combined Lipo balancer/charge guard, diy?

[hul]

I was thinking about building a combined charge guard / balancer for a while now, because I'm not quite satisfied with what's available commercially or diy. TP205 balancer with TP1010 charger comes close but I don't want to buy a new charger and I'm not totally convinced of their build quality.

This is what it should do:
- charge guard to stop the charge if any cell voltage is greater than 4.3V
- charge guard to wait for operator interaction before restarting the charge (Polyquest charge guard resumes charging when cell voltage has dropped below the trip point, if the charger doesn't shut down quickly enough)
- charge guard to survive the output of my charger 8A charge current, 65V maximum terminal voltage (Polyquest charge guard's FET is rated for something like 30V only and blows when it tries to stop my charger)
- charging to be through the packs discharge leads, not through the balancing connection to avoid voltage drop at higher charge currents
- balancer to act like Thunder Power's TP205, to equalize cell voltages once they're above 4.0V by using the lowest cell's voltage as a setpoint (this is to avoid the charger frying the balancer if the balancer's setpoint isn't quite right)
- balancer to limit cell voltage to 4.25V independent of the lowest cells voltage
- Multiplex 6 pin balancing connectors for a maximum of 7s packs (7s needs 8 pins total, 6 Multiplex and the 2 discharge leads)
- the same balancing connector is to be used for 2s to 7s packs
- the device must be able to operate with 2s to 7s packs without manually selecting cell count etc
- must indicate what it's doing (tripped on overvoltage, balancing etc)

I believe, I can build the charge guard with relatively simple analog electronics and there are several balancers around (Suzanne's, Dan Baldwin's etc).
But it would probably be better to integrate charge guard and balancer into a common concept.

Any ideas? Comments?

Thanks, Hans

[Dan Baldwin]

I am working on a charge guard/balancer DIY project at present that uses my high current balancers. It meets most of your criteria, but not all. The charge guard is based on the assumption that balancing a normal healthy pack should never require very much current. If the balancing current on any cell get too high (650 MA on prototype), there is a problem of some sort, and it trips out. It uses two sets of 2 amp contacts in parallel on a latching relay, so it would be rated at 4 amp charge current (larger latching relays are probably available). Once it's tripped out, the user must reset it to start the charge again. Since it's relay contacts, it should be able to handle the 65 volts. I have it operating, but I'm still tweaking things trying to get it perfect before I post it.

I had only planned on making it capable of handling up to 4 cells.

Dan

[Zlatko]

Hi Hans and Dan,

Koichi Tanaka made a multi cell charger / balancer but for smaller cells.
His details here http://babelfish.altavista.com/babelfish/trurl_pagecontent?lp=ja_en&trurl=http%3a%2f%2fblog.goo.ne.jp%2ftoko0131%2fe%2 ffc1ac06d54fc0f3a460e984afe7fdaf6

Charges each cell sequentially from 5V supply...

I hope that is of some help.

Cheers
Zlatko

[hul]

your design is discrete / analog, no PIC? Monitoring the balancing current is a smart idea, doesn't subject the cells to high voltage at all.
It should be relatively easy to replace the relay with a suitably rated FET with a flip-flop circuit. I have added a relay to a Polyquest charge guard to make it survive the 65V but the coil current comes out of the pack, which makes the charger not terminate on small packs (because current doesn't decrease below the cut off point). Using a FET would fix that.

The bit I'm struggling with is to make this thing work on any pack without having to change settings for the number of cells.

Any ideas on the balancer using the lowest cell voltage as the setpoint for all cells?
I built your balancer, works really well. But cell voltages vary by roughly 0.03V at the end of a charge. This is because my chargers voltage setpoint seems to drift around a little. To cope with that the balancer setting needs to be a little higher than would be necessary otherwise, now using 4.22V. On my 3s packs the high cell is 4.22V, the low cell usually 4.18V and the last cell 4.2V.

Thanks, Hans

[Dan Baldwin]

The balancer/charge guard design doesn't use a pic or any ICs for that matter, except for the TL431's. I used a latching relay specifically so that it would not draw any current (about 1 MA for the balancers) while charging. The relay contacts also mean that there is basically no voltage drop from the charge guard. I re-checked the specs on the latching relay I'm using, and it's 5 amps per side, so it should be able to handle 10 amps total. The contacts are rated for 250 volts, so 65 volts isn't a problem.

I'm not sure what you could do to make my balancers work well with a charger that drifts in voltage. What kind of charger are you using? How much does it drift?

You might be able to put a small resistor (perhaps .1 or .2 ohms) in series with each balancer, and set the balancers so that they just barely turn on at a low current when the charger is at it's highest voltage.

Dan

[hul]

it's a Schulze 636, it puts out between 12.60 and 12.63 on the 3s Lipo setting. Not sure what makes it drift. Noticed this when it turned on all 3 balancer LEDs, measured voltage and had 4.21V on each cell using a Fluke meter. Pretty sure it's the charger, Fluke and balancers always agree, charger doesn't.

Hans