Paul & Sabrina's cheap DIY 144v motor controller

[jcp123]

I am so lost here.

[jedsmd]

Originally Posted by MPaulHolmes

Uh oh. You can get a data stream by typing this through the serial port:

rtd-period 100

That will dump a data stream on the screen 10 times per second (100 milliseconds). It will show things like current feedback, throttle, temperature, etc...

I will need to download realterm and figure this out.

Does it start as soon as you start driving?

Yes

How about in neutral?

I don't know I will test tomorrow

Does it have a heat sink?

No I just force air across the bottom. Max temperature of the plate is 45 degC. I can hold my hand on the plate after a long drive. It doesn't seem to be temperature related anyway.

I have rigged up another 5k pot to substitute for the throttle pot. Friends came over before I could test it and I shouldn't drive it tonight. So I will test in the morning.

[MPaulHolmes]

I have a theory... if the throttle ramp rate is sufficiently slow (on its max setting) that the controller is still delivering current for an instant after the throttle contactor opens, the big ring capacitor voltage will go to zero. It will start filling up again due to the precharge resistor, but if the throttle closes before the cap is full again, the capacitor would be hit with like 10,000 amps for an instant. If this happens a lot (like in stop and go traffic), maybe the cap could get damaged. Even if that's not the reason for the strange behavior, your code needs another modification, due to the contactor on the throttle. I know what I could add to yours.

[jedsmd]

Originally Posted by MPaulHolmes

I have a theory... if the throttle ramp rate is sufficiently slow (on its max setting) that the controller is still delivering current for an instant after the throttle contactor opens, the big ring capacitor voltage will go to zero. It will start filling up again due to the precharge resistor, but if the throttle closes before the cap is full again, the capacitor would be hit with like 10,000 amps for an instant. If this happens a lot (like in stop and go traffic), maybe the cap could get damaged. Even if that's not the reason for the strange behavior, your code needs another modification, due to the contactor on the throttle. I know what I could add to yours.

So you need to make sure the PWM is completely off before the throttle switch opens?

I substituted a 5k pot I had laying about for the throttle pot. Everything was smooth again. Switched them back and forth a couple times to make sure.

Decided to clean the throttle pot with tuner cleaner but when I got into the potbox I found that the pot was a sealed unit so cleaning didn't happen.

The throttle pot is new Curtis, I bought it for this controller. I wonder if it is borderline noisy when it moves and if that is something that could be filtered in software?

[MPaulHolmes]

***breathes a huge sigh of relief*** Haha. There is actually quite a bit of filtering on the throttle. It averages out about the last 128 samples or so. The ohms must have been jumping around pretty significantly for you to feel it. I'll send the hex file for the fix that is probably not a concern, but could be at least in theory. It would just be a single extra line of code.

You could watch the resistance change on an ohm meter as you move the throttle. My guess is it jumps around quite a bit.

[jedsmd]

Originally Posted by MPaulHolmes

***breathes a huge sigh of relief*** Haha. There is actually quite a bit of filtering on the throttle. It averages out about the last 128 samples or so. The ohms must have been jumping around pretty significantly for you to feel it. I'll send the hex file for the fix that is probably not a concern, but could be at least in theory. It would just be a single extra line of code.

You could watch the resistance change on an ohm meter as you move the throttle. My guess is it jumps around quite a bit.

My DVM isn't fast enough to catch them.

It feels like drop outs. Maybe ramping the power down as we discussed and rejected earlier would mask them.

[MPaulHolmes]

Large resistance would be read as zero. So, if there are places where the wiper loses contact for an instant it would constitute a moment of zero throttle. It would race to zero, only to turn around and head back once contact was made again. That would make for a jerky ride.

[jedsmd]

Originally Posted by MPaulHolmes

Large resistance would be read as zero. So, if there are places where the wiper loses contact for an instant it would constitute a moment of zero throttle. It would race to zero, only to turn around and head back once contact was made again. That would make for a jerky ride.

That's my theory of what is happening. I think the wiper float times are short enough that momentum keeps them from being too jerky.

[MPaulHolmes]

You have to type rtd-period 0. A separate program must take the data and extract the relevant numbers, and graph it.

[Astro]

Originally Posted by MPaulHolmes

Large resistance would be read as zero. So, if there are places where the wiper loses contact for an instant it would constitute a moment of zero throttle. It would race to zero, only to turn around and head back once contact was made again. That would make for a jerky ride.

Paul would that be reversed for throttles that were 0k = no throttle, 5k or higher = full throttle? Would they experience moments of full throttle if the wiper in the pot floated too much?
Would it be possible to hook up one of these hall effect throttles instead to avoid the mechanical wear on such a crucial part.
EV Works Hall Effect Throttle Box 3-wire, 0-5 volt
The hall effect throttles have to have a 5v supply and i believe they output a 0v to 5v signal. Would this be possible?