Paul & Sabrina's cheap DIY 144v motor controller

[DJBecker]

Originally Posted by MPaulHolmes

Also, I wrote some gcode, and made 3 etched boards with a couple holes in each one so it will bolt right onto a bus bar. I also soldered a little high magnetic field current sensor to one of them.

With this change in hardware, have you considered changing the current sensor reporting polarity?

We started by using our current sensor in positive output orientation (2.5V->5V with increasing motor current), just as with Cougar.

We recently changed to negative output, 2.5V->0, so that we have the option to use a 3.3V or 3V A/D reference. This allows us to use faster microcontrollers that only run at 3.3V or lower.

Implementing this is easy -- just put the current sensor on backwards. We still use the same 5V supply, since most current sensors won't run on less than 5V.

We did need to make one other minor change. Our circuit has a high value resistor (100Kohm) on the A/D converter input so that we can detect when the current sensor isn't hooked up. (When disconnected, we fall back to proportional PWM mode.) We changed this to bias to ground instead of +V. Although either extreme raises a fault, biasing to higher current protects against a flaky connector.

[MPaulHolmes]

That's a really good idea. My little board is symmetric, with 2 mounting holes for the bus bar, so you can flip it either way, and the current sensor will still be in the same spot above the bus bar. That's the only 5v specific part I think. What a good idea!

I wrote the Gcode for the 2 pcb isolation strips for the 1000amp controller and cut out all 6 of them, 2 per controller. Those little buggers are 11.245"x0.75"x0.08", in case anyone was wondering. I think the metals will be here Monday.

[MPaulHolmes]

I ordered a bunch of control board and power board components today for the 3 prototypes. I have a BOM and stuff that I can post, but the wiki is uneditable now I think.

[MPaulHolmes]

Oh no! Active low signal? just kidding. Yes, that's what the 2C and the new board have. So it's just POIFECT for the IGBT boards. ya!

The cool thing is, just adding a resistor from pin 2 to pin 3 of the HCPL-4506 footprint gives you the pull up you would need, since the dang chip would be gone in the igbt application! ya! Now tell me I wasn't sneaky... go ahead. tell me. hahaha

[MPaulHolmes]

FOUR?! didn't think of that one. hahaha. each one can drive 1400amp igbts. What are you making?

[MPaulHolmes]

Hello! I don't have any driver boards. That's mr bigh, but I think he's been working nights and has been busy. The atmega pin can only do <= 25 mA, and I think that <= 18mA is a better idea. I think there's a way with an op amp or maybe a comparator to get larger current capability of the PWM signal, but that would have to be added. I think it would be pretty simple, but on a separate tiny little board.

[jackbauer]

Bad news lou. You would need to derate 50% to have any hope of the devices surviving. So a 400amp igbt is really good for 200amps or 250 at a stretch. My 500amp controller in the car uses "800amp" igbts.

[MPaulHolmes]

If you are going to use the hex inverting thing anyway, why not just invert the signal, adn then feed that signal into 4 different hex inverters? Or if you are going to add the hex inverters, why not just add a little op amp circuit instead, which can then supply like 200mA, and be following the exact signal as the input?

[sawickm]

Originally Posted by MPaulHolmes

I have a BOM and stuff that I can post, but the wiki is uneditable now I think.

Hi Paul,

Your right, the wiki is uneditable: http://ecomodder.com/forum/showthrea...ews-15987.html

I have an account with onlinefilefolder.com and have your AC & SR Controller documentation parked there. If you want I can upload your new DC Controller documentation there too. The link can be added to the wiki when it get's fixed. Let me know ....

Your new DC Contoller sounds good... but I am still holding out for your dsPIC AC Controller !!!!!

-Mark

[DJBecker]

Instead of a hex inverter or comparator as a driver, you might consider a non-inverting gate driver. Common ones work down to 4.5V, with a few even lower.

They are inexpensive, come in small packages, have input thresholds that work with almost any logic family, and can easily drive almost anything. Easily. Why calculate if an inverter or op-amp will have an acceptable rise time and continuous rating. A $1 TC4422 non-inverting gate driver will put out 9 amps peak and 2 amps continuous.