PINNED - Lithium battery conversion/info

[shirley_hkg]

That's why quality of cell is essential.

[Burgerman]

The cell quality that you can get today from the likes of docanpower is great. Certified, garanteed, with capacity certificate and matching bar code etc. All the EV A grade cells sold as 230Ah actually measure 245Ah or a fraction more. All close matching low impedance.perfectly flat cells, good terinals etc. Its not like it was a few years ago.

Also not sure active balance is wise unless you can stop it doing anything below 3.45V absolute minimum. It should ideally also balance only when voltage is held steady. So while on charge. Too much gap between cut off and restart charge voltage. If you make that small enough then it helps some.
In a chair they just take too much space. In the battery area theres not a mm to spare unless I fit small cells. Which as you know destroys many of the advantages. Solar, yes. So will get one for that.

[shirley_hkg]

These 4 parameters together will do the job, to some extent at least.

I have the 0.4A smallest one on a 125Ah pack for 6 months now. So far so good.

[shirley_hkg]

youtu.be/rQgOet7AH0I

It balances by charging a capacitor from the highest cell, and then charges the low cell by the capacitor.
This is done one by one. Works all day round.

Though it's 2a, unlike PL8's 1a on 7 cells together, it's much slower

[Burgerman]

But thats bad.
Because voltage does not work as a proxy for state of charge.
So as soon as the voltage drops below 3.45V it actively unbalances the cells all day long...
For e.g the warmest end of the battery will push many Ah into the coolest end if parked in the sun...

And even at the same temperature the voltage of a cell can INCREASE slightly as it becomes more discharged due to internal resistance microscopically warming the cell. That then "appears" as the most charged cell when it is the opposite. So active unbalancing begins...

And the BMS itself, doesent have perfect voltage calibration across all cells. If its 1mV out, on say cell 2 then it will spend all day trying to equalise it. In the flat part of the curve where the voltage doesent drop with state of charge.

Even worse, balance at say 70% charged, is completely different to balance at 100% charge. So it makes things worse!
So active is bad on LiFePO4. Works great on lithium ion like a laptop or a tesla.
With LiFePO4 the ONLY time you should balance is at the top, and above 3.45V. So during charge only.

[shirley_hkg]

No.
You set the "start balance volt", and the delta V.

It won't balance below these.

[Burgerman]

So it can be correctly set to balance when basically above 3.45V per cell, not global voltage?
If so then it will basically only balance when on charge. Again thats the correct thing to do.
So it can only balance at 2A. But I am charging at say 3A or 30A...
High cell will then go super high because its still charging at full power Or does this also then turn off the charge current at 3.55V? I dont think it can.
So it cant work properly and will let the high cells go way over the safe voltage under charge. And it shouldnt do anything if not on charge?

[shirley_hkg]

OVP will disconnect charger. Any cell.

Any cell is ≥ StartBalanceVolt, balance continues. (if delta V meets ). Charging or not charging.

Charge resumes if all cells below OverVoltprotectionRelease.

All volts are cell volt, not pack volt.

[Burgerman]

ON A BMS YES. The endless on off bounce...

But not on this:
Active balancer has no way to disconnect charge? In your pic above. Lets say I am charging at 40A...

https://www.wheelchairdriver.com/board/ ... 26#p205660

[shirley_hkg]

A new wheelchair powerBank, 8P7S, using LG M58T cells.

Just finished connections , and is undergoing initial charge.

Will see what would show up tomorrow morning.

[Burgerman]

Thats not LiFePO4, all sorts of balancers work on other lithium chemistyr where state of charge = voltage. Thats different to LiFePO4 where this is not the case.

[shirley_hkg]

This is the smallest BMS with 0.4a balance current.

First picture is settings ,
charge volt @4.2v
Charge resumes @4.135v
Balance ≥ 3.95v

Second picture shows it hits 4.2v four and a half hours ago. Charger disconnected (would not resume as cell 2 is higher than 4.135v) , but balance continues.

Balance is pulling High cell down.

[shirley_hkg]

5.7 hours from last charge, it resumes charge when all cells are ≤ 4.135v.

A few more cycles to go, before a full balance.

It's slow, especially when SOC is obvious in an initial charge.

It will do in one go during daily use.

[Burgerman]

Important!!! Q. There are no large connections on the active balancer in your image. How does start and stop a charger?

The difference between 4.315 and 4.200V in lithium ion chemisty is massive. about 90% charged. And with lithium ion you CAN balance at a lower voltage. So I would expect that to work.

Not so on LiFePO4 though. Because voltage is not equal to state of charge below 3.45V and at lower voltages the difference between cells can be extremely small. So if it allows it to try and balance with charger "off" then it will not actually balance anything because the resting voltage, off charge is already less than the recovery voltage. So set say 3.550V and then turn off charge and it will all drop lower than the recovery (restart charge?) voltage. So at this point balance will be excrutiatingly slow!

Try it with a big 200Ah lifepo4 set to 3.550V charge max.

[Burgerman]

Charger disconnected (would not resume as cell 2 is higher than 4.135v) , but balance continues.

HOW was charger disconnected?
Are you doing it manually or are you using a BMS, or just an active balancer?
Or are you using a BMS AND an extra active balancer?



And instead of setting 4.135, reconnect charger which will work on a Lithium ion, but need many cycles... Or it wont end up full. Why not set it to 4.198 and over voltage to 4.201? Because thats whats really needed.

Because on a lifepo4 havig a huge 85mV gap between charger on, and off and attempting balance at this wide spread wont work well. If at all. at least not accurately or inside a week...

[shirley_hkg]

HOW was charger disconnected?
Are you doing it manually or are you using a BMS, or just an active balancer?
Or are you using a BMS AND an extra active balancer?

..

It's a JK BMS just, with build-in active balancer @ 400mA. (It doesn't balance by resistors)

I made one the same BMS for "jefferso", that he can plug in his chair's Dsub, and charge with dumb charger when away from home.

Tested with a 180Ah pack and results as follows.

[Burgerman]

I thought you were talking about an active balancer. Not a BMS with one.

Because your image / video, shows only a balancer with no large connections.


youtu.be/rQgOet7AH0I

Only difference in practice between active and non active in a bms is that theres less wasted energy so less heat. Has all the same BMS issues as before, take up lots of space, bounce up/down and fire risk etc. So you can use one on a chair if you dont care about that, and also use 160Ah cells instead of 230Ah ones.

Its an option, if you really must. But compromises are not worth it to me. In a chair. As I said before, I may still use one on my solar system. As theres no better alternative for that battery. But I want a PC version, not a rediculous telephone one!

[shirley_hkg]

Sorry. I just wanted to tell it takes 2 steps to get energy from A to B, not 1.

[shirley_hkg]

14 hrs, and it's done.

[Burgerman]

14 hours? How big was the battery? 160Ah? Or less? Holding high for 14 hours = swelled cells. Its like charging full and holding it for half a day without relaxing. Especially on prismatics or lipo.

Why so long?
If it only does .4A and one cell at once(?) then you actually need an 8A active balancer?

[shirley_hkg]

A new wheelchair powerBank, 8P7S, using LG M58T cells.

Just finished connections , and is undergoing initial charge.

Will see what would show up tomorrow morning.

[Burgerman]

It will work fine on those cells, as they are just common lithium ion laptop style inr 21700 cells. I use these things in a quadcopter and with a cheap BMS in my RC transmitter. Because on the typical cobalt type lithium ion cell voltage does = capacity. You can balance them at almost any voltage above about 3.3 volts all the way up to 4.2V per cell. And they are only about 5Ah. (Mine are tesla 5000mAh panasonic? 21700 ones).