GrauchoMarx:
Since the driver was getting SO hot (also the motor) it totally seems that is a overheat problem.
Now my doubt: is it possible to run a stepper with that driver? What voltage / current?
Thanks!
I have ordered a bunch of A4988 for some new NEMA17 steppers with the shafts on both sides, so at least I can put some yumo coders on the back end of them. I used a L298N to drive a small YM2754 uni-polar stepper reconfigured for bipolar operation, and was going to "try" to switch the Enable A and Enable B via software (like use PWM control on them), to see if I can reduce the power going to the stepper, since the stepper was getting hot to touch after a while. It should be possible to drive a stepper with a L298N if one really wanted to. I'm just going to try it out for the NEMA17 while waiting for the A4988 modules to arrive. If efficiency isn't a concern, then there is such things as those mini fans that they mount on raspberry pi's etc, for passing air over the L298N heatsink. That would be to see if extra cooling sorts out the L298N cut-out issue.
As for determining voltage/current to use. I'm thinking..... just operate with usual voltage or current specified for that stepper. And if things get too hot, then gradually reduce something, like the voltage, or the current. And figure out the operating envelope..... see how the unit behaves over a range, and pick a condition that works well.
They used to do current-control for steppers with great big series resistors in the days before switch-mode
power electronics was cheap - the old method does the same job of limiting current but with
massive amounts of wasted power (80 to 90%).
Robin2:
With an A4988 set the correct current limit at the start and there will be no need to reduce something
...R
I agree Rob. Current limiting will make things so convenient.
I ended up trying the L298N with NEMA 17 stepper, with the ENA and ENB pins of the L298N driver board each connected to self-configured PWM outputs of the arduino MEGA. The PWM duty cycle was just chosen to be 20% while feeding a voltage of 12V DC into the L298N (to be used to supply the motor). This was just driving the shaft only. The L298N heatsink feels only barely warm for this case. No problem at all. The PWM frequency was just set to 490 Hz, using the advice very kindly provided by: http://forum.arduino.cc/index.php?topic=72092.0
I confirm that feeding full 12V DC into the L298N won't work, because the motor coils require some minimum amount of current, which the L298N cannot supply (due to it's maximum power V*I constraint). To get around this, the PWM will reduce the average DC voltage, and at least the coils can get enough current to move the shaft properly.
I'm only driving the NEMA 17 shaft only - no extra load on it. And just using bread-board and vero-board. It's only a test set-up, and always keeping an eye on things like temperature and power dissipations in certain spots.
For all those that want to know if the L298N can drive a bipolar stepper (or even a 5-wire uni-polar stepper converted to a 4-wire bipolar stepper). The answer is yes - you can. Although, using other drivers like A4988 will improve things, like efficiency and size of driver circuit (more compact).
I didn't use PWM on the ENA and ENB pins of the L298N for YM2754 motor. For that case, the ENA and ENB pins were 100% enabled using the L298N board factory-provided jumper pins.
Robin2:
I agree Rob. Currently limiting will make things so convenient.
I prefer Robin.
I was also referring to an A4988, not an L298.
...R
Thanks Robin2. Was abbreviating your nick/alias/username.
I was referring to "for any case" - current limiting will make things convenient. It will be easy to do current limiting on L298N. But if the A4988 can do it all easily already, then might as well use A4988.
For those that have a L298N board on hand, then a L298N can be used to drive relatively small stepper motors - just depends on requirements of the motor. The L298N has been driving my NEMA 17 really well. I just needed to have a low enough motor supply voltage to get the coil current to be high enough to drive the motor properly, and yet not overheat the L298N chip at the same time. But not too low a supply voltage where the L298N electronics can't function.
I decide to test the L298N (as I have only used the MonsterMoto shield - which work fine).
My notes:
Used a Duinotech "Stepper Motor Controller Module" which is typical of the boards out there.
Ran a 17HS19-16844-PG5 (a geared Nema 17 stepper rated at 1.68A and 2.8V per phase (1.68 ohm and 2.8mH)).
My supply voltage was 12.5V using a Li-Po battery.
Initial stepper speed was 1000 steps per second and the current draw was 0.6A total.
The heatsink was 50 degrees C.
At 500 steps per second the total current was 1.76A and the heatsink could not handle it (the chip shut down after a few seconds).
With an 85% PWM (at 20kHz) the current fell to 0.17A total (this I don't understand - should be proportional!?).
Power supply capacitor
The board (that I am using) has no supply capacitor!
So this means inductive currents run through either the power supply (good) or the chip (not so good).
So with battery supply should not be a problem but with an "electronic supply" it probably wont work well (i.e. chip overheating) without the supply capacitor (suggest 470u at 25v).
Heat sink
As the transistor saturation voltages a quite high (~3.7v at 2A) the chip will need quite a bit of heat sink capacity.
The tiny heat sink will not cut it - 50 degrees C at 0.6A (supply current) - No chance of 4A as per the stated capacity.
PWM inputs
The PWM inputs (ENA and ENB) do work but not very well.
Barely at 20kHz, some operational envelop at 10kHz.
The datasheet suggest it should be okay at +20kHz so I am a little confused why the test was so poor!
Did not bother testing at frequencies below this, you should get reasonable results using analogWrite at 980Hz or 480Hz providing you don't want to go too fast.
Recommendation
To use the capacity of the chip a MUCH LARGER HEATSINK is required.
Basically I am very disappointed with this board and will just through it away.
