Paul & Sabrina's cheap DIY 144v motor controller

[xtian999]

Do you know how big of a motor your controller will run? Or, I should say, what size motor would be a good match?

[MPaulHolmes]

Great question! That's where the testing will come in handy. If I had to guess, I would say that it works fine with anything from a small permanent magnet bike motor, to a 6.7" diameter 60 pound Series wound DC motor, to an 11" diameter warP series wound DC motor.

It has been tested on the bike motor, the 6.7" diameter motor, and Ben has started initial testing on his motor, which I don't know the diameter of. I think it's a 8" or 9" diameter. Also, Ian from Zero Emissions Vehicles Australia has used a home-made controller with a similar power section on his 11" diameter direct drive motor.

[bennelson]

My motor is a 10" 48V Nissan forklift drive motor.

I drove the car to the bank today. There are three roads that parallel each other out that direction. The middle one is flat, but the other two roads have BIG hills on them, which I have avoided in the Electro-Metro.

Since I and testing the Open Source Controller, I thought I would give it a try.

I went up the one hill to the bank, and up the even bigger one on the way back.

The car did great up the first hill - plenty of power and speed, didn't draw way too many amps.

The other hill is not only big, but LONG as well. To start up it, I was crusing at 35 mph at 100 amps. Going up the hill, my speed started to drop. I increased amps to keep speed constant. I got it up to a little less than 300 amps. Keep in mind that if this was at 72V, that it would have been 600 amps! Which my old controller (nor my batteries!) would not have been able to put out, especially not for very long!

I also tried another burn-out test. THIS TIME, I put on the parking brake! That back tires did still spin though. Not sure if that's because I couldn't pull the parking brake handle hard enough, or because of the 700 pounds of batteries directly over the rear axle!?

I also noticed that the "whine" seems to happen at around 250 amps while accelerating, NOT when maxing out the ammeter.

We need to investigate this "hurkey-jerky" thing the controller does while starting accelerating. It feels like the car doesn't know if it wants to go or not. From the drivers point of view, it's just not a good thing.

Also, anything we can do to "unleash" the amperage with this controller is good. I like on the Curtis that you can use a pot to adjust your throttle response and power output.

Overall, the controller is very nice. I'm mostly comparing it to the Curtis and Alltrax controllers that I have experience with.

http://www.youtube.com/watch?v=0nIIPsVfUm8

The "burn-out" doesn't really look all that impressive. Several people online have already been making fun of the first one. I think part of it is that the video was shot at 24P, which sort of distorts the "time" of it. Also, there is NO ENGINE NOISE! If I dubbed in some really loud engine sound effects, I'm sure it would be much more impressive!

[MetroMPG]

I'd call that an impressive burn out.

[MPaulHolmes]

Originally Posted by bennelson

I also noticed that the "whine" seems to happen at around 250 amps while accelerating, NOT when maxing out the ammeter.

We need to investigate this "hurkey-jerky" thing the controller does while starting accelerating. It feels like the car doesn't know if it wants to go or not. From the drivers point of view, it's just not a good thing.

Also, anything we can do to "unleash" the amperage with this controller is good. I like on the Curtis that you can use a pot to adjust your throttle response and power output.

Overall, the controller is very nice. I'm mostly comparing it to the Curtis and Alltrax controllers that I have experience with.

Are you measuring battery amps or motor amps? The current limiting is happening when motor amps goes over 500 (according to the current sensor), and the battery amps is the MOTOR_AMPS*PWM_DUTY. So, at low duty, you can have low battery amps, but very high motor amps. Actually, that's another reason why we need to get the signal cleaned up out of the current sensor! If there's a 15 or 20% error on the current sensor reads, then the hardware current shutdown might be coming on anywhere in the motor amps range of 400 to 500.

And another reason to clean up that signal!!! I'm using the current to limit throttle. So, if I'm trying to set the current so that it matches the throttle position, and the dang current sensor is spazzing out, +/- 20% error! haha! You are going to feel the "not sure if it wants to go or not" phenomenon. It wouldn't be the same issue at higher RPM, because the motor current would change more slowly, because of the inductance, so I'm thinking it would be more accurate current sensor reads then.

I have gotten some really good suggestions about how to fix it. In fact, I might be able to fix it without any new components added to your controller, Ben. I'm almost done with version 1.000001 (few changes from the previous one, so I can drive to the alternative energy fair in a couple weeks, and use my car once again!), and I'm going to try some software ideas that have been sent my way. I could then send you that chip, and you could take off the top of the controller, and plug in the new one, and see if there is any improvement at the low RPM.

[bennelson]

My ammeter shunt is in series in the string of batteries, so I am measuring battery amps. I always felt that's the number to watch, because the batteries care a lot more about how many amps you pull than the motor does!

Anyone have a spare analog 500 amp ammeter kicking around? Mine only goes to 300!

I think I am starting to get why it only does wierd stuff at low speeds.

Once the car gets going, it's pretty slick in either second or third gear. Both have really nice acceleration and a good speed range.

Clean up the signal to smooth out acceleration and max out the amperage, and THIS IS GOING TO BE ONE NICE CONTROLLER!

The heat sink does get a little warm while in use. My 72V Curtis never got warm, but it only had half as much power going through it!

Also, I haven't hooked up that 12V computer fan to it either. I plan on that this weekend.

Or better yet, just cut a whole in the hood. That way I can show off the controller AND keep it cool!

[MPaulHolmes]

Originally Posted by bennelson

My ammeter shunt is in series in the string of batteries, so I am measuring battery amps. I always felt that's the number to watch, because the batteries care a lot more about how many amps you pull than the motor does!

Anyone have a spare analog 500 amp ammeter kicking around? Mine only goes to 300!

You can get like +/- 5% accurate using the copper cable from the controller M- to the Motor as the shunt. Then, just use a cheapo $2.99 volt meter from Harbor Freight set on the 200mV setting, and convert the measured voltage across that cable into current by doing I = V/R, where R is the resistance of your cable. It will be a tiny resistance. How long is that cable? What gauge cable?

The reason the controller senses motor amps is to protect the mosfets and diodes. It's also a really good idea to measure battery amps, to keep them safe too. That should be a feature we add later.

[bennelson]

Less safety!

MORE POWER!!!!!!!


ARR rrr RRR RRR RRR!

[MPaulHolmes]

Ben, you are a good person to test this, because you...are...
INSANE! HAHAHA!

[nickfarmer]

Great work, can't wait to get one of these on the electric 306.
Should the finished version be ready by July?
Can't wait to burn rubber and not oil!