Paul & Sabrina's cheap DIY 144v motor controller

[MPaulHolmes]

I ran into a similar problem with programming the AC controller. It turned out that the colorful ribbon cable wasn't allowing it to be programmed correctly. Too dang long. Once I switched to a 6 wire shielded cable, everything was OK. It did work the few times I tried it with yours, but maybe it's one of those hit-and-miss things. Sometimes works and sometimes doesn't. I actually just bought some shielded 6 wire cable. I could mail you some. I don't have any 5 or 6 pin connectors though. I'd need to order that. In the meantime, it's extremely annoying, but you could take the enclosure off, and plug the pickit 3 right in to the 6 prongs with no cable. That's much more reliable because there's no cable involved.

I'll still looking for what could be causing the problem. I haven't found anything yet. I'm going to send you a simplified hex file, which doesn't use the EEProm at all. Just your programmed settings hardwired in. You can try programming that onto the chip using the ribbon cable. Worst case, you may have to remove the stupid top and plug the pickit 3 right in. dang it!!!

edit: I just sent a new hex file that skips reading from the EEProm, and just loads a default value in. BEfore, it would read from the EEProm, and if there wasn't a valid setting, it would load a default. But who knows. Maybe there was some left over valid programmed values that were messed up. So, let's avoid that difficulty at the moment, and just load defaults that I know are good. I hope this works!!

[duncan]

Hi Paul
I think I'm being a numpty
Can you send me a list of the steps I need to take in updating the PiCkit

I think its
Set device
Import your program
- it says successful

connect to device
- it asks about 5v and then target detected device ID revision = 1004

Then (and I didn't do this before) hit the "program" button
it says - programming complete
Pass Count 1
I haven't tested it in this state except to try the RTD - which operates, communicates
But does not seem to find anything

If I then press the "Verify" button
I get
Verify Failed

Can you lead me through what I should be doing - please

[MPaulHolmes]

1. Set device
2. Import your program
3. Programmer->Select Programmer->6 PICkit 3
4. Turn on the 12v power, which does the following:
connect to device
- it asks about 5v and then target detected device ID revision = 1004

5. Programmer->Erase Flash Device
6. Programmer->Program
7. Programmer->Verify

Try that! Maybe it needed to be erased first?

[duncan]

Hi Paul
Not done the erase thing yet
Put your latest program in
Now the car could climb the slope!
Felt really lively - I was careful with the throttle until I got down the hill and onto tarmac
Then I throttled
Quick surge and then it cut out - no response
I shut it all the way down
Back on - power again
But as soon as I hit the throttle - shutdown

Being New Zealand in the spring it was now hailing!

It drove back up the hill - a really steep gravel hill - with me being careful with the throttle
Not sure how much current but I saw 100 battery amps at one stage

Feels like some sort of over-current shutdown

[MPaulHolmes]

Crap, I wonder what I had the hardware overcurrent set to. There's a little cube shaped pot on the control board. I can't remember how far I have it twisted clockwise. Fully Clockwise should cause a hardware overcurrent at 600amp per IGBT. It's on the left side of the board, and is blue. It needs a small phillips screw driver head, or a fingernail?

I would download realterm. I don't think this controller is compatible with RTD Explorer. Get that here:

RealTerm: Serial/TCP Terminal | SourceForge.net

Then, set it to 8 data bits, No parity bit, and 1 stop bit. Also, 19.2kbps. It will display the text from the controller's serial output. Then you can see what fault is happening. It probably is the hardware overcurrent. If the small blue pot is already twisted fully clockwise, then we probably need to tune the PI loop. It is likely overshooting right now if the blue pot is fully clockwise already.

What voltage are you running?

We can also clear the hardware overcurrent at like 1kHz or something, but it shouldn't be tripping at all. Fully clockwise is 1800 motor amps for hardware overcurrent.

[duncan]

Hi Paul
I turned the wee blue cube thing clockwise

It's magic!
The tires slip on take off - then grip
I think I could even use the 1200amps -
But that's for later,
I'm hitting up to about 800 battery amps now!

Great!

There is a motorsport event in Christchurch on the 30th November
Evolocity » eMotorsport
I'm entered - now I'm good to go!

Thanks for all of the help - now lets go and see if I can keep up with a Tesla

I'm using Headways - 16Ah (nominally but they are actually a good bit less)
With four strings of 40 cells
So about 130v
They are 10C continous - 640 amps
and 15C burst - 960 amps
But they do saaaag

[MPaulHolmes]

Yahoo!!!!!!!

I can send you 1200amp code tonight if you want. I will do that. And you can use it if you want.

[duncan]

Hi Paul
Been driving the Device - its great fun - feels very lively without trying to kill me

I notice a hiss/whistle - I think its from the controller although it "sounds" like its coming from behind me
Not at all a nuisance in fact it's useful as it seems to be proportional to current

My speed signal to the Cycle Analyst seems to have gone again - I will have to fix it after the weekend sprints
I will try and mount my GPS somewhere - just in case I overspeed the motor

Anyway
Thanks again for a real fun controller - I will keep you guys updated about how it is doing

[MPaulHolmes]

hello! That's great to hear. The hissing is the spread spectrum switching so there won't be a piercing 8khz tone. Will the races be filmed? I'm curious to see you in action!

[duncan]

Paul
What frequency did the old 500amp controller use?

I will see if I can get a film of me missing the apexes and generally doing a terrible lap
Not deliberately - I'm just not that good a driver

The guy I used to mechanic for could go much faster than me - even in my car!