Speed Problem with Stepper

[mykill]

Hello this is my first post.

I've got a bit of a problem. I am working on a project where we need to use stepper motors for motion.

I designed an L297-L298 combo driver for the stepper based mainly on design for the controller of the reprap machine. I interfaced it with arduino and it works fine. The stepper motor I am using is KL34H280-45-8B.

www.kelinginc.net/KL34H280-45-8B.pdf

The only problem is that the maximum speed I am able to reach is of 170 rpm. This is too slow!! I am sure the motor is capable of faster speeds. Above that speed the stepper starts vibrating and produces no motion.

I am current supplying 12V to the motor via an old pc psu. Is this the reason why I am not able to achieve higher speeds or is it because a max current limit of 2A is too low? Or is it something software related?

I attached the code.

Hope you can help me out.

Thanks.

[johnmac]

I think you need to replace the PC power supply with one that produces more voltage. Do not exceed the voltage rating of your stepper driver. More amps will give you more torque, too. Again, the driver may be the limiting factor.

John

[ger21]

If it's wired Unipolar, then you're getting less than half the rated torque at 2 amps. And for maximum rpm, you'll want about 48V, and you'll need to supply 4.5amps.

[James Newton]

Actually, if you wire it bipolar, make sure you use parallel bipolar, and not series, because series bipolar will actually top out at a lower RPM than unipolar; all other things being equal.

Unipolar looses torque less rapidly at higher speeds than people think. In fact, unipolar and parallel bipolar actually are pretty close to each other on top end speed in most cases.

See:
http://techref.massmind.org/techref/...onnections.htm for comparisons of the different types of connections and the advantages and disadvantages of each.

The primary limiting factor for top end speed, as far as I know, is how fast the coil can change it's magnetic field strength to pull the stator to the next position. A motor with lower impedance will turn faster. Higher voltage, as John and Gerry suggest, will allow the same driver to push the field to it's desired strength faster.

Another feature that will increase speed is active current regulation. If the driver measures the current flowing through the coil, and turns the FULL available voltage on until the current (and so the field) has built to the desired level before backing the voltage off to the correct level, it will allow faster movement with the same voltage.

I know the L297-L298 senses current in order to limit max drive, but I'm not sure if it actively regulates current buildup between steps. Anyone know?

[serriadh]

Originally Posted by mykill The only problem is that the maximum speed I am able to reach is of 170 rpm. This is too slow!! I am sure the motor is capable of faster speeds. Above that speed the stepper starts vibrating and produces no motion.

Sounds an awful lot like 'mid-band resonance'; a problem that can effectively paralyse steppers in the slow-to-fast transition. Google for it (many of the results will be from this site!) to see how other people go about solving or avoiding the issue.

Off the top of my head, using less than full steps (I'm pretty certain your setup will do half-stepping with no additional effort on your part) helps a reasonable amount. Also try just pushing it faster to get out the other side of the resonance band.

[wildwestpat]

Note that you say 'we' need more speed. Are you developing the motor drive as a separate part of the project? I only ask as stepper motors work best when they have a load to work against. With no load the spindle of the motor can twitch as the magnetic fields are switched. The action is to oscillate about the step mean position and this rapidly develops into a larger oscillation that eventually stops the motor rotating. This has been pointed out by Serriadh.

There is not much point in testing further without the mechanics as the addition of the load will alter the resonance tendancy.

Try accelerating hard past stall point by demanding say 300 rpm. If the motor rotates then you have a mid band resonance problem and could just program out any demands for speeds in this band.

There threads on this site giving details of dampers (rattlers) or just tighten the mechanism.

A quick test is to hold onto the motor spindle (wear leather gloves if it is a big motor!) between the fingers to add load and try accelerating to see if the stall speed is increased. If the speed is increased by pinching the motor shaft then you have a case of resonance to sort out. Add daping in the form of friction or mechanical inertia.

Hope this helps Regards Pat

[davidmb]

Hi mykil,
one thing to watch out for is the 5v logic supply must be supplied to the 297/298 before the high voltage, otherwise the 298's sort of disappear ( speaking from experience ).

[James Newton]

serriadh and wildwestpat are absolutely right, and I forgot to mention it as well: Resonance could also be causing your problem. Resonance is part of any stepper drive system just due to the fact that they step! In addition to mechanical damping through inertia or friction, moving from a full step mode into a microstepping mode will reduce resonance. Also, some drives, like our Linistepper, can further reduce resonance by smoothing the transitions between microsteps to practically eliminate resonance electronically at certain speeds. I believe the Gekko drivers do this as well, although with a different method since linear drive is not practical at the higher powers they support.

[mykill]

I've tried giving the stepper more voltage however I did not see an increase in speed. The stepper is part of a pick and place mechanism and it makes part of the motion system of the x and y axis. The stepper is coupled to the screw thread via a sprocket and chain. One thing I noticed is that when I give the motor certain rpms the motor doesn't turn and when I help it a bit with my hand it starts the rotations at that speed.

I assume that it needs more torque to overcome the initial inertia. I will try to build a higher current driver.

Also I read about accelerating the stepper to a desired speed however I am uncertain how to go about this. When I tried implementing such a system on software I did not achieve the desired results. Any ideas on this matter???

Thanks.

[irving2008]

All of the above comments may well be true, but I'm guessing you are running this motor in bipolar series configuration? If thats the case you'll never get anything like the speed you want with a 297/298 design simply because of slow decay characteristics of the L298 driver and the inductance of the motor...

I have motors that should in theory run at 600+rpm but in practice wont get about 300 before they stall (running on 36v which is the max for the L298 driver), and, before anyone says it, I know its not midband resonance. Look at the traces in the pictures..

Trace 1 is at 150rpm - see the 0.7mS delay before the current ramp... thats the L298 recovering from the back-emf of the last reversal. Then you can see the nice ramp up of current till it hits the current limit and the chopper maintaining that limit (in this case 1.4A).

Now look at trace2 - 225rpm. Same 0.7mS delay, same ramp, same chopping, but less of it of course...

Trace 3 - 270rpm. Same 0.7mS delay, same ramp but now we dont get any chopping because theres no time left before the next cycle...

Anything faster than 270rpm with that motor and it never gets enough current into the coil.. even off load it stops spinning and makes graunchy noises at about 290rpm but you can give it some hand assist and it will run. But 300rpm is the absolute limit! Adding damping or load makes no difference at all. It simply that the bipolar transistors in the driver cant decay the current in the coil fast enough because at no time do the two bottom transistors turn on together to 'short the coil'. I have other (MOSFET based) drivers that will run the same motor at a much faster rate.

[mykill]

Originally Posted by irving2008 All of the above comments may well be true, but I'm guessing you are running this motor in bipolar series configuration? If thats the case you'll never get anything like the speed you want with a 297/298 design simply because of slow decay characteristics of the L298 driver and the inductance of the motor... I have motors that should in theory run at 600+rpm but in practice wont get about 300 before they stall (running on 36v which is the max for the L298 driver), and, before anyone says it, I know its not midband resonance. Look at the traces in the pictures.. Trace 1 is at 150rpm - see the 0.7mS delay before the current ramp... thats the L298 recovering from the back-emf of the last reversal. Then you can see the nice ramp up of current till it hits the current limit and the chopper maintaining that limit (in this case 1.4A). Now look at trace2 - 225rpm. Same 0.7mS delay, same ramp, same chopping, but less of it of course... Trace 3 - 270rpm. Same 0.7mS delay, same ramp but now we dont get any chopping because theres no time left before the next cycle... Anything faster than 270rpm with that motor and it never gets enough current into the coil.. even off load it stops spinning and makes graunchy noises at about 290rpm but you can give it some hand assist and it will run. But 300rpm is the absolute limit! Adding damping or load makes no difference at all. It simply that the bipolar transistors in the driver cant decay the current in the coil fast enough because at no time do the two bottom transistors turn on together to 'short the coil'. I have other (MOSFET based) drivers that will run the same motor at a much faster rate.

Thanks for the explanation. As a matter of fact I am using the stepper in bipolar series configurations. What mosfet based drivers are you using to run your steppers? Are any of those open source or have you bought them?

Thanks.

[irving2008]

Originally Posted by mykill Thanks for the explanation. As a matter of fact I am using the stepper in bipolar series configurations. What mosfet based drivers are you using to run your steppers? Are any of those open source or have you bought them? Thanks.

I borrowed them but they are a commercially available unit similar to these and to be honest, for the price, I wouldn't build another set again, there's little difference in cost.