Paul & Sabrina's cheap DIY 144v motor controller

[MPaulHolmes]

I almost forgot. The sequence of events that keeps things safe is

1. Turn on 12v supply to controller
2. Pre-charge resistor.
3. Close main contactor.

To turn off, I do

1. Open main contactor,
2. Turn off 12v supply to controller.

It probably doesn't matter anymore, since I think I took care of the problem, but I'm only 99% sure. I fried the mosfets a long time ago by closing the contactor before turning on the controller. I highly doubt it could happen now, but you know.

[bennelson]

I got the controller physically in the car - not actually bolted down yet.

I hooked up the power cabling, including brushing the copper bus bars nice and shiney, and using matching 1/2 inch bolts.

I like that the 12V and go pedal connections use insulated spade connectors. I hate that they are a different (read: incompatible) brand from MY insulated spade connectors. I had to make mini connectors from scratch to be able to hook them up.

12v to the controller will be controlled from the original key-switched 12V to the ignition coil. 12V negative is just a ground to the car body.

Potentiometer wires are hooked up to the signal in wires on the controller with just some short jumpers.

I still need to get the controller into a position where I can bolt it down AND close the hood.

I also have to build a "rip-cord" going to a giant manual battery disconnect in the back seat.

[Blue Bomber Man]

You also have a rip cord for the parachute brake, right?

[MazdaMatt]

Paul... re: throttle. 0ohms is 0 throttle... 5kohms is 100% throttle... what happens if the connection breaks loose and there is infinite ohms? Infinite throttle?

[wjdennis]

I think he has a 4.7K pull-down resistor on the line. So if the throttle breaks or comes unplugged from the PCB, the line gets pulled to 0 volts.

Bill

[MazdaMatt]

I wish that there was a way to add a thread "header" so we could keep links together on every page. I'm going to skim for the latest schematic and have a look.


EDIT:

I just looked at the schematic 1 page back and I see the following...
as resistance goes to zero (no throttle), the voltage at the pin goes higher. As resistance goes to 5k (full throttle), the voltage at the pin goes lower. So if the throttle is disconnected, the pull-down drops it to zero volts and "lower" is associated with high throttle. Paul, you may wanna check that out. Ben, you may wanna extra-double-tripple-check your throttle connections.

[bennelson]

I know that the Curtis controllers use a 0-5kohm input, and that if you go HIGHER than 5K, it just stops working all together.

It actually doesn't start doing anything at all until up a bit from zero and maxes out somewhere around 4700 ohms.

I think this is both for safety and to account for potentiometers which are slightly out of spec.

I suppose I could always put the car in neutral, turn it on, then pull out the throttle wire and see if anything happens?



The controller spins the motor! Yea!

This controller "feels" different than the Curtis.

The Curtis has a really nice smooth slow start when you accelerate very gently. The open source controller has a bit of a shutter if you try to do a really slow start.

I believe that's the current-limiting feature?
Makes me think that a switch to turn that on or off might be pretty cool.
I guess it just depends how much you like doing things manually instead of having a computer do it for you. In a car, I like manual.

Photos for you!

Here is the open source controller next to the Curtis controller. You can see that the open source is somewhat larger. It has a lot of space inside the box and it has a heat sink on the bottom as well.


I removed the Curtis, and made some room for the open source.


Here, you can see the controller, with the M- cable on the left and the B+ and B- cables on the right. Do NOT stick your hand between the B+ and B- cable when forgetting to disconnect the main battery shut-off! Ouch! 144V bites harder than 72V!


The 12v wires and throttle wires use insulated spade connectors. I also have insulated spade connections on my PB-6 throttle, but they are exactly slightly different enough that I had to make some short custom jumpers to plug in the throttle and 12V power.


In case of emergency....Pull this rope.
I screwed down one half of the battery disconnect to the back seat battery box lid, and tied a rope to the other half. Give it a yank! No more power!


I now have a 144V battery pack, but only a 72v charger! One way to solve this is to charge half the total pack at a time. I got another 50 amp Anderson plug, which matches what's built into the charger. I wired that plug to one half of the battery pack, and the original connection to the other.

[MPaulHolmes]

If the throttle reads too high of a resistance (like if it comes loose, becoming infinite resistance!), the software checks it 16 times, and takes the average of those values (to make sure it wasn't just a single spurious TOO HIGH read), and if the average resistance is above that TOO HIGH value (I can't remember how high, but let's say 10k or something), then the micro-controller kills the throttle, and shuts down the controller. You have to then turn off the 12v power and turn it on again. That gives you a chance to go buy a new dang throttle! hahaha!

Here's how the throttle control works:

The first 10% or so of throttle will give no response. It's "the dead zone"! Duh Duh Duhhhhhhh...

Full throttle is around 4400 or 4500 Ohms. Above that, the throttle just ignores the extra ohms, and treats it as if it was exactly 4400 Ohms (or whatever the heck the specific number was).

But it doesn't ignore it if the ohms get too high! Then it makes you pull off to the side of the road, because it kills the throttle (as described above).

[MPaulHolmes]

I think the shudder is probably not the hardware current limiting (unless you were pulling 500 amps). I'm guessing that it is the noisy signal from the current sensor. I'm partially using the current sensor to control throttle, so you are probably feeling the variation in the reads in the input to the A/D converter. I have gotten advice on how to clean that signal up, so it shouldn't be a problem on version 2.

I wish I would have tried this test too, to see if my smaller motor behaves the same. I only drove the car, never spinning the motor in neutral, and I didn't notice a shudder while driving. It will be interesting to see if you notice the shudder while driving or while taking off from a stop.

EDIT: Actually, it's not just noise from the current sensor. The problem is that the motor current isn't constant. It goes up and down, but I'm not reading the current at the same part of the PWM waveform each time, so the current sensor is spitting out values that reflect different parts of the motor current variation. I'll also fix that in version 2.

[bennelson]

Paul, what you just said about how the ohms works on the open source controller is exactly how it works on the Curtis.

Nice and simple and safe.


I am currently test piloting the vehicle at 144V. I can't give away too many details yet. I am hoping to make a video on Saturday. I'm pretty busy with work this week, so I won't get to it until at least then.

One thing I can say. Acceleration in second gear is pretty nice.
I will also need to install some sort of "anti-gravel" right behind my front tires.

I believe that gel cell batteries are not designed for high amperage use, they may end up being the limiting factor in my system.

Also, I may need to move where my main contactor is, so that I am able to split the pack in two for charging.