Paul & Sabrina's cheap DIY 144v motor controller

[flores]

Yeah.. like these guys

https://www.youtube.com/watch?v=SOM7zE3j7-A

[jedsmd]

Originally Posted by MPaulHolmes

To change the time to 4.6 seconds before the contactor closes (this is only useful if you are precharging during the 4.6 seconds. If you just have the capacitors charged all the time, you could set it to something NONZERO but small, like 1. if it's set to 0, it's disabled, and I think it won't close the contactor. I'll have to doublecheck that though):

pc-time 46
save

That applies to me I will give it a shot.

[jedsmd]

Test run today with revised max current (800 amps) and accelerator dead zone (6%). Both setting are good!

Really nice -- Very smooth fine control for parking, no valet switch will be needed.

Acceleration is also very smooth and controllable.

8Khz whine is quite noticable not quite annoying but reminds you that you are driving an EV.

Drove about 15 miles with some stop and go and some hills. Base plate temperature rose to 150 degf. Any thoughts on 150 degf? At what temperature does it become a problem?

[MPaulHolmes]

150degF is 65degC. That's not too bad. The thermal cutback comes on at 75degC, and smoothly reduces max throttle until 85degC. The capacitor can be at 105degC all day for years. However, a finned heatsink would not be a bad idea. I was warned by Otmar (zilla) that it would be necessary, possibly with a fan, or liquid cooling. He made IGBT module controllers in the 90's and they were unreliable due to heat. The caps could have failed, which won't happen in our case. It could have been the IGBTs though. Do you want to try the spread spectrum switching? Or we can try higher frequency too (inaudible), and see the affect on heating.

[jedsmd]

Originally Posted by MPaulHolmes

Do you want to try the spread spectrum switching? Or we can try higher frequency too (inaudible), and see the affect on heating.

I'm a little confused here, I thought I had the spread spectrum already? If not I would certainly like to give it a try.

I do have a heatsink It just laziness that it is not in there now.

[MPaulHolmes]

Nope, no spread spectrum switching on your current hex file. Just plain old dumb 8khz. Let's try it. I'll send you the spread spectrum hex file tomorrow morning.

[MPaulHolmes]

Be gentle with your driving at the moment. I had the temperature protection disabled by mistake. I'm almost done making several changes. It should be ready in a few minutes.

My wife just got done doing our state online taxes, so NOW it should be ready in a few minutes!

BTW, if there is a single hardware overcurrent event, the controller will shut down, and you will have to cycle power. It would be preferable to clear the fault after some amount of time, but just for now during debugging I thought it would be best to shut it all down if that happens. You don't want to keep clearing hardware overcurrent faults if they happen in rapid succession. That blew up one guy's controller a couple years ago.

One thing I'm very curious about: Will the driving smoothness be affected by the spread spectrum switching? The period keeps changing, which may complicate the PI loop convergence. But it might be OK. One test is worth a 1000 theories, as they say. haha

[flores]

I got the rs232 shields today.. did some research and decided that I'd better take a raspberry pi.. haha...

More on that later..

[jedsmd]

Originally Posted by MPaulHolmes

One thing I'm very curious about: Will the driving smoothness be affected by the spread spectrum switching? The period keeps changing, which may complicate the PI loop convergence. But it might be OK. One test is worth a 1000 theories, as they say. haha

Just took it for a 4 mile test run with the spread spectrum software. Drive-ability was unaffected. It was just as smooth as with the 8 khz.

The sound was quite different instead of a whine it now makes a hiss. The best I can describe it is the truck sounds like it is leaking air. It is about the same loudness as the 8khz whine but it blends in with the background noise and is much easier to ignor.

It didn't seem to be heating any different than with 8 khz pwm but data logging and more driving would be required to be sure.
Jerald

[MPaulHolmes]

We can also try a higher frequency (say, 15KHz) if the hissing gets annoying, but there will be more heat loss. Instead of spread-spectrum, I wonder if we could do the Jetson's car sound? I bet we could. It might take quite a bit of memory, but we have the space.