Lithium Cell Balancing at Nowty Towers - Resurrections

[nowty]

My Pylontech's seem to be behaving well with the new cell in series with it, just as I expected. The Pylontech battery string had a partial discharge overnight and has now charged up again to the new higher voltage and gone into standby.

I have installed a remote battery display in my lounge next to the main battery bank voltage display so I can see if I need to do any manual adjustments to the cell voltage.


I have checked the voltages with a more accurate meter which I have compared with a calibration voltage source. The battery banks are 56.16V, the new cell is 3.31V and the Pylontech cells are (56.16V - 3.31V) / 15 cells = 3.52V.

As I have no active balancer on the new cell, I am kind of happy with it running at 0.2V lower and being a larger cell it should not fall too much at the bottom end. I might add another 0.1V to it but will just monitor it for now.

[nowty]

I’ve been running for over a week now with the extra cell in series with the Pylontech's, but I do have a problem to sort out.

I was running with the extra cell at 3.3V at the top end and that was OK for a couple of days. Then I decided to up the cell voltage to 3.4V but the next day it had risen to 3.5V. I bled off some charge with a 1 ohm power resistor back to 3.4V but the next day it was up there again.

I thought about it a bit more and realised that the Pylontechs obviously use some power to run their internal electronics but my new cell only has a LED voltage indicator draining only about 50mW. Now I don’t know how much the Pylontechs are using but if they are using more than my LED voltage indicator and I'm sure they do, the voltage of the extra cell will inevitably keep rising even if it’s a slightly larger cell.

Obvious really,

I cannot use the newer active cell balancer's on the extra cell because there are no other accessible cells to balance it with. So what I need is an old skool top end balancer which simply discharges excess charge through a power resistor. Had difficulty finding a commercial one and thought I might need to build my own but I have managed to find two on AliExpress.

This one operates at 3.5V and presumably shorts the cell through the 3 resistors in parallel and dumps 2.3A.



This one operates at 3.6V and can do up to 4 cells, I am thinking I could run all 4 inputs in parallel to dump 4A.



I have ordered both to have a play with, I intend to initially test them with my variable bench power supply. Tracking shows the first has arrived in the UK yesterday, not bad as I only ordered it on the 18th.

[Oldgreybeard]

Somewhere I think I have some individual cell shunts I built years ago, just a TL431 and a darlington transistor, with a 10 turn trimpot to set the conduction point. Pretty sure they were made for LiFePO4 cells, Winston's, I think. Haven't a clue where I stowed them, though.

[Stinsy]

Great solution to your problem. But it isn't really a "balancing board" more of an "overcharge protection board".

[Joeboy]

Getting very impatient with watching these 4 cells charge in parallel, 3.5V x 10A = 35W
And my remaining Growatt battery banks are already at 55V with todays crazy sun.

So I've temporarily reconfigured them in series to make 14V x 10A =140W
Four times the charging POWA.

Just got to be careful not to short them out with the copper strip.

Damn, wish I could see your work Nowty! I am enquiring on.price incl shipping for 15 of 3.2V lifepo4 prismatics to a total capacity of 200Ah. I shall report back with the price.
At the moment my stack of 6 us2000 is 300Ah and on the insane prices seen would cost me around 9 grand. Another 200Ah would be a further 6 grand. Top end prices but just for reference....

I enjoyed seeing the photos today. Looks great!

[nowty]

The first of my battery protector's arrived today, not bad 6 days delivery time from China.


But I'm not happy with it as its supposed to operate at 3.5V but it doesn't.
The LED starts to flash at 3.6V but I can hardly measure any current through it, it seems to just pulse at a very low power, it might even just be the flashing LED.

It turns on full with a 2.1A discharge through the power resistors at 3.8V. Although nothing really bad is supposed to happen with LFP batteries until 4.2V, 3.65V is supposed to be the normal full charge voltage and I'm not content with having it going up to 3.8V every cycle.

Hoping my second one which should only be a few days behind arriving here actually operates at 3.6V as its spec.

[Joeboy]

That is a bugger. Always the risk with board type Chinese builds (gulp).

[nowty]

My second battery overvoltage protector has arrived so I am going to test it today. The one I already installed seems to work better than when I tested it with a variable bench power supply. It seems to keep the voltage at around 3.6V but I need more more current shunting, so I have been manually using a 1 ohm resister and limiting the max battery bank charging voltage for now.

[nowty]

I have now tested the second overvoltage board and it works exactly the same as my first one but only shunts 1.25A compared with the first one shunting about 2.1A. To parallel up the circuits, on the new board I would have to cut several tracks and link the negative rail across to each individual circuit. Its the sort of thing I would have done in the past, but to be honest I cannot be bothered with these new boards as they have surface mount components and my patience and eye sight is not what it was.

So I have ordered 3 more of the first type and I will fit 4 of them onto the new cell.

[Joeboy]

I have now tested the second overvoltage board and it works exactly the same as my first one but only shunts 1.25A compared with the first one shunting about 2.1A. To parallel up the circuits, on the new board I would have to cut several tracks and link the negative rail across to each individual circuit. Its the sort of thing I would have done in the past, but to be honest I cannot be bothered with these new boards as they have surface mount components and my patience and eye sight is not what it was.

So I have ordered 3 more of the first type and I will fit 4 of them onto the new cell.

Are the boards inexpensive? My delivery date for the new battery buckboard is 28th Oct! Only excuse is that it was in Turkish when I ordered.