PINNED - Lithium battery conversion/info

[shirley_hkg]

You seem happy !

[Burgerman]

Yes cells have resistance so loading them up to drive a scooter warms them a few degrees. Lithium cell resistance reduces, nd voltage drop under load follows on so reduces less. And warming the pack increases voltage slightly.

Nw you know why C rate and resistance matters!

Are you wearing out faster than the battery yet?

[shirley_hkg]

Is the discharge curve steeper than just lithium because it's a discharge curve for a combined lead & lithium pack?

That is one big single lithium battery , meant for INVACARE ARROW's G24 compartment .

CCS stopped responding at the 8th hour , so you only see the last 4 hrs of the discharge .

[expresso]

i re done the wiring with cables instead of the copper buss bars - what i found was - 3 loose connections - they were a starting to be stripped so i used a longer screw on those 3 with a lockwasher to thicken it and it tighten down nice now - the rest i didnt use lock washers and shorter screws - all tight - its charging now - so far so good - as long as they stay this way -

also changed the fuse to 100A - this is the 80ah ADD ON for the bounder -

Do you find it better to make long cables than short ones?
Is there any advantage?
Any better than buss bars?

Mine take the shortest route.

i didnt find making the cables hard - they are better than the buss bars i had - mines were very thin - actually i recharged after and the graph was better at the top balancing part - lines were tighter -

BUT i think i found the issue and dont even think it was the buss bars - i still like the cables better- but the Cells - the posts you screw down into - its very soft metal - and easily stripped - i noticed i had 4 that were getting stripped -

on those four - i added a lockwasher and a longer screw 12mm - which worked - it screwed in nice and tight - the rest i didnt add the lockwasher - and used 8mm screw - everything was triple checked - i didnt use locktite or lockwashers - just flat washers

only the 4 that were loose - getting stripped - if i keep removing and installing them - they will get totally stripped - now i will install and cross my fingers i dont have to touch it again -

if i notice something is wrong - i will remove it and check - if i have to add lockwashers on all of them - i will using 12mm screws - if that dosnt work - then not sure how to fix it - maybe locktite and lockwashers -

or what scooter man did - grub screws first locktite them in - and then never remove them - and use lock nuts on top flange -

[expresso]

Nice on expresso, neat job.

You don't want short straight links Terry, they need to be loke expresso's . So if cells move slightly when hitting bump is doesn't loosen the or strain the terminal connections.

On vibrating machinery you never terminate with straight cable, you always put a helix in the cable.

thanks - i didnt known that about straight cables - good to know - it was just too short of distance to work really - i have just one straight to connect them both - - if i knew i would have made a new one - now that i think about it - makes sense

i hope it dosnt give me a problem now - i may end up using those grub screws down the line if i have more issues - i would locktite them in there first -

[Dogsbody]

Hi,

Been lurking a bit..

Don’t think this is a BMS, but balancing circuits... any chance you could take a look for an opinion?

http://www.deligreenpower.com/deligreen ... ad-image-2

Also Shirley can you help me with A TaoBao purchase please?

Dog

[Burgerman]

Yes it balances at some small unknown current all the time. No spec provided. Regardless of voltage. Not sure why you would ever want that. Certainly no use for charging, as it cant control the charger output. And you should only balance LiFePO4 once above 3.45V anyway as it will UNBALANCE the pack all day long the rest of the time which is bad.

[Scooterman]

You seem happy !

It's good Shirley, I'm pleased I did it (with your help)

Yes cells have resistance so loading them up to drive a scooter warms them a few degrees. Lithium cell resistance reduces, nd voltage drop under load follows on so reduces less. And warming the pack increases voltage slightly.

Nw you know why C rate and resistance matters!

Are you wearing out faster than the battery yet?

Ah! thank you that clears that up

And as to your question, YES I am wearing out faster than the battery. Born out by the fact that I had to spend all day yesterday on the bed, I felt like this But it's great that the 1-day range of the scooter far exceeds my physical endurance. I reckon my endurance and concentration could manage 60 miles at a push (4hrs AM, 4hrs PM) on a warm sunny day, but that would be an exception. And the lithium range exceeds that.

That is one big single lithium battery , meant for INVACARE ARROW's G24 compartment .

CCS stopped responding at the 8th hour , so you only see the last 4 hrs of the discharge .

Ah, got it

i may end up using those grub screws down the line if i have more issues - i would locktite them in there first -

I'm sure while browsing the web I've seen lithium cells with studs terminals instead of tapped holes. But I think they were on cylindrical lithium cells?

Btw I had a bit of a lightbulb moment yesterday while laid up in bed. I was trying to think of a alternative/better way to stop a lithium pack moving around in a battery compartment, cos I think movement is what loosens connections.


and I thought of,


drum roll......................

Wadda you reckon or

[LROBBINS]

Clearly that device is unsuitable, but I do wish someone could find a cell balancing board that would: (1) start balance only near or at the end of CV, (2) dissipate or exchange enough current to balance reasonably quickly (would 3A be enough?). It should be rugged enough to be mounted in the chair and not need any user attention (after initial programming). As it would only be doing something when connected to a charger, it can be a passive balancer that just dissipates energy from the highest cell - the mains watt-hours wasted would be tiny in any case.

There are lots of cell monitor/balancer boards out there that use an Analog Devices or TI chip that can do the above, but that can pass only mA of current. One of those chips plus 8 MOSFETs and 8 resistors would seem to be all that we really need, but I'm not up to designing and testing same (at least not with too many other projects underway). So, with all the "stuff" that's out there, isn't someone making a simple board like this?

The only thing that has kept me from going LiFePO4 has been the fact that we can't have a setup that only I can hook up for charging. It's hard enough to get people to remember to push the rocker switch on her dumb charger (and even I have forgotten once in a while), no less follow a sequence of who connects to whom followed by one or more button presses. It's not just that I'd rather not have yet another daily chore. Perhaps if I were 40 or 50 y.o. that would be OK, but I've already got 73 behind me and Rachi is going to still need her chair well after I'm gone.

[Burgerman]

Clearly that device is unsuitable, but I do wish someone could find a cell balancing board that would: (1) start balance only near or at the end of CV, (2) dissipate or exchange enough current to balance reasonably quickly (would 3A be enough?).

No it must be equal to or greatr than the charge amps. If one cell hits the top voltage allowed, while the rest are all below that point the charger continues to charge all cells at full power. Thats why you need a PL8 style charger that sees this at a cell level, not the total voltage, and so reducesoutput power to whatever level is needed to allow the balance current at your proposed 3A to keep that cell below the CV per cell voltage.

And a BMS tries but cant do this, either (obvioulsly) and so it balances, and repeatedly disconnects the charger in a ON/OFF fashion while its insignificant current pulls down the high cell. One problem with doing it this way is that it must have a high hysterisis. Because of resistance. (The volts drop as soon as it cuts off the charger and visa versa). So a BMS allows anything up to 3.8V (!) to happen, then disconnects the charger completely till the balance current can drag the cell down to below the CV per cell voltage. So to do this it waits till its down to 3.5V... Then rinse and repeat. For hours. Because the BMS is in the wrong place, and has no proportional control over charge current or balance current.

[foghornleghorn]

Shirley has kindly got me 16 of the EVE-LF90 cells which after a bit of time wasting by UPS arrived yesterday

Going to be using them in my wheelchair as a 2P8S pack.

I've maybe overloaded myself on too much information from reading all the previous threads/posts and would like to clarify the best way to proceed.

• Charging each cell to a fixed value. Leaving them for a few weeks to self discharge and measuring the loss. Making 8 pairs with similar self discharge over the set.
• or Charging each cell to a fixed value. Discharging to 3 (or 2.9 volts) to measure capacity. Making 8 pairs with similar capacity over the set.

I have a PL8 bought from Revolectrix in November.

If it has any bearing over any of this my usage is likely to be almost every day once the weather is a bit warmer doing a decent amount of miles, but during the winter not so much unless I now have so much spare capacity I can add heating to my seat.

[Burgerman]

When you have many small cells like in the headways, its worthwhile doing that, both capacity and more importantly self discharge measurements and then building closely matched strings. With only 16 cells its a lot of work for little gain. So you could just build it and do a single test with all cells before fitting.

[foghornleghorn]

So just build the pack, test it as charging and discharging fine once assembled, then if ok fit to chair and start using.

Not worrying about matching cells up in pairs at all?

[Burgerman]

Depends on how much time and patience you have.

The idea is that you would make 2P groups with the same self discharge rate primarily. So charge all full. Then remove say 5Ah exactly from each. Then store for a month or more. Then charge each one individually. Note Ah returned. There will be a few mA difference between cells. So if one is high, match it with a low one. So its average like all the other groups.

You can do the same with capacity too. If you want. Thats not so important with headways as they are remarkably similar and with say 72 cells they average themselves.

[foghornleghorn]

Does the self discharge indicate an imbalance that will be amplified in use, or is it more a factor affecting leaving the chair sitting unused for a week? Just trying to get my head around whether pairing self discharge is more important than pairing capacity.

[Burgerman]

If 7 of your 8 groups lose 20mAh per day, and one loses 120mah per day, and you charge daily, not a problem After a month without chaging, thats 30xdays 100mAh so one group are now 3Ah out. On subsequent charge it will take 3 hours to reblance with the PL8.

It may be closer or further apart than this. Only measuring can tell.

As far as capacity testing goes, then as long as you dont run it low, it wont ever make any difference. So a large pack, not a problem Small pack, where you may empty all cells to 95% but one group is 100% empty then you have issues...

So with a big enough pack, capacity testing and matching doesent matter much. But charging weekly, or monthly the self discharge becomes more important. Esp if unused and stored for 6 months...

[shirley_hkg]

If used everyday , self-discharge rate shouldn't be a big concern , or the cell were severely defective otherwise .

Capacity matching is more meaningful to me .

Set up two 1P8S batteries . Perform full discharge down to 2.98V via PL8.

THEN ASSIGN THE STRONGEST CELL OF ONE GROUP TO THE WEAKEST OF THE OTHER .

I'd get the lowest terminal voltage of pack , before the weakest cell pair reach 2.98V. This will make everyday use more comfortable .

[shirley_hkg]

Don’t think this is a BMS, but balancing circuits... any chance you could take a look for an opinion?

Also Shirley can you help me with A TaoBao purchase please?

Dog

Only balance LiFe when they are full, and big balance current always help .

http://chargery.com

[expresso]

looks like it does it all - is this the next step up over the PL8 for plug and play situations ?

[Burgerman]

I would say its much more complex connections and requires a bunch of modules, than the PL8 is, so less plug and play. And as always the devil is in the software and firmware details. And we dont know any of those. And no PC control or CCS type setup either. Not for me. But its an untested alternative if you want to charge more than 8S with one charger.

[Scooterman]

The power of Apple Machintosh

[Burgerman]

What you going to use that for?

[Scooterman]

Re possible self-discharge imbalance during winter months. What wattage load resistor would you recommend for giving the PL8 a bit of assistance in pulling down a high cell?

[Burgerman]

Depends how much.

[Scooterman]

Depends how much.

How long is a piece of string

Fair enough, it obviously depends on how much out of balance.

I know WD.com probably won't approve but as I'm only a part time power chair user and me and my powerchair are half the weight of most powerchair users when I do lithium I thinking of going for a smaller 8 x 90Ah lithium pack if Shirley can still source cells. It's easier to wire up, no worry about fitting in battery compartment, self-discharge won't be such an issue, cheaper, about twice the range of lead, light enough that chair could be lifted into a vehicle by two people, or one person by tipping back and loading castors first, and it's not for long road trips as I would rather use the scooter for that.


small and v.lightweight

[expresso]

you can do easy 150ah in there with same no worries - light enough and will fit - there are many Cells - Sizes etc, and cost will be about the same - dont rush doing 90ah if you can fit 150ah - you will one day regret it - plus works less - will last longer in years !!!

i bet it wont even cost more if you pick the right seller - i like that chair simple small compact-

[Burgerman]

200Ah fits.

16 of the cells you just recieved expresso.

[expresso]

yes they do - i just test fitted them in my battery tray from the 646 chair - now - i would have to cut one end of the tray off - the short end - front and back side - i think i can do just one cut - and thats it - leave the front alone with the handle on the tray - cut the back which goes in the chair - will be near the Rnet 120 CM

take a look at the pictures

[expresso]

more

[expresso]

more