Paul & Sabrina's cheap DIY 144v motor controller

[jyanof]

Originally Posted by MPaulHolmes

The lights flash rapidly only when throttle or current is out of the expected range.

What happens when the current sensor is not connected? I know we disabled output, but I'm not sure what happens to the LED. Does it just give the 'current-out-of-range' error and blink rapidly?

[MPaulHolmes]

It should just blink fast again. It would read less than 2v on A/D input. A single LEM output below 2v causes the controller to lock up, and make the light blink.

[MPaulHolmes]

OK, I've tested all the main circuits for the SR controller now, since there are 2 of them for SR, and 6 for AC, and the AC version works (so the SR version will work). The only thing for the SR version to really test is multiple gate drivers. I think it will be fine.

[The Toecutter]

We tried trouble shooting it some more.

You specified the components that should be checked over the phone with the owner of this board. We checked them. The current sensor read 3.03 volts and you said it should read 3.33V.

One curious item stood out; we gave resistor R4 a voltage measurement and it read 0V. This does not sound like it should be reading this. It was 4.7k ohms as specified, but it has no voltage across it, and therefore no current going through it. I highly doubt it is supposed to operate this way.

We are thinking there is a short somewhere in the board itself; all of the soldering points are good and there are no bridges anywhere.

What do you think it could be?

Well, it turns out that I got a call from the owner of the board as I was typing this. A pulldown resistor that was specified for this board was too high of a value; he replaced it with a lower value and now the controller works.

We will see what happens when he gets his WarP 9" motor back from being rebuilt(failed under warranty) and back in the truck it came from, to let you know how the product works.

If it works as advertised, I will be buying one myself, for a 216V/500A setup in a Triumph GT6 with a race-modified Prestolite. Maybe your controller will be suitable for flying around at 120+.

[MPaulHolmes]

That has ALMOST zero volts across it. It's very unlikely that your volt meter can measure the voltage drop with such a tiny current going through it. It's part of a low pass RC (R = 4.7kOhm. C = 0.1uF) filter to keep the throttle signal clean.

The current sensor A/D input also has a resistor that has almost zero volts across it. (R6) It's low pass filter is R = 4.7kOhm, C = 0.022uF.

Did you mean the throttle should read at least 3.33v? It sounds like it's not a
0-5k pot, or the board was populated with at least one incorrect resistor value.

Which pulldown resistor is too large?

[The Toecutter]

Yes, I meant to say throttle in place of current sensor.

All of the resistors gave the values demanded from the online schematic when measured(+/- 2%). None of them were faulty.

I'll have to get specifics on which resistor was changed. I was not there when he took it out and replaced it, but with it gone, the error light no longer blinks on and off.

Just how low is the voltage across those resistors supposed to be? The multimeter wouldn't register any voltage at all even on its lowest setting.

On a side note, When I built filters during school, the voltages across the resistive components used were never lower than the mV range, and the passband calculated was oddly enough way off from the measured results, even though the calculations were done correctly. I suppose designing working filters is at least as much trial and error as calculation. The difference between measured and calculated results was always frustrating and made the task of designing filters seem a waste of time, despite the fact that they are ubiquitous in today's electronics.

I can only imagine the effects the stray high frequency signals might have on the throttle, that being said...



Attached are pictures of the board.

While the flash of the camera destroyed the view, when held up to direct light, the rings of the solder points are clearly visible.


There may be a possibility it doesn't like the pot. I believe it is a 0-5.6 kOhm pot, but I don't see why that would make a difference if "no throttle" input is 0 ohms.

It's an interesting problem. Almost everyone else has been able to get this to work except for us using this particular board. Do you have a list of which voltages are supposed to be read along each component? That would certainly aid in trouble shooting it.

[MPaulHolmes]

I think you are right. It's not the pot. 3.33v was the absolute bare minimum that it had to be for 0 throttle. Assuming the resistors were absolutely perfect, and assuming that you just shorted across the throttle to guarantee 0 Ohms, the voltage at the microcontroller would have been:
4.7k/(2k+4.7k) * 5v = 3.51v.

I said 3.33v since there's built-in dead throttle. Anything in the 3.33v to 3.51v does nothing. Now, to get 3.03v, assuming R10 was 4.7k:

4.7k/(R8 + 4.7k) * 5v = 3.03v
So that R8 = 3.06k.

So, you may have used a 2% tolerance 3k resistor instead of a 2k resistor for R8.

I just noticed from the picture that R6 is a 4.7k 5% tolerance resistor from the color code. I don't see R8 anywhere though. Can you include a picture of that one too?

[snakub]

Hello I recently put together a control board and talked to you (Paul) over the phone about the yellow error light flashing. I measured the voltage at pin 25 to pin 22 of the preprogrammed atmel which was what you told me it would be 2.5 volts. Then I measured the voltage at pin 23 to pin 22 of the chip and it was a little lower than expected 3.03 volts rather than 3.33 volts. So I changed out R10 to a 10K and R4 to a 4K and still got the blinking light then I put a 6.74K ohm resistor in parallel with the 10K in R10 and now the yellow light just stays on. Is that bad? The voltage at the chip is now 3.35 volts mabye a little more. Now when I turn the pot I am able to turn the motor but the throttle is really abrupt it goes right to about 8.5 volt to the motor from a 12 volt battery. I had all the resistors of the right values in the first place I don't understand why this is happening? For R10 I had a 4.7K and R4 I had a 4.7K.

[MPaulHolmes]

Was R8 3k by mistake? (It should be 2k) I'm willing to bet $1,000,000,000 that it was. hehe.

[MPaulHolmes]

Throttle is proportional to current, so if you give it 50% throttle, the controller will do everything it can to get 50% of max current. Max current is 500 amps, so it will attempt to get 250 amps from that poor little 12v battery. It will fail in that attempt, so it will run the PWM duty all the way to 100%. The software is designed for controlling higher voltages and heavy vehicles. 50% of full throttle when you are just taking off feels nice and gentle. But it's still 250 motor amps. It just doesn't seem very ideal when driving a little 12v motor in a garage.