Sunday, February 18, 2007

Battery Equalizer Schematics Rev 1.0



I'm currently working on a battery equalizer for the 914 EV based on the whitepaper at the following link:

http://www.smartsparkenergy.com/pdf/batteq1.pdf

The above two schematics show the circuit that I'm experimenting with using 12V batteries. It seems to work fine on my bench. The circuit above costs less than $10 per battery in parts and I hope to make this available to everyone to help people make their batteries last longer.

By clicking on the images, you can get a much larger version of the pictures. Any feedback on how to improve the circuit would be welcome. I'll have to modify the circuit with a voltage doubler in the final stage to work with 8-volt golf-cart batteries for the 914 EV kit. Big thanks to Max at and all the other supportive engineers at HP for designing this circuit.


ADDENDUM: recently, many people have been asking me about this circuit and how well it's working. After some experiments awhile ago, I found that the capacitors simply do not transfer enough current to make this circuit viable, especially for the large amp-hour batteries needed for an electric vehicle. I currently drive an electric Civic and use resistive shunt balancers on it that seem to work well. The large 3-ohm shunt resistors are mounted away from the batteries on an aluminum base as a heatsink since they dissipate quite a bit of heat. The shunt balancers were purchased from Belktronix, but you could probably make them yourself with a low-current comparator that drives a FET to switch in the power resistor at a specified voltage (like 14.4V).

7 comments:

Unknown said...

In a blatant attempt to steal your traffic I improved on your circuit and posted it on my blog :)

Seriously though, I potentially found a way to eliminate the gate drivers from the circuit making it even cheaper. I ran some quick spice simulations on it and it seems to work.

Have a look at it here: www.electric-lemon.com

Nice blog by the way, keep up the good work.

-Peter

TimK said...

Hi Peter. Thanks for your comment. It looks like electric-lemon.com is down for the moment. I'll check back in a bit to see if the site is back up. I'd like to see your optimization! Regards, Tim.

sparc5 said...

Hi Tim,

What if you used a very efficient DC-DC converter to bump the capacitor to the battery's float voltage? That way you'll be able to transfer current faster between batteries especially needed when they are <500mv apart.

TimK said...

Hi Tschaff,

That would work, but if you had a DC-DC converter for each battery that would be far more expensive than just a couple FETs. If you have one small isolated DCDC converter for each battery, you might as well just have Soneil individual battery chargers for the system. Since I'm running flooded lead-acid batteries now, they just boil over to equalize and they're fine. Cheers, Tim

RickT said...

Has anyone built this circuit yet? If so does anyone have the layout file so I could etch a board for it? Using a breadboard would be really hard with the number of wires in this circuit.

ronv said...

I am getting ready to build this circuit and have made a few changes that I would appreciate input on. Added 12 volt supply developed from either the 2nd or third battery in the string (depending on 12v or 8v batteries).I returned the anode of the boost diodes to the 12 volt supply instad of the next battery in the string. I think this gives me 12 volts of gate drive at each stage. Used a flipflop and 2 and gates to develop out of phase and non overlapping clocks. This allows use of only 2117 ic's. Added an on off switch. Added a resistor from each battery to precharge the caps to reduce the high current in the first turn on before the caps are charged.
I have schematics and layout but i'm not sure how to attach. e-mail me for them.
Ron V.

ronv said...

Forgot to leave my e-mail from the last comment. Ronv@wbhsi.net
With the parts I ended up with i anticipate 10 amps peak and 1 amp avg. with a 1 volt difference. Seems like this should work as good or better than the 3 ohm shunt that only works during the last part of the charge cycle.