Is this feasible at all?

My idea is to have a battery of 112 Li Ion cells in series, ending up with a nominal voltage of about 414V. This is apparently how car batteries are configured as well and it’s the easiest for an EGS002 inverter.

Charging however is quite difficult as stepping up to those voltages is not trivial.

The idea is to have a charge pump circuit on every single cell. 2 mosfets for charging, a capacitor, 2 mosfets for discharging. The capacitor would charge from a 4.2V supply and discharge into the battery.

The mosfets on the battery side can be controlled with optocouplers. The gate voltage is provided by cells further up in the chain so the gate always has at least 12V on it.

Is this a bad idea? Things that come to mind are the mosfet losses. Essentially the on resistance counts 4 times per cell but usually the rDS on is in the 20 mOhm range, so 80 mOhm per cell.

Balancing is much easier this way.

Perhaps a compromise that charges several cells in series and watches the balancing could work too if efficiency is the biggest reason against this.

I’m aware of the safety considerations.

  • Rolivers@discuss.tchncs.deOP
    link
    fedilink
    English
    arrow-up
    1
    ·
    11 days ago

    The idea of the battery pack is to provide energy for a household. The reason I’m considering such a large amount in series is that it would make an inverter much more efficient and easy to make. An EGS002 based circuit for example relies on a 400V DC, which I find very difficult to make out of lower voltages. Especially when high wattages are involved. There are no DC-DC converters commercially available with around 55V input (16s li-ion) and 400V output at 4KW. There are however commercial inverters available such as the Victron Multiplus II 48V that can deliver about 4000W of AC. This would suit my needs however I want to DIY as much as possible. I also have plans for a self made solar charge controller and possibly even a wind controller as extra modules next to the battery charging circuit and inverter.

    Solving the issue of converting high power 55V to 400V DC would remove the need for a high-series battery pack as well.

    Indeed blindly dumping the charge into a battery is not the intention. There will be additional circuitry to prevent cells from overheating and to limit the inrush current that they receive but the general idea is a capacitor that floats between 4.2V and Vcell.