First Controller Test

[twotires]

Hi all, need some advice.
I completed my controller assembly (500A, 144V version) and am trying to test it. I've got it hooked up on my workbench using (8) AA alkaline batteries to power the board, and a 12V lead-acid lawn mower battery as the traction pack. The motor is a blender motor which pulls 0.8 A running full speed (hooked directly to this traction pack.)

So, nothing happens.

Is this traction pack / motor combo too small to test the controller? I would have expected to push the throttle just a little, and the controller just ramping up very quickly to full speed - trying to force the amps up past 0.8 .... maybe the LEM HASS can't read amps that small?

The motor doesn't even begin to turn.
I can see the PWM signal with my oscilloscope, sort of, and it looks like it quickly goes to full throttle after beginning to turn the throttle just a few degrees off zero.

So, why doesn't the motor spin?
Thanks!

[apowers]

When I made mine I tested with just a 12 volt light. You can check continuity between bat neg and motor neg. Recheck the connections from battery and motor. Try jumping straight to the motor to see if it works.

[twotires]

Thanks, there is no continuity between M- and B-. The motor works when jumped straight to the battery, it draws 0.8A at 12V DC when jumped straight to the battery.

The motor should be connected straight to the battery on the '+' side, and connected to the MOSFET Drains on the other side (M-), correct?

When you tested yours with a light, were you able to use the throttle pot like a dimmer? My understanding of the code that is in mine is the ATMEGA wants to see current develop (a couple hundred amps at least), and if it doesn't it just keeps ramping up the PWM until it does, or it reaches full throttle.

So, if you had a light hooked up, I would think as soon as you started turning the pot, the light would begin burning dimly, then shoot up to full bright pretty quickly. Is that how it would work?

[twotires]

Ok. After a little more investigation:
I've got good solid +5V coming off the regulator so all chips are on and working. I'm getting a solid yellow LED which tells me the brain is OK. The +12VDC coming out of the CINCON DC-DC converter, .... however, is not so good. It's 255kHz garbage. A triangular wave pattern.

Any ideas if that is due to a bad DC-DC converter, or is it probably coming from somewhere else?

Thanks!

[apowers]

I don't remember what the light did exactly. That was 2 years ago. Did you get continuity when you throttled up? The bat pos might need to be connected to controller also. I have no idea about the dc to dc converter.
Alvin

[DJBecker]

Originally Posted by twotires

Ok. After a little more investigation:
I've got good solid +5V coming off the regulator so all chips are on and working. I'm getting a solid yellow LED which tells me the brain is OK. The +12VDC coming out of the CINCON DC-DC converter, .... however, is not so good. It's 255kHz garbage. A triangular wave pattern.

Any ideas if that is due to a bad DC-DC converter, or is it probably coming from somewhere else?

Thanks!

The DC-DC converter will have a high frequency triangle wave of 10mV-100mV superimposed on it's output. If it's higher (it likely is, or you wouldn't have mentioned it), check for a problem in the gate drive circuit that would cause a high power draw.

[twotires]

Yes, it looks like the triangle wave superimposed is 13V p-p .... I think .... I'm still realizing how rusty I am with an oscilloscope.

Approximately how may amps should the control board be using?

The Cincon is putting out the correct +12V ... I was trying to find it with reference to the digital ground, not the power board ground ... the gate signal is not there at all, just the triangle wave stuff.

Thanks so much for your input and ideas.
Michael

[twotires]

OK. Upon further investigation - please, I hope somebody will read this and help me to get on the right track! - the opto-coupler on my controller board (labeled 'U6') ... the input ... measured against the digital ground - looks good. The PWM signal from the ATMEGA pulls pin 3 'low' with each of the peaks (well, from 5V at zero throttle, to 3.5V at full throttle) - so that places a 1.5v potential across the photodiode on the input side and makes it conduct and pass the signal to the output diode.

Here's where the problem is, I think. The output diode, when 'On', should pull pin 2 of U7 to ground, otherwise it should be +12V with respect to the power board's ground, right? It's not. It's 0.7V at zero throttle and 0.29V at full throttle. There is no short to ground at R16 or between pin 6 of U6 and pin 2 of U7.

I think the input to pin 2 of U7 should be basically an 'upside-down' ATMEGA PWM output with a peak-to-peak voltage of 12V, is that right?

I'm thinking right now that U6 is bad and I have to replace that.
Anybody concur? Am I reading this all wrong?

Thank you sincerely,
Michael

[Freyguy]

twotires,
Glad you figured out to measure DC/DC output side reference power ground. I burned myself on that one when I first bench tested my controller.
The +12V draw of the controller should be in the 110mA range (zero throttle).

I think your U7 might be bad versus U6. Reason I say this is that U7 is an inverting MOSFET driver. If you see 0.7V on U6 pin 6 (and therefore U7 pin 2), the output pins of U7 should be +12V and driving those mosfets on. U7 pin 1 measures +12V, correct? I think U7 pin 2 is internally shorted and may be pulling R16 low. Remove U6 then power up the board. U7 pin 2 should measure +12V. If it’s still at 0.7, then U7 is bad and U6 is probably fine.

[twotires]

BRILLIANT! You, Freyguy are a gentleman and a scholar. my U7 is on an IC socket and is therefore much easier to remove. I indeed have +12 on pin 1 and when I remove U7 and power up, I have +12 at pin 2. Furthermore, the signal at pin 2 looks like a PWM signal ... drops to 0V DC more and more as the throttle is applied. So, it looks like U7 is the culprit.

New one is on order.
Got my fingers crossed - I'll update once it is replaced.

Thanks again,
Michael