EasyDriver v3.1 torture test

Archive SparkFun Site Questions/Comments
1

For anyone who’s interested in the Allegro A3967 stepper motor driver chip (upon which the EasyDriver is based), I just got a new bench top power supply, so I had to finally test the EasyDriver to see how fast it could go.

Previously, using a small 12V wall wart, I could get around 20,000 steps/second maximum speed. This is with a small NEMA17 stepper motor that is similar to the one SparkFun sells. I drove the EasyDriver with a frequency generator.

At 12V, the EasyDriver gets hot enough that you can’t keep your finger on it. And it smells hot. But it never reaches it’s thermal cutoff of 150C at the die. Surprisingly, even at 20,000 steps/s, it still feels very torquey.

So tonight, I decided that I have enough EasyDrivers lying around here that I could push the input voltage and if it blew up, well, then I’d know where it blew.

So I ran it at 30V. At 10,000 steps/s, the EasyDriver was pulling about 200mA out of my power supply, even when it was set at maximum current. And, guess what. It gets -really- hot. My temperature gun measured the top of the driver at 135C. And then it started shutting down every couple seconds on it’s own. Due to over-heating, I believe. It would then start back up on it’s own after a short time. So the die was getting to 150C.

So, to keep going with the test, I put a small PC fan pointing at the EasyDriver. Then the device temperature was cool enough that you could keep your finger on it no problem. Fans are awesome. Also, without the fan, the little voltage regulator got up to about 110C at 30V input.

At 30V input, I got it all the way up to 44,000 steps/second. If I grabbed the shaft hard, I could get it to stall, so there wasn’t a huge amount of torque up there at that speed. But that’s what you’d expect.

The interesting thing through both of these tests was that because my motor resistance is pretty high (4 ohms per coil), I had to turn the current limit on the EasyDriver as low as it would go in order to get the motor to run smoothly. Oh, it ran at max current, but it was very rough. Gritty, like sand in your scrambled eggs. At minimum current, the EasyDriver was able to generate all 8 microsteps, and even at 44,000 steps/s, it was so smooth you could not even hear it. Couldn’t feel it. No vibration at all. Just a little tape flag that spun so fast it was a blur.

The other test I did tonight (based upon a user’s e-mail question today) was to see if I could fry an EasyDriver by supplying 5V to the STEP and DIR inputs without supplying any power to the EasyDriver. I tried and tried, but was unable to kill the chip. So that doesn’t seem to be one of it’s failure modes, at least the way I was testing it.

Anyway, I’m confident that if I unplugged the motor from the EasyDriver (this was all done on a breadboard) while it was powered at 30V, it would kill the driver chip. However, at 12V, I have been unable to get this driver chip to pop by unplugging the motor while hot, even with hundreds of attempts.

So, this is all just a little bit of info for anyone who’s considering using the EasyDriver for their application. If you need 20-40K steps/second, be prepared to use >12V input and a fan. :slight_smile:

*Brian

PS Just for kicks, I tried plugging and unplugging the motor to the EasyDriver while running at 3K steps/s at 30V input. It turns out that if you plug and unplug it slowly (maybe in for 500ms out for 500ms) then it is fine - it never pops. But when you plug and unplug it as fast as you can (simulating an intermittent open in the wire to the motor) it pops RIGHT away. And it really popped - huge puff of smoke from inside the driver chip, black liquid oozing out from the top of the chip, and a melted spot on the top where the smoke came out. My whole basement smells horrible now. Why-oh-why did I have to push it?!?

2

This is very good info, thanks!

I was wondering about the speed of EasyDrivers just the other day.

3

I hope that this does not become a torture test or a destruction test but a comment I’ve seen on one of the threads about the chip having the ability to do other modes than 1/8th stepping got me doing some thinking and exploring of datasheets and the EasyDriver board.

I’ve not tested it yet because I want to finish some other work first but tonight I took to the EastDriver with soldering iron and lifted the two pins which control the step mode and attached my own wires to them. I used solder wick to remove what solder I could then while heating a pin lifted the tab with the smallest of my jewelers screwdrivers. Some fine hookup wire soldered to the pin’s with my smallest heatshrink over the connection and I’m hoping that I will be able to control the step mode.

I’ve not yet worked out a way of providing physical protection for the connections, still thinking about that but in use the controller and motor will be housed away from accidental bumps.

I’m also think of ways to provide heatsinking for the board. I’ve got some thermal double sided tape here and am thinking of sticking the chip to a narrow heatsink via the upper surface of the chip. I’ve not worked out if I should be trying to heatsink the voltage regulator as well. My intended application is driving a telescope mount where I want very find control of the speed for tracking and fine adjustments and much greater speeds for slewing the telescope to other parts of the sky.

4

Bob,

In the very near future, an updated version of the EasyDriver will bring all of those pins out to headers so you can do with them what you want without fear of damage.

For heatsinking, if you can’t use a fan (which I have found completely sufficient and easy) then clamping the ED down with an aluminum bar across the top should take care of getting the heat away from the driver chip. As far as the regulator - hmm. That’s a tougher problem.

On the other hand, if you bring the driver into Sleep or Disable, you can vastly reduce the current consumed (and thus the heat) and only turn it on when you actually need to take a step or two. Unless you are stepping it for long periods of time, that will prevent heat buildup.

*Brian

5

Brian bringing the pins out will make it a lot more versatile. I’m planning to use the two I have at the moment in the drive for a telescope mount so one will be stepping at a low rate continuously and the other only used occasionally. I’ve not worked out if I can put one driver into sleep mode. I’ll be using worm gears on both axis but I also want it to hold rock steady on the axis that’s not stepping.

Bob

6

Bob,

It should work out quite well. With a worm gear, and enough tightness in the mechanical motion, there shouldn’t be any reason to need the stepper to remain energized at all times.

This sounds like a problem who’s best solution is mechanical in nature.

*Brian