorry Ed, I recently played this game and it’s not fun. You could modify the drive with a POT to replace the dip and resistor combo but that kinda defeats the purpose doesn’t it? You have to know up front what your target current is, open every data sheet and determine if the settings get you in the range you want to be in.

   1. getting caught in the trap of labeling.

      Methinks you’re reading too much into a list of eBay drives… [grin]

      A stepper motor’s current rating isn’t a hard-and-fast limit. If you need less torque, you can use less current. If you need more torque, you can use more current and verify that the motor temperature doesn’t exceed its spec. All that’s within reasonable bounds, of course, and when you’re pushing the limits you need to understand what’s going on around the motor, too.

      The 8 current settings for the M542 drives differ by about 500 mA, upward from 1 A. That means the power dissipation increases by factor of 2.25 from the 1 A to the 1.5 A setting, which may or may not raise the temperature too much. It depends on whether you’re sealing the motor in a plywood box [grin], bolting it to a big metal frame, applying active cooling, or just slapping a “Caution: Hot Surface” sticker on it.

      Running a 2 A rated motor at 1.9 A, the nearest setting, costs you only 5% torque and drops the disspation by 10%. Running it at 2.4 A (the next setting) gives you 30% more torque at a cost of dissipating 60% more power. I’d favor the lower current, but that’s a tradeoff.

      Now, I’ll agree that running a 0.5 A motor from a 1 A drive wouldn’t be a Good Thing, but I don’t plan to do that.

      More current resolution would be nice, but not having to build well-protected drives may be much nicer.

2. I will say this, the 5XX series drivers are from an a well proven design. Both my laser cutters have them and I got a set specifically for that Ultimate Thing-O-Matic upgrade frame, but haven’t yet tried them out. BTW I like the pricing and service from http://www.lightobject.com/CNC-CO2-Laser-C11.aspx
   They are very fast to respond to emails or any question and never had a problem. Hands down, my go to place for stuff like that.

   1. And they have little bitty drive boxes with little bitty steps between the current settings! [grin]

3. You might want to look at the American Precision CMD-260 could be worth searching for. I got several when the basement storage at our HP facility was being cleaned out. (Said cleanup announcement came after they dumped a test system and wafer prober I had had transferred from another division. Arggh!) (Not selling any of mine–I have some projects in the round tuit bin that want them.)

   Anyway, these are DIP programmable from 0.25A to 8A per phase, microstep at 16 semi-logically spaced resolutions (maximum 50,800 steps per rev), need 24 to 80 VDC. A fast search on Google shows information, but I managed to misplace my only data sheet. These date to the mid-90s and are industrial type. Form factor is 5 x 5 x 7/8″ or so. No idea of speed. There’s a more information out there via timgoldstein dotcom site and the rest of the search engines.

   For what it’s worth, the Princeton Monitor responded nicely to a capacitor transplant. I’m looking at my old Mac Classic II and it’s ugly. The system board has a bunch of surface mounted electrolytic caps, and most, if not all leaked over the board. I’m not set up for surface mount repair and I’m not sure it’s worth a permanent fix. Apparently the board can be recovered for a while by washing it, but beyond that, it’s still the odd computer in the bunch. I’m not willing to put an Appletalk network together to get data off the beast, so, it’s adios, I think.

   1. The gotcha is that I want this setup to be reasonably easy to replicate, so the usual weird crap that I’d lash together from my collection isn’t appropriate, much though it pains me to say that. Besides, those antiques would need a re-cap job to make them useful… [grin]

      The new brick drivers with low-resistance MOSFETs are a wonder to behold, readily available, and reasonably inexpensive. Hard to argue with that!

      recovered for a while by washing it

      Eeewwww!

      The guys in the local hackerspace do wonders with a hot air gun, but … at some point you (well, I) must take a deep breath, admit it’s not worth winning that particular contest, and drop the corpse in the (electronics recycling) trash.

   2. FYI, the Mac II uses Perfectly Normal SCSI drives containing Perfectly Normal HFS volumes. Very easy to connect to any modern system (well, any modern system worth talking about: Linux, BSD, or OS/X) and copy your data off. No Appletalk needed.

4. “A stepper motor’s current rating isn’t a hard-and-fast limit.”

   I counter that well, then why do we worry about tuning stepper drivers so that we don’t clip current on microsteps and end up with noticeable artifacts from the current clipping in motion? For CNC cutting, so maybe you take a little more bite (feed speed variation VS the bit assumed constant speed), but in FDM, the correlation seems to have more of a noticeable affect.

   Maybe there is more headroom and to be honest, I haven’t tuned one via o-scope, but in general I know both from noise and output, that 5% under rated spec is usually a good spot to be.

   And, I’ll be dead honest here, I’m seeing artifacts in my laser cutter on the Y axis. That particular motor is an 8 wire version. I have the coils in parallel. It is noisy jogging and has caused enough vibration to loosen the mirrors in the mounts once. I thought Oh, maybe I’m over driving it but lowering the current didn’t help. The X motor has identical wiring and no such issue. It could be that I need to abandon parallel and either only drive one set of coils, or drive it in series. It’s the 57J1880-830 http://www.novanic.co.uk/catalog/files/255_2M_57_E.pdf
   The motor never gets even warm and both X and Y use the same pulleys. I will say that looking back, pulley size/teeth count, is not optimal in that we are assuming midly accurate microsteps to get reasonable resolution. Basically, I see a pattern only in Y, more noticeable when I cut acrylic that appears to be microstep clipping related, hence periodic in nature and the sound too is telltale.
   Or it could be, looking at the data sheet, the low resistance and inductance is more than what that driver can handle in controlling microstepping and current.

   I also acknowledge the obvious answer would be incorrect wiring/phasing, but I assume I would have a bigger issue in that it simply wouldn’t turn or would be so bad as to be impossible not to notice from a noise alone perspective. And it’s been double tripple quad checked.

   All I’m saying is, I don’t exactly agree with the thought the current limit is there for heat only. I believe there can be saturation issues and thus impact velocity, especially in a critical application.

   In my specific case, I’m either looking into changing wiring, or new motor (4 wire only), or going to the 2:1 reduction drives to increase my physical resolution. Due to mechanical issues, just changing the pulley on the shafts in this case is not feasable as the simple fix. (being the new versions use reduction drives and my old smaller laser cutter used 0.9 degree motors, I have a feeling that is the answer).

   1. so that we don’t clip current on microsteps

      That harks back to the original MBI stepper debacle, where the motors were designed for H-bridge drivers without current limiting, then used with a microstepping driver from their rated voltage. Under those conditions, the driver can’t do microstepping at all.

      For a properly designed microstepping system, that doesn’t happen: the driver limits the winding current to the current microstep value. That’s not sensitive to small variations of the current, stepping rate, or anything else.

      there can be saturation issues

      That won’t kick in until the current is far over the spec. I’m talking about minor variations around the nominal rating, not huge multiples.

      I suspect the Y axis on your cutter has a broad mechanical resonance that’s excited by the step frequency; it will be much different than the X axis. It’d be interesting to change the microstep settings (and adjust the steps/mm) for just the Y axis: double or halve the number and see how that affects the results. The simple study I did on resonance may provide some useful suggestions. I’d also look at the mechanics for problems like loose nuts, because I’ve been bitten by that far too often.

      Also, that’s a pretty beefy motor. For a parallel connection, you need 1.4 times the winding current for full torque, so you’re looking at nigh onto 6 A and 13 W dissipation. If the motor is barely warm, then it’s not disspating anywhere near that much power, it’s not getting anywhere near the required current, and it’s not delivering the torque you expect. While it’s not missing steps (you, of all people, would know that [grin]), it may not have enough torque to stabilize the load: if there’s any resonance, you’ll see it.

      The parallel inductance is half the separate winding inductance (roughly, as there’s some mutual inductance), which means the current limiter will be running faster. I don’t know if the drive has a minimum inductance limit, but at some point the poor thing can’t keep up. If all else fails, rewire the Y axis for series operation (inductance roughly doubles), drop the current to 0.7 x the rated value (about 3 A), and see what that does.

      Nothing like a bit of system tuning to keep your mind occupied, sez I…

5. Ed, thanks for the thoughts and experience!!! And this is somewhat why I was shopping for a new motor. I don’t know but the very low resistance and inductance may not have been the best choice by the supplier of the system. I was going for a known good 4 wire version. And honestly, for the headache of trying different things, replacing the motor for around $40 is far cheaper than my valuable time. And if we’ve learned nothing from the beaten path, sometimes replacing the motor is the best decision we can make (hmmm wonder where that came from).

   1. for the headache of trying different things

      Well, heck, at least try changing the microstepping: that should be a couple of DIP switches and a config file tweak. If that changes the vibration pattern, you know you’re on to something.

      If you’re going for a new motor, you’ll be digging into the wiring. Before doing that, rewire the existing for series operation, use the same microstepping settings, and cut the current in half (depending on what you had already). It’ll certainly turn backwards the first time, but after you tweak the config file one more time, you might just be on the air.

      Cheaper and easier than getting a new motor, doing the mechanical stuff, and then finding out it’s a current setting!

6. Pingback: Stepper Driver Specs: 2M415 « The Smell of Molten Projects in the Morning
7. Pingback: Stepper Driver Waveforms: Current Control | The Smell of Molten Projects in the Morning

Comments are closed.