Paul & Sabrina's cheap DIY 144v motor controller

[MPaulHolmes]

I tried the smooth transition from throttle controlling current at 0 throttle to throttle controlling pwmDuty at max throttle. It felt really nice. I didn't do any hard testing, but it felt really good around the block.

The rate of transition from throttle controlling current to throttle controlling pwmDuty could easily be changed too. If people like the curtis feel, they could set the full transition to happen much sooner than full throttle, with the added benefit of not having a jerky start.

[JayC]

Just thought I'd share this interesting tidbit I just read:

Heat pipe - Wikipedia, the free encyclopedia

The advantage of heat pipes is their great efficiency in transferring heat. They are a much better heat conductor than an equivalent cross-section of solid copper.

And here is a guy making his own heat pipes (haven't read this yet).
Benchtest.Com - Heat Pipe 1

[MPaulHolmes]

awesome jayc! very nice read.

[jyanof]

Originally Posted by MPaulHolmes

It fixed the voltage problem you were having? You mean the sudden cutoff of 80 amps? or something else. the 80 amp thing has been keeping me up at night!

it fixed the problem where my voltage dropped to zero because my contactor opened before the PWM turned off and it drained the caps. Now, no matter how fast I lift off the pedal, the PWM is off before the contactor opens.

Still getting the 80 amps minimum thing... I'll take some video and post it. We should try the serial data logging too, but I'm mostly positive that I can't get below 80 amps. I tried going through the code to determine if it were possible for the throttlePos to go from 0% to 16% (80/500), but didn't get anywhere...

I'm glad it's not a problem for you - likely means it's not a controller hardware thing...

[adamj12b]

Hi Guys,

So here is something I have been working on. I have been working on a version of Paul controller capable of very high power output. I started with a the heat sink from a car stereo amplifier. I then designed a power board that fit in the heat-sink. The power stage board holds 21 Mosfet/Diode pairs. I have upgraded the diodes from the 60A STTH6002CW 's to 90A APT30S20BCTG 's. The capicators have been upgraded to 820uF from the 470's. There is 42 capacitors. Totaling 34,440uF. The controller should be capable of about 1680A continuously, With the pairs capable of 6720A for 8.3ms. At 165V that is 1.108 Mega watts of power. The continuous power is about 303kW of power. The controller will have the option of being water cooled if I need it. If it can stay cool by just the air moving over it, I will not using the water cooling, Saving weight.

I plan on using this controller at 162V, driving a Warp 11.
Paul thinks that there is a chance that the 200V components will be able to handle the extra voltage.

I have not tested the mosfet/diode pairs yet at 162V to see if they will hole up, but I have some parts on the way so testing can begin.

So here are some pics of the heat-sink and some renderings of the completed controller.


Heat Sink:




Renderings:





No Water Blocks:


With Water Blocks:


-Adam

[esoneson]

Thanks Adam!

The one big question that I have is the same one you have: will it be able to handle 168 Volts? I guess you will find out soon enough.

Great looking heat sink packaging

Eric

[MPaulHolmes]

Go Adam, it's your birthday, light a candle, throw a party! Adam's going to test the voltage limits. I'm really excited. He has 24 12v SlA batteries for testing. Ya!

Beautiful pictures! I have got to get a program like that! I want water coolign too!!!!! not fair!

[jyanof]

Originally Posted by esoneson

Thanks Adam!

The one big question that I have is the same one you have: will it be able to handle 168 Volts? I guess you will find out soon enough.

Great looking heat sink packaging

Eric

hm, so a couple things I haven't told anyone about yet... I've found this all out by accident, fortunately no big deal.

1st, the controller survives being turned on and off while the caps are precharged. One time I screwed up my sequence of 1-turn car/controller on 2-precharge caps. Several times, I was in traffic and the micro froze (this was before we figured out the problem), so I had to cycle the car on and off to reset it. didn't have time to drain the caps and precharge again.

2nd, the controller can handle at least 175V because once I forgot to disconnect the pack with my manual circuit breaker before I charged it. It probably got up to at least 175V. Embarrassing, but the controller still works!

That being said, if you charge to 2.4V per cell for SLA's, then that's just a hair over 200V. You may have to wait a bit for the voltage to drop before connecting your pack, not to mention that a cap's voltage rating has a tolerance and the rating apparently decreases with age...

I bet if you're careful, you'll be fine!

[MPaulHolmes]

Joe, you sneaky sneaky little bugger! hahaha!

[jyanof]

here are two clips from a trip I took on saturday. i guess they aren't clips - it's a 20 min video, and youtube doesn't accept videos longer than 10 mins. It was a total of 34 miles, but I don't show it all. I had just fixed my temp sensors (needed to upscale some variables before doing some calculations), so I wanted to make observations and see how the controller performed - I hadn't gone more than a few miles til then.

Conclusions:

The temp probes were touching the top of the mosfet and at 200A on the highway, were reading about 65C, or about 25degC above ambient (42C). Unlike what I say early on in the video, this seems to be worse than typical city driving where there's ample opportunity while coasting or waiting at lights to cool down.

at ~240 amps continuous in 4th gear, the temps rose about 10degC to the mid 70s. I don't know if the temp limiting was coming on because 240 amps is still significantly lower than 500 amps and the thermal cutback reduces current in increments of 12.5%.

I think these temps are really just ballpark numbers and it's likely that the internal mosfet/diode temps are a bit higher, but it's interesting nonetheless.

The caps have significant thermal mass and no active cooling. They take a while to warm up, but when they do, they're just as hot as everything else, maybe just a couple degrees cooler. When back in city driving and waiting at a light, everything else 'cools down' but the caps are still hot. Maybe a fan to pump some air through there would help - if we determine it's needed - at the risk of blowing dust and other crap in there. maybe a fan and a filter? AZ is very dusty...


overall, the controller performed great!


So, if you want to watch 20 mins of cruisin, here's part1:

http://www.youtube.com/watch?v=mDQLjwKT7vY

And here's part2: (which if you view mithin minutes of me posting this, is probably not completely processed by youtube)

http://www.youtube.com/watch?v=Z8JpIhaZRy0