PINNED - Lithium battery conversion/info

[shirley_hkg]

It's hard to find M4 lugs that accepts 10mm2 wire.
Steve

This is the biggest barrel for a 4mm lug.

[shirley_hkg]

APPEARS TO TRY AND BALANCE REGARDLLESS OF VOLTAGE. YOU DONT WANT THAT!

No. It doesn't.

Balance Start Voltage,
Cell voltages diffidence ,
are user defined. Balance will start only when BOTH are true.

[shirley_hkg]

This one seemed promising. It seems configurable. Most pricey….of course. Thoughts?

https://www.amazon.com/gp/product/B0CMP ... A0R09&th=1

Thanks

Not that pricey on Taobao , <$70 US.

[Burgerman]

Thats great. Set accordingly!

[shirley_hkg]

BM, would like you to advise on the charge and discharge voltages for lithium-ion pouch cells, compromising longevity and practical usage.

BTW , below is the basic settings of a JK bms.

[Burgerman]

BMS work OK on lithium ion cells. Unlike LiFePO4. As long as they dont fail and cause a firework explosion...

Voltages.
If you want maximum capacity then the usual 4.200V full. (up to 4.3v on some HV high capacity ones - read specs)
And 3.4 to 3.5V low end. GENERALLY.
On SOME high capacity types this can be as low as 2.7 volts, see spec sheet. Be sure you get this part right or you waste a lot of capacity and run the risk of killing your cells if its wrong in the opposite direction. They all vary by type of lithium ion. And cell manufacturer. Theres now many slightly different chemistries used. So no hard and fast accurate rule of thumb possible.

They live much longer (many times longer) if you only use say 70% of rated capacity. I generally charge to 4.100V per cell on hobby stuff like transmitters. And run them down to 3.7V if I have to. On flight stuff, where apacity matters I charge to 4.200v and discharge to around 3.4V if needed in an emergency. To get back on the ground gently.

Its all a compromise. If you use less capacity they last a lot longer. If you charge and discharge at a low rate they loe a lot longer. If you keep cool they live a lot longer. If you store below 85% charged by capacity as I do on my laptops then instead of getting a year and a half they last a decade or more.

Tesla charge to 85% in normal use. And read "empty" at 85% discharged. And that way their lithium ion last the life of the car with around 20% loss of capacity at a decade of normal use.

So its completely up to what you are trying to achieve.
If I am going to use a ithium battery laptop at an airfeild all day, I will charge it to 100% at 4.2v. Most of the year its sat plugged in to the AC and at between 80 and 85%.

So you can have the best of both worlds if you understand the compromises and adjust according to circumstances.

[Burgerman]

Electric bikes. On news just now. This last year (2023) there were 149 house fires in england only, attended by fire services. To put out fires caused by ebikes stored or charged indoors. 4 deaths, 22 serious injuries. 60 injuries needing hospital treatment. Of course theres a lot more that were not indoors, or didnt need the 999 services so not included.

This is why I would not use lithium ion in a wheelchair. We have bigger packs, and cant get off and run away. And when a small pack decide to go its violent and sudden. Just like a massive firework display. Ending is a fire.

With LiFePO4 cells that we all use in wheelchairs that cant happen as the cells themselves dont burn - at least not without external heat applied (such as in a fire). But if you use a BMS then this CAN cause a fire on its own. Every LiFePO4 fire that can be, is inevitably traced back to a BMS failure. And its not the cells that burn but the wiring, and the balance wires melt and cause the main wires to become shorted. Then everything else that can burn follows suit. That then causes the cells to be "in a fire" which boils the oily electrolyte out. That stuff is like cooking oil. It can burn if hot enough. But needs an ignition source.

So as long as we charge with hobby type chargers then theres no BMS to fail and short out anything either. So we do things in the safest way possible. Actually safer than lead batteries. Since when overcharged or faulty those give off hydrogen and oxygen which is a bad thing! The hindenburg was full of hydrogen... Theres many photos of exploded lead bateries on the internet. When they do they throw acidic gel or AGM matting out as well as start fires... But we accept this as we are acustomed to its risk level.

[shirley_hkg]

Always bear in mind.

It is another power bank for wheelchair, and it is not a fixture.

The ultimate add~on for dummy turns out to be more versatile than we thought. It does not just extending range when we are out for a longer ride. Many, who are waiting for new batteries, borrow these power bank to resume their daily lives. Application for new batteries usually takes months.

Will use 21700 cells this time.


Pouch cells are my project.



https://www.wheelchairdriver.com/board/ ... +for+dummy

[Burgerman]

I have the samsung 21700 5000mAh ones here for various things. 50G. Very good cells. Especially regarging self discharge. Absolutely no change on any of them over a year and a bit.

https://www.thunderheartreviews.com/202 ... rison.html

Note that THESE cells are capable of being discharged to 2.5V although I limit this to 2.7V. Which seems extreme but they dont seem to mind! And to get near to 5000mAh you must.

So 7 of these will be 29.4v charge (around 29v afterwards) down to just 17.5V cut off and so wheelchair own display should be fine. And the chair will stop probably long before a cell goes too low anyway.

[Burgerman]

Or you can use say 8S and 4p 21700 cells and build a physically small firework 20Ah 35V battery in a small bag carried on the chair push handles. And use this, to give say constant 27V and 3 to 5A limited CC/CV charge via XLR all day long. So adding 50% or 60% to the range. https://www.ebay.co.uk/itm/226087828265

Been planning to build one to lend out myself for years. Never get around to it!
It would extend the lifespan of lead by a long time, maybe double or better if set as suggested above.
4 or 5 Amps isnt enough for continuous rolling. But it would still more or less keep up as it catches up when you stop for a few mins or eat etc.
Remember that 4A charge is 20Ah restored every 5 hours. About the rate most people use up the battery. So low DoD average.

[fishinjunky]

BM I've been charging my 230ah pack 40amps and 3.60v should I be charging 3.55v or is 3.60v ok?

[Burgerman]

Both are OK.
Both will end up 100% full.
3.650 is also OK.

But the battery should last longer if charged to the lower voltage as lithium doesent like high voltages. In the same way it can be discharged down to 2.500V. But again if you limit this to say 3.00V you lose only a tiny amount of capacity and give the battery an easier life. So again it will last longer.

But its not that important since these big packs will last a very very long time anyway as long as you never exceed the 3.65, and 2.50 limits. Even doing this will still get you around 1500 to 2000 cycles. But be super kind and get 3 to 5 thousand instead. But by then you will likely have a different chair and fitted 400Ah batts...

[fishinjunky]

i was looking at the graph after it finished charging today. The BatteryRI says 13.5 (I think this is total battery resistance). The CellRI says 2.4 ( I think this is the resistance of the cells in total). The individual cells RI is 0.30. In the graph the green line is BatterRI you can see at the end of charge it went up from 9.5 to 13.5
its a 8 cell 200ah pack

Should I be concerned about the 13.5 BatteryRI ?

The previous charge a week ago the BatteryRI was 9.4

[Burgerman]

I wouldnt care at all. You should only really measure resistance as an AC impedance anyway, and only at around half charge too. Thats where it is lowest. The charger cannot measure resistance accurately at all. At best its a very very very rough guide. It can tell you if a cell is say 3 or 4x higher or lower and that may indicate an issue with cell or connections. And even then only really with hobby length connections like 3 to 5 inches...

[fishinjunky]

I wouldnt care at all. You should only really measure resistance as an AC impedance anyway, and only at around half charge too. Thats where it is lowest. The charger cannot measure resistance accurately at all. At best its a very very very rough guide. It can tell you if a cell is say 3 or 4x higher or lower and that may indicate an issue with cell or connections. And even then only really with hobby length connections like 3 to 5 inches...

thanks BM your reassurance always makes me feel better.

[Superchunk]

Hey everyone,

I spent the last couple of days reading this entire thread and my God what a journey that was! So I thought I'd make a post summarising what I understood from all that and what the situation is in 2024 with lithium so you don't have to read the whole thing.

BATTRY CELLS AND PACKS:

In the past cylindrical cells were the best choice as they had the best energy density and reliability. The commonly used cylindrical cell here was the Headway 40152S 15Ah.

Prismatic cells however have caught up these days and are now the cells of choice (the difference is the cuboidal cell shape so they can be arranged more easily). The EVE or CATL cells are commonly used now, where you just get the biggest Ah cells that can fit in your chair's battery box (usually the height is the limiting factor, also check the height with the welded studs).

CELL CONFIGURATIONS:

So now that you decided which cells you're going for, you need to decide on their configuration to make a 24V battery for your chair. There are 2 ways you can wire multiple batteries (of the same capacity/Ah), either in series - where the voltage is added up and capacity stays the same; or in parallel - where the voltage stays the same, but capacity is added up. This is where the S and P come from when combining cells. The EVE cells for example have a nominal voltage of 3.2. Therefore, to get the 24V needed for a chair we need a minimum configuration of 8S. If you can fit 16 cells in your battery box you can double your capacity by going for an 8S2P configuration etc. Here is a good place to play around with and visualise battery configurations and wiring: https://scriptasylum.com/rc_speed/lipo.html

When physically putting all your cells together you want to put a piece of acrylic or some other insulating material between each cell and compress all the cells together. Probably wise to put insulating material underneath and on top too.

BALANCING AND CHARGING:


Not all cells are made equal and even if they're from the same batch (which ideally they should be), they will not be in the same state of charge (SOC). Some of the cells will have more or less charge than the others. This is a problem because when charging, the most charged cell with be full first and will become overcharged while the other cells are still charging. Or when discharging, the most empty cell will die first.

So when making your battery pack you normally want to "top balance" the cells. The easiest (and dumbest) way to do this is with a standard bench power supply or just by leaving them all together. Simply use the bus bars to connect all the cells in parallel, pop your power supply on like 3.55/3.6 volts and leave them. Would take months to balance this way properly.

The other and most recommended way is to use a balancing charger. The charger will detect the SOC of each cell and balance as it charges. The most recommended charger for LiFePO4 batteries for our use case in this forum is the revolectrix cellpro powerlab 8 (PL8) with the FUIM usb adapter and burgerman's (BM) presets. This charger is discontinued now (although I believe shirley may still be selling some) however, so some users have been playing around with another hobby charger, the icharger 4010 duo.

The third option that isn't really an option just yet for us is to use a balancing BMS. I will leave BMS discussion to the BMS section.

BMS:

Ahh here we go, this dreaded topic. I'll try to be brief but essentially a battery management system (BMS) is designed to limit the environment in which the battery can be used so that the cells do not get damaged. In most cases, it really sucks at doing so. When discharging, the BMS has a maximum continuous discharge amperage of usually around 100-150A. In a modern power chair, driving up a ramp for example can suck up to 240A continuously. So if you had a BMS you will get cut off and will therefore be compromising your safety for the safety of your cells.

It is also important to note that LiFePO4 cells have a max continuous discharge rating such as 1C where C is the capacity. Therefore for a 300Ah cell the max continuous discharge would be 300A if it was 1C or 600A for 2C etc.

So now we established that a BMS will actually cripple your chair when discharging during use. What about a BMS just for charging?

Well, theoretically a BMS should balance your batteries as they charge with a non-balancing DC charge input. However the way it does this is flawed. Essentially given that the BMS is inserted after the charger in the circuit (between the charger and the battery), all it can do is switch off the charger for a bit if one cell is getting overcharged, let them balance and discharge a bit, then turn the charger on again and repeat until they're balanced. This ruins your cells because it uses up your cycles unnecessarily. Not only that but a BMS uses MOSFET (basically electronic switches for the purposes of this) to turn on and off the charger which are not reliable, and a broken BMS = no power = broken chair. The only advantage with a BMS theoretically is that you can do 2 wire charging through your standard charging port on your chair. (Although I'm not actually sure how much current the r-net joystick/bus cable can handle, if BM could shed any light on this, would be appreciated! )

I think that's everything you need to know if you just arrived here and want to try lithium in your chair in 2024! Please do scream at me if I got something wrong or if there's something I should add.

[Burgerman]

Prismatic cells however have caught up these days and are now the cells of choice (the difference is the cuboidal cell shape so they can be arranged more easily). The EVE or CATL cells are commonly used now, where you just get the biggest Ah cells that can fit in your chair's battery box (usually the height is the limiting factor, also check the height with the welded studs).

They havent cought up. They have incresed in energy density that we can fit into the same battery space by around 250% but these cells are lower C rate. As long as you install ADEQUATE capacity this doeent matter.
So with Headway cells you could typically get around 96Ah in that battery place. Now we it 230Ah in the same volume giving a HUGE range and longevity increase of say 2 decades and 500% range gain. Mostly because of the capacity gain means lower average depth of discharge daily. And or charged less frequently so the same cycle life lasts twice as long or better.

CELL CONFIGURATIONS:

So now that you decided which cells you're going for, you need to decide on their configuration to make a 24V battery for your chair. There are 2 ways you can wire multiple batteries (of the same capacity/Ah), either in series - where the voltage is added up and capacity stays the same; or in parallel - where the voltage stays the same, but capacity is added up. This is where the S and P come from when combining cells. The EVE cells for example have a nominal voltage of 3.2. Therefore, to get the 24V needed for a chair we need a minimum configuration of 8S. If you can fit 16 cells in your battery box you can double your capacity by going for an 8S2P configuration etc. Here is a good place to play around with and visualise battery configurations and wiring: https://scriptasylum.com/rc_speed/lipo.html

Yes gfor education. But realistically you want 8x 230Ah cells in series for the best energy density in pace of 75 to 80Ah lead bricks in common use. These JUST fit if laid out correctly.

BALANCING AND CHARGING:


Not all cells are made equal and even if they're from the same batch (which ideally they should be), they will not be in the same state of charge (SOC). Some of the cells will have more or less charge than the others. This is a problem because when charging, the most charged cell with be full first and will become overcharged while the other cells are still charging. Or when discharging, the most empty cell will die first.

So when making your battery pack you normally want to "top balance" the cells. The easiest (and dumbest) way to do this is with a standard bench power supply or just by leaving them all together. Simply use the bus bars to connect all the cells in parallel, pop your power supply on like 3.55/3.6 volts and leave them. Would take months to balance this way properly.

You can set a power supply anywhere between 3.45V and 3.6V and they should all end up charged pretty fast if you set a high current. But you are charging a 230Ah x 8 battery... So about 1000Ah if half full. At say 25A that will take 40 hours. But you dont know how full they are! So charge at 50A if you can.

BMS:

Ahh here we go, this dreaded topic. I'll try to be brief but essentially a battery management system (BMS) is designed to limit the environment in which the battery can be used so that the cells do not get damaged. In most cases, it really sucks at doing so. When discharging, the BMS has a maximum continuous discharge amperage of usually around 100-150A.

Anything from 5A to 1000A depending what you buy. Half the rated value is more realistic.

In a modern power chair, driving up a ramp for example can suck up to 240A continuously. So if you had a BMS you will get cut off and will therefore be compromising your safety for the safety of your cells.

Its more complicated but yes. But it can be down to cell resistance and voltage drop etc. And if the cells you chose are BIG ENOUGH then its not required and voltage drop is lower too.

It is also important to note that LiFePO4 cells have a max continuous discharge rating such as 1C where C is the capacity. Therefore for a 300Ah cell the max continuous discharge would be 300A if it was 1C or 600A for 2C etc.

Higher discharge rate cells are available. But are physically bigger. So less capacity so a larger discharge rate capability is then needed...

Well, theoretically a BMS should balance your batteries as they charge with a non-balancing DC charge input. However the way it does this is flawed. Essentially given that the BMS is inserted after the charger in the circuit (between the charger and the battery), all it can do is switch off the charger for a bit if one cell is getting overcharged, let them balance and discharge a bit, then turn the charger on again and repeat until they're balanced. This ruins your cells because it uses up your cycles unnecessarily. Not only that but a BMS uses MOSFET (basically electronic switches for the purposes of this) to turn on and off the charger which are not reliable, and a broken BMS = no power = broken chair. The only advantage with a BMS theoretically is that you can do 2 wire charging through your standard charging port on your chair. (Although I'm not actually sure how much current the r-net joystick/bus cable can handle, if BM could shed any light on this, would be appreciated! )

12A max RMS which isnt really enough even for a small lead battery...

Heres a XLR connected PL8 charging at 12A the max allowed on a lead gel battery.

[shirley_hkg]

Prismatic cells are much easier to diy, and far more economical .

Though BMS might not be the best, the good one is an alternative.
We can get rid of the sudden cut-off anxiety by discharge bypassing it.
To avoid over discharge, monitor it by voltmetre through rule of thumb , and with capacity more than one could drain in a day .

[Superchunk]

12A max RMS which isnt really enough even for a small lead battery...

Heres a XLR connected PL8 charging at 12A the max allowed on a lead gel battery.

Interesting that it really only is 12A through R-Net. That really sucks.

I'm just about to pull the plug on some EVE MB31s for my Permobil m5 (gonna be a real tight squeeze but hopefully they go in . Might have to grind some of the studs down I think.

Probably gonna go for the icharger x8 and a 30A PSU. But ngl, I am still tempted to whack in a high current BMS and charge through the r-net XLR port. I mean, what sort of decrease in lifespan would I practically get?

[Burgerman]

Its not about the decrease in life so much.

Be careful as permobil are doing everything possible to stop you actually getting 230Ah cells in. Height issues... Permobil are problematic in lots of ways but I am not going there again!

Also the 230Ah ones are 50mm deep. The ones you are looking at are 70mm. So I would be very surprised if those fit!

These are the correct physical size, where 8 can normally replace a pair of grp24 batteries. If arranged correctly. But this will be a bit too tall in a permobil.
230/240Ah. https://www.evlithium.com/LiFePO4-Batte ... -cell.html
https://www.evlithium.com/catl-battery- ... ttery.html

[Superchunk]

Also the 230Ah ones are 50mm deep. The ones you are looking at are 70mm. So I would be very surprised if those fit!

Ahh schmucks! was too concerned with the height and width, totally missed they are different depths too, thanks for pointing that out! I was too dazzled by the capacity

Its not about the decrease in life so much.

Is it more the mosfet reliability/hazard?

[Burgerman]

Its about 30 different things.

The problem with a BMS is that its mostly doing a load of things we do not want at all, that in a correctly SIZED pack isnt wanted. And if its programable enough then all those protections and cut off etc can be adjusted until they do not do anything, and so do not trip you up and stop the chair. But then you just might as well not have the thing in the first place. All you are doing is adding a bunch of unreliable electronics in the wrong place. The control system we are using is already current and volttage limited and it does this in a proportional and smooth way. So all the protection stuff isnt wanted.

If you use it to control cell balance ONLY then you again have a device that cannot control the charger.
Q. How does the charger know whan the battery is charged, and balanced?
A. it doesent. So what point does it turn off? If it stays on it holds voltage too high for too long = short service life. If it turns off before balanced then battery can get more and more unbalanced over time.
Thats just ONE issue.
The other problem is it lets the high cell reach a too high voltage then cuts off charge while the normally too weak balancer pulls down the high cell. Then rinse and repeat for hours to days depending on if the charger stops or not, and it may or may not end up balanced...
Another issue is that when the mosfets fail the wiring melts balance or main, and everything shorts out and catches fire. Not the battery just the plastics etc. This is how most lifepo4 fires begin.

And theres more. But if desperate charge with a power supply set o a low enough voltage that it will not cook the battery, not the recommended 3.65V and stop it when you know all cells are balanced, and when residual charge current has fallen to say 1/320th C. Which is what I usually set the PL8 to. But the PL8 reduces output to 10mA if needed, while the balancing happens. It chosoes the correct current as it flls up and balances so that the 3.55V I choose is never exceeded on any cell. With a BMS expect repeated 3.8V or so over and over as it keeps reaplying full charge current...

Anyway just believe me when I say a BMS is a thing designed to stop cells going back under warranty when installed in the wrong device or undersized...

[ex-Gooserider]

Mildly interesting note... Several years back I bought a bunch of the 15Ah Headway cells intending to do what was (at the time) the best available LiFePO4 setup... I did the suggested charge, let sit for a month, and recharge to measure the self-discharge and build a 'balanced' pack... Then other things in life happened, including that bug going around, the maker-space move, and so on, and the cells were left sitting, fully charged...

The ancient Odyssey knockoffs I've been using for years have started really showing their age in the form of shorter distances before losing LED's on the joystick... Since I have them, I just recently started building out the Headway pack... I haven't tried charging yet, but looking at the voltages, so far every cell has been at 3.315 - 3.330V after sitting for YEARS in the OEM boxes...

I'm impressed.....

ex-Gooserider

[Burgerman]

Sitting full isnt a wise idea. They dont like it. Esp as self discharge is really low so the stay that way.

Imagine breathing in as deep as you can. About to burst. Then holding your breath for years... Thats how they feel!

[shirley_hkg]

20% loss in capacity, I guess.

They are 10C rated so can handle wheelchair loads well, though capacity is reduced.

[ex-Gooserider]

Sitting full isnt a wise idea. They dont like it. Esp as self discharge is really low so the stay that way.

Imagine breathing in as deep as you can. About to burst. Then holding your breath for years... Thats how they feel!

I wasn't PLANNING on having them sitting full for anything like this long.... However life happens and it's what I have to deal with. Probably still better than the lead bricks I'm sitting on now. Given what I paid for them back then, I'd like to get at least some use out of the cells... I am a bit concerned about the cotton balls I'm finding under the screws on the anode ends, but hopefully that isn't a problem (see my other post on this..)

ex-Gooserider

[Burgerman]

Might be because you charged full then stored for years... Some pressure inside leaked out.

[ex-Gooserider]

Might be because you charged full then stored for years... Some pressure inside leaked out.

That makes some sense if it was on the cathode end, as that is where the pressure relief valve is, but supposedly the anode end where I am finding the cotton is supposedly sealed.... It also ignores the question of where the cotton came from if it wasn't from manufacturing...

FWIW, I'm finding the balls in every cell that I've looked at so far, which I'm actually finding a bit of a relief as it makes it seem more likely to be a manufacturing thing if it's consistent...

A second question that I'm wondering about, now that I'm actually building, is what is the best way to do the power lead connections? Balancing wires are easy, just a crimp terminal, but what about the power, as it seems a bit of a high resistance point to just connect to one terminal screw... I'm using AWG6 power wiring (probably overkill) and it bothers me a bit that the cable is thicker than the terminal screws....

I'm thinking of bending a couple of the link strips in half so that I'm at least bridging across a couple of terminals....

ex-Gooserider

[Burgerman]

Dont crimp...
Or if you do solder it as well.

Yes your cables are overkill. I just use a ring terminal and solder it same as balance wires and add a 150A MIDI fuse.

[ex-Gooserider]

Another question - I'm going to put the pack into the chair w/ it sitting on the plastic building block "pegs" which don't seem all that sturdy to me. I'm wondering about putting some sort of padding / cushion under them to help soak up the bouncing around they'll get from rolling and so on... Any suggestions for material? Since it will also be resting on the junction plates and screws I'm a bit concerned about moisture (especially once the road salt and grime gets in) causing a short - any good way to prevent that? I do have a sheet of some sort of plastic I can probably put between the padding and the plates which might help.

A few options I have on hand -
1. A foam "Yoga Mat"
2. I have a bunch of sheets of the white packing foam the batteries came in
3. I still have a chunk of a closed cell foam mat that was originally used on a bone density scanner (x-ray machine)

If these wouldn't be good, any other suggestions?

ex-Gooserider