per cell charge controller

I'm liking this for Arduino: dual core ESP32

comments?

_____

Yes, delays aka "hysteresis" is required, ideally adjustable like the open/close V setpoints.

I'm thinking somehow maximum per-cell current limiting as well, since as cells's relays disconnect from the charge buss, the remaining cells could draw higher amps than what's good for them wrt longevity.

____

Suggestions for OTS devices would be **very** welcome.

Some "hobby chargers" for electric aircraft have some similar features, but haven't seen any that get above 6-8A per cell, many do not allow for adjustable voltage, and all are very limited in cell count, so going to say 48 cells ends up costing thousands.

______

I am looking (ideally, eventually) to eliminate the need for "bulk charging" at any specific bank-voltage. Every charge cycle being a per-cell top-balance charge if you like.

So 12V portable pack, or 96V EV propulsion bank, same charging setup handles it all.

I would like to limit the topic here to the specific per-cell relays + regulator functionality outlined, referencing a collection of N standalone cells, not connected in any xPyS pack configuration.

AFAIK there is no inverter involved in this scenario.

______

The per-cell relays do need to handle the maximum **charging** ampacity, which yes will be pretty high depending on cell Ah capacity, since voltages are so low, in the 3.2-4.0V range.

Maybe start out assuming 8-12A? So, for prototyping purposes, looking to use ideally 20-25Ah cells, but anywhere in the 15-40Ah range.

I mean, we **could** start with baby cylindricals like A123 cells, what 2-3 Ah? Which would put rates at say 800-1500A per cell.

I'm just afraid then components will be very different, and maybe test results won't extrapolate well when moving up to the bigger sizes, maybe cylindrical LFPs have different behaviour responses than pouch/prismatics.

The design must eventually scale up to 180Ah cells, ideally even over 1000Ah.

Or if that's not realistic, one workaround for large Ah cells, is to bulk charge "rough" still wired as a xPyS pack, up to say 95% SoC, depending how well they stay balanced that way, and then do the per-cell "fine finish" per-cell charging for just the last 10-20Ah, to restore close as possible to a well-balanced state.

______

Finally, note that I am **not** looking for "a BMS" as a possible OTS alternative here. I am considering this a voltage-based per-cell **charge regulation** challenge.

To me, the cluster of functionalities encapsulated in "a BMS" should all be oriented toward "last-ditch" redundancy protection of the cells, against failure of the normal "usage-level" controls involved in normal daily cycling.

> There is a fairly common method for limiting high current flow for AC inverter outputs.

Yes that's what I'm talking about. Maybe crude but cheap, if it works and can be varied with cell size as needed, great.

Ther is no bank, no xPyS, just a collection of isolated cells wired to a charge buss.

The cell gets taken offline to let V drop. If V stops dropping below the chosen Full setpoint, that means it is Full, charging has stopped. That is the only intent.

Each cell is isolated, no series, no parallel, each V independent.

Current does not need to be measured at all here, as long as from the beginning, preventing going over a rough ceiling, .5-1C.

LFP prismatics, as stated

> 20-25Ah cells, but anywhere in the 15-40Ah range

for prototyping purposes, design intended to be useful for much larger ones.

> Adding a high value resistor or small fuse per wire is common in sense circuits to limit any short circuit current.

If needed for safety then I'll do it.

But just sensing voltage going on, no ammeter here.

I would **really** prefer, for purpose of this thread, to focus on the HowTo accomplish what I've outlined, if members here have the knowledge & skills to help me without putting in a lot of work, I'd really appreciate it.

Not so much on the "big picture" why's, theory of charging, what my intentions are, etc. Yes this is "unconventional" even experimental, and yes if maybe a waste of my time and money so be it, I will learn a lot and try something else.

Or maybe come back and share some interesting results that somehow extend our knowledge or help other DIY types.

So, again, the purpose is

Achieving top-balancing by low-voltage, independent CC-only charging at the cell level, using voltage-based termination only.

And ideally if possible, using this for every-day charging in normal cycling.

There is no dynamic-while-charging regulation of per-cell currents. Periodic charge disconnection is based strictly on hitting the top voltage setpoint, then re-connecting each cell to the buss based on the low-volt setpoint and hysteresis delays, basically a CV-only delta-peak style algorithm comparing cell- vs the buss circuit-voltage

Remember this is not "a bank" nor even "a pack", just a bunch of individual cells. When the pack is ready to be put back in service, its cells will be returned to whatever xPyS full-bank topology delivers the voltage and Ah capacity required by the intended use case at that time.

But that "stage's" configuration / wiring, details on the overall use case(s) is IMO moot and off-topic here; I think getting into that would divert the focus of the thread.

While the packs are "broken down" like this, the cells are **not** paralleled in the usual sense, or rather, **only** while the "cell-level regulator" is holding that cell's relay open to accept the buss' charging current.

Those relays **may** well get opened (isolating the cells) at a "closely similar" time, but if so, only at the first opening time.

But even then, I suspect not at **exactly** the same time, since the starting-point SoC and SoH (health, resistance, age, residual capacity) of each cell may be quite different from each other.

Furthermore, depending on the programmed number of minutes' delays, and the degree of hysteresis between those top- and bottom-voltage setpoints, each cell will differ in the time it takes during their "resting" period for the relay to subsequently cycle **closed / connected** again, and subsequently between closed and open, until the ultimate goal is reached:

To ensure as close as possible balance of resting voltage (as a proxy for SoC) between all the cells on that charge buss, without using measurement of trailing current, egg-timing the CV/Absorb stage, etc. As stated earlier

> I am looking (ideally, eventually) to eliminate the need for "bulk charging" at any specific bank-voltage. Every charge cycle being a per-cell top-balance charge if you like.

> So 12V portable pack, or 96V EV propulsion bank, same charging setup handles it all.

For any given cycle, if charge time needs to be shortened, exactitude of balancing can be postponed until the next leisurely opportunity.

looks like quoting still does not allow inline editing?

I was trying to guide the thread topic so as to **not** derail, but certainly not my intention to cause anyone to take their ball and go home.

My apologies if I offended.

Obviously I will not be able to conduct these experiments without help.

If I need to pay for that help, so be it I guess.

Suggestions on such resources, other forums, learning materials to help me self-teach etc would be most welcome.