Paul & Sabrina's cheap DIY 144v motor controller

[twotires]

I've made a dumb move. I'm building a controller for a smaller application - a VW Beetle running at 96V with a 7.5" salvage forklift DC motor. I substituted 150V Mosfets and dropped down on the power caps' size to 470 uf. I messed up and ordered 30 mm diameter caps, though (only cost $20 instead of $60 for the 820 uf's !) so all sixteen won't fit. I can stuff thirteen of them on the power board by changing the hole pattern.

My question: Do you think (13) 470ufs will be OK - they should fit OK like the pic below - or should I spring an additional $60 and get the 25mm dia 820 ufs?

Thanks,
Michael

[adamj12b]

Originally Posted by twotires

My question: Do you think (13) 470ufs will be OK - they should fit OK like the pic below - or should I spring an additional $60 and get the 25mm dia 820 ufs?

Are you reducing the current as well? Or do you still want 500A and 96V?

If you still want the 500A, you will need the correct caps. They are designed to filter the current ripple which will not differ with voltage.

-Adam

[MPaulHolmes]

I recently read about the Kelly 600amp controller. Evidently, they use ten 330uF caps. Take it for whatever it's worth. 13 of the 470uF caps might be fine? The only reason I originally used 16 of the 820uF originally, and then 16 of the 1000uF was because it was the most I could fit in the space. I figured if that didn't work, then I was out of luck, since there would be no room.

[isaac_alaska]

i realize that the power board would have to be re-designed and the layout would have to change to incorporate the higher voltages. would any changes have to be made to the control board to handle the higher voltage on the power board?

the reason to go with a higher voltage battery pack would be to reduce current for the same amount of power. right now we are running 200 volt strings, and have 9 of them in parallel. i'd rather go with 400V strings, and only run 4 in parallel. this would reduce the parasitic losses in the battery pack and i would think it would reduce losses in the controller by keeping the current smaller, but maybe i'm wrong.

re-designing the power board wouldn't be a huge issue for me as i have access to a circuit board prototyping mill and a full machine shop for milling bus bars and things like that. not that it would be an overnight project, but i think it would be possible. Aside from re-designing the power board, what other drawbacks can you forsee from going with a higher voltage design?

i guess it's also worth mentioning that since it's a snowmachine, it has a heat exchanger built into it, i was considering adding a small 12v pump and making the controller liquid cooled. also, physical size isn't really an issue, as i have about 2 cubic feet to fit it into. it's hard to imagine needing any more room than that. (1 ft, by 1 ft, by 2 ft)

[twotires]

Originally Posted by adamj12b

Are you reducing the current as well? Or do you still want 500A and 96V?

If you still want the 500A, you will need the correct caps. They are designed to filter the current ripple which will not differ with voltage.

-Adam

The current will also be much less than 500A. The current will be what the current will be .... that is, I have no idea exactly how much current this motor will consume at 96V max, but it is a tiny-ish motor compared to what most folks are using. It's 7.5" in diameter and weighs maybe 75 lbs. (est.) - I lifted it in place by hand. It's designed to run at 48V, I've been told, so I advanced the brushes 10 degrees.

Current being a function of voltage and resistance, and since I'm running a small motor (higher resistance to the flow of current) and low-ish voltage, then I'm making an assumption that the current will be somewhere in the 300A maximum range. I've not been able to find any data on the motor - I do know it's from a 1977 Yale lift truck. I got the SCR controller with it (the one the lift truck used for proportional control) - I'd love to get that up and running and use it in the car - in the words of one of my blog readers, it's very "steampunk". This Open Revolt controller is going to be way better, though.

[adamj12b]

Originally Posted by twotires

The current will also be much less than 500A. The current will be what the current will be .... that is, I have no idea exactly how much current this motor will consume at 96V max, but it is a tiny-ish motor compared to what most folks are using. It's 7.5" in diameter and weighs maybe 75 lbs. (est.) - I lifted it in place by hand. It's designed to run at 48V, I've been told, so I advanced the brushes 10 degrees.

Current being a function of voltage and resistance, and since I'm running a small motor (higher resistance to the flow of current) and low-ish voltage, then I'm making an assumption that the current will be somewhere in the 300A maximum range. I've not been able to find any data on the motor - I do know it's from a 1977 Yale lift truck. I got the SCR controller with it (the one the lift truck used for proportional control) - I'd love to get that up and running and use it in the car - in the words of one of my blog readers, it's very "steampunk". This Open Revolt controller is going to be way better, though.

Dont be surprised if you motor pulls more then 300A. My golf cart uses a 6.7" motor and I can draw almost 200A continuous. With peaks as high as 450A. This is in a vehicle that is only 1000lbs. In the car, it will most likely be twice the weight.

Also remember, the controller is current regulated, you can have very large amounts of current at low duty cycle because voltage is low.

Have you measured the resistance of the motor yet?

-Adam

[jyanof]

Originally Posted by twotires

I've made a dumb move. ...

My question: Do you think (13) 470ufs will be OK - they should fit OK like the pic below - or should I spring an additional $60 and get the 25mm dia 820 ufs?

Thanks,
Michael

You could always build it and measure the temperature of the caps while in operation. If they get really hot, like anywhere near their rating, they probably won't last long.

I think the originals are rated at 105C and I measured a typical operating temp of 65C (about 20C above ambient) while cruising at about 200A, for reference (that's off of memory though, the numbers are somewhere in a previous post).

BTW - I'd think that short bursts of high currents would be ok on the caps. Most controllers don't have enough caps to handle their max current, but enough for the continuous current.

[DJBecker]

The capacitors also have to do work. If they don't get warm at all, they probably aren't doing much good.

We've experimented with big surplus caps and found that they are mostly ineffective. They are optimized for filtering 120Hz ripple. At 15KHz they just slowly absorb a some of the energy and ring -- the ESR and inductance makes them a tank circuit. It only takes two smaller 330uF caps designed for switching power supplies to match the peak clipping of the 4,900uF and 18,000uF monsters we've tried, with much less (and higher frequency) ringing.

That said, those two caps work very hard. They get warm with just a minute or two of operation. You need more caps to spread the energy absorption. And with more caps, there is a lower combined ESR, therefore much less total heat, and much less ringing.

It does take a while to internalize that it's not about bulk capacitance. You do need enough capacitance to absorb the inductive spike, but it's really about how quickly you can push the energy into the cap.

Another concern with giant caps is the heat build-up. Besides having less surface area for their volume, the heat has much further to travel to get to the surface.

[MPaulHolmes]

OK, I got one 1000amp controller assembled with a precharge resistor and relay onboard. So far I tested that the pwm signal makes it to the output of the hcpl-4506 optocoupler... PHEW! That means basically everything works like it should. Next I'll calibrate the voltage monitoring. The initial testing phase just about gives me a heart attack. I have no plan B. hahaha.

I also heard the precharge relay click closed. After "x" seconds, the main contactor closes, and then opens the precharge relay. No more emails about "what is with this precharge resistor?"! I'll just say, it's not important. It's allllll been taken care of. hahaha.

[isaac_alaska]

i'm extremely interested in your new 1000 amp controller. any ideas on when the schematics and parts lists will be available? also, is the voltage still limited to 144 volts? i'd like the option to be able to go with a higher voltage if we decided to go that route, but if it's only a matter of re-designing the power board parts and layout, i could probably handle that.