Paul & Sabrina's cheap DIY 144v motor controller

[Roger Heuckeroth]

Sounds good guys. Keep the creative juices flowing!

[MPaulHolmes]

I got the car "drivable" with a PI loop, but it was really crappy and rough. It would take a very long time to discuss all the things that happened, and all the stuff I tried, so I'll skip it all, and just say that I found a nice, no PI (gasp) super smooth (smoother than my PREVIOUS Curtis controller. it sold. ya! $432!) It's sort of hard to explain, but I basically came full circle. It doesn't have to respond super fast because of the hardware current limiting, although I haven't been able to get the hardware current limiting to come on with normal to hard driving, because I just don't use 500 amps for goodness sakes! Call me a conservative driver if you want to, but only if you want a Karate Chop Sandwich!

I'm going to make up a new box out of Lexan, because the Plexiglass cracked, and the Lexan is AWESOME and unbreakable. I'll ship it to Ben before the end of the week. I just hope it runs as smooth on Ben's bigger motor. It will be fun to find out.

On version 2, I'll really spend some time making the current sensor output really smooth, with greater resolution. I only have 0.6v of resolution this time that spans 0 to 500 amps, and the noise is like 0.12v! That's bad! So using that current with the best PI loop in the world is still crap.

I'm going to spend some serious breadboard time next on version 2 finding the right layout. Also, anyone that wants to make it better, feel free. I know basically nothing about electronics. I know that sounds like false humility, but it's true. I'm just winging it and trying to follow people's advice.

We were going to go to the Lacey Alternative Energy Fair, but they charge money this year! Man, that's messed up. $25 for a bike? Maybe I'll ride my electric bike there, or the electric car (come on, controller version 1a!) and just look at all the alternative energy vehicles. hehe. Wow... An "electric car", I didn't know they had those!

[blowncurtice]

I have just blown my curtis 1231C so am following this with interest

[Roger Heuckeroth]

Paul,

What model LEM sensor are you using? I thought their output range was 0-5V, not 0-0.6V. What does the output look like on the scope?

Roger

[wjdennis]

Three things, only one of which is related to Paul's controller, but I hope y'all don't mind,

1) Paul, what words do the P and I in "PI loop" stand for? Thanks.

2) bennelson, the link on your web site to your build thread has some invalid characters at the end, so doesn't work (I got there by viewing the source of your page and finding the real address)

3) I'm new to soldering and I can't find go info on how to do the following: I've got a DC/DC converter with a pins that are a little over 1/4 inch tall, and .080 inches in diameter. I need to attach a #8 wire to each pin. How do I go about soldering wire to pin?

Thanks all.

Bill Dennis

[MPaulHolmes]

Roger: the output of the LEM is 0-5v, but that covers the range -900 to +900 amps. I think it's intended for use with AC stuff too. 2.5v is considered 0 Amps. Then for the 500 amp version, 2.5v + 0.625v corresponds to 500 amps. They want to leave extra room for higher currents that may be encountered I guess. There isn't one that I have found that is 0 amps at 0v and say 900 amps at 5v (for safety). That would be too obvious! Tamura has one that has an output of -5v to 5v, but is powered by a 15v and -15v supply. It would be nice to have a 5v or 12v supply line to keep things simple.

It's a LEM HASS 500-s, from digikey.
http://www.lemusa.com/docs/products/hass_e.pdf

I tried to video tape the oscilloscope output, but it didn't really show up. Maybe I can get some pictures. I need to go outside and figure out the value of the frequencies that are showing up that look like noise to me.

PI loop means proportional gain and Integral gain loop. You don't really need the D (derivative) in this context. Wikipedia has a good shower hot water analogy about it.
PID controller - Wikipedia, the free encyclopedia

For soldering the pins to the wire, you might want to try to attach some sort of MOLEX connector to the wires, and then plug it in to the pins. That would be my guess.

[bennelson]

Thanks for letting me know, Bill!

I fixed the link.

I will also do some posting on my site once the controller comes out my way.


Also, what are our thoughts on a pre-charge resistor for the main contactor once the car and controller are running 144 volts? Should I just double the ohms of Curtis' recommendation for 72V?

[MPaulHolmes]

One thing to note is that this controller has more capacitance than Curtis uses. They use 35 220 uF if I'm not mistaken, and this has 24 440 uF, so it's like a capacitance of "48" in comparison to Curtis' capacitance of "35".

Some people use the same lightbulb for their high voltage as for low voltage. I have a 40 watt 2.5 Ohm resistor (4 10 watt resistors in parallel),

Let's say we want a pre-charge time of 3 seconds, so that it will be charged to like 99% full (say 5 time constants)

5*R*C = 3 seconds
5*R*0.01056 = 3 seconds
R = 57 Ohms.

So, maybe something around 57 Ohms?

[bennelson]

On my Curtis 1209-64xx controller, Curtis recommends a 600 ohm 10 watt resistor.

I looked up the specs for the 144V 500 amp controller (1231-86xx) and that calls for a 750 ohm 25 watt resistor.

57 ohms sounds really low by comparason.

Are the controllers really that different?

Does anyone else have a thought on the importance of really "nailing" the right resistor?

I just don't want to fry Lil Miss Mosfet/The Cougar/Oprah.

[MPaulHolmes]

I posted the question to the EVTech list. I should hear back shortly! ya! man I love that place.

I am almost done with the box. oh ya.