PINNED - Lithium battery conversion/info

[Burgerman]

Yes. It tries to start and fails as the bat voltage rises fast with just low Amps. Once charging you may at times get a small wobble on the graph if it needs to change from boost to buck. No error.

[expresso]

i should have my Pl8 and the rest of the cable to connect it to the 12V PSU and will charge my chair - first time in a week - i used it very little indoors once on thanksgiving and outside for a short trip to bike shop to do my tires -

so should charge for a bit of time - to see how it handles it - will post results graphs when done - if all goes well -

using the same profile i use now - 25A

[Gnomatic]

In fact its hard to make it actually give an error unless battery is high resistance or cables long.

When I got this error I was charging off brand Gel batteries that have ~2X the IR of something like MK's. Why is this boost/buck startup error more likely to happen charging batteries w/ high resistance?

[Burgerman]

Because you add a tiny bit of current and the battery voltage jumps up (high impedance). So the charger has to go from buck, to boost as its starting up. As it does this the charge stops for an instant, so batt voltage drops again, and then the charger needs to go from boost back to buck and rinse and repeat...


Read...

https://en.wikipedia.org/wiki/Buck%E2%8 ... _converter

[Gnomatic]

Interesting.

So compared to my generic gel which have an IR of 7-8mOhm, what is the IR of 100+Ah Headway pack? Is the pack IR the same as an individual cell IR, or is that figure divided by the number of parallel groups or something like that? Just curious. Still have lots to learn.

[Burgerman]

Its not as simple as resistance on lead. A lithium yes, it is that simple.

With lead you have another bigger issue. The surface charge. That means that AC impedance which is the figure you refer to is far, far lower than DC resistance.

On lead the DC resistance, increases, and voltage under load drips over time. During a discharge. Or charge does the opposite. As the lead plate surface is in contact with the conductive electrolyte only on the surface. Discharge it for an instant and the layer of atoms (molecules) in direct contact with the plate turns to water. Which doesn't conduct. Charging does the opposite. The initial charge charges just a few atoms thick, and voltage rises almost instantly.

And it happens fast. So the only way you can see a low resistance with lead is to measure it with AC. So you cant compare lead and lithium in that way.

[expresso]

off topic - but funny - this is in Central Park - they have them all over the benches - this one is good

[frogsuk]

My kind of guy

[expresso]

I still didnt receive my new PL 8 - i was waiting to use it on the 12V PSU to charge - got tired of waiting and decided to just use my first PL 8 and its charging now


i forget to max out the Amps in the Supply Tab of the PL 8 - its at 50A now - i am using the 25A preset but its only charging at 20A steady -

So worse case would be 20A Max to charge with the 12V PSU - or when i max out the PL 8 to 60A - it may charge at 25A ?

either way - its a fine for me - 20A is fine to charge my ADD ON with Lead - and even my full pack - so the 12V is perfect for the size and works fine -

will post results when done -

[Burgerman]

It will do 25A or more.

[expresso]

ok 25A is decent - i have to set the supply tab on the PL 8 to 60A max ?


its at 50A now

[Burgerman]

ok 25A is decent - i have to set the supply tab on the PL 8 to 60A max ?


its at 50A now

Yep!

[shirley_hkg]

:(

My 45Ah pack with 1.5 mm SS box now weights 40 lbs.

Out of my expectation, it's heavy and high CG on the backrest .

[expresso]

Looks nice and neat - i would be ok with that -


here is the results with the 12V PSU - i updated the PL 8 to 60A in the supply tab - will see what this 12V PSU can do next time - i may just use it as my steady PSU for now at least

[Burgerman]

Looks OK to me. 21A when the batt voltage was lower.

[expresso]

next time i charge i should be able charger higher - i updated the PL 8 to 60A max in supply -

and as usual the post office i think lost my package with the new PL 8 - bag - wiring etc, - it should have been delivered yesterday - when i track it - says in transit but i know what that means - means they can find it or it went to wrong place

happened before same way - hope its packed well

[Burgerman]

Your postal service is as good as ours...

[Gnomatic]

:(

My 45Ah pack with 1.5 mm SS box now weights 40 lbs.

Out of my expectation, it's heavy and high CG on the backrest .

Shirley I love that little fold up stand on the back of your chair for a passenger!

[shirley_hkg]

It fits into the vertical posts, making it as close to the back as possible .

[expresso]

Looks very nice - sealed - safe and small compact - compared to mines - i just hang it in a Messenger bag in the rear

yours is nice flush to the chair - you never remove it ? i do like the convince of removing my ADD On off the rear and on to another chair in 5 mins - nothing to it

but if you never have to move it - which in your case you dont seem like you do - the way you mount it - its pretty much perfect -

[shirley_hkg]

Your chair is not to board a plane, with something swinging .

It's 40lbs after all, and takes 4 bolts .

[expresso]

thats ok with me - i am not big on planes - but true - needs to be bolted and not move if using a plane - i have it hanging and also a couple of those bungee cords - if thats good enough for a plane - i dont know - and worse case - dont have to take it - just unplug it and leave it -

either way i wont have that problem - but i like how you did it - compact - and stable bolted etc, - but the case is metal - how did you protect the Cells inside there so it wont touch the metal case etc, - i though that would be a issue

did you wrap the whole pack in a rubber case of some sort first - ?

[Dogsbody]

Every post read and understood...

That being said, why no BMS?

*stands on doorstep listening for a head exploding to the south*- joking...

I understand it mostly, need the practical skills...

I am slightly reminded of 'high fidelity'

[steves1977uk]

viewtopic.php?f=6&t=4460&p=84126

A BMS is not needed as it cuts power if you're trying to climb a steep hill or turn on thick carpets/wet grass. Also the balancing current is too weedy at 100ma, compared to the PL8 at 1000ma! Lots of reasons why it's not needed.

Steve

[Burgerman]

*stands on doorstep listening for a head exploding to the south*- joking...

[Dogsbody]

If I may say so from reading this currently 105 page thread all I can say is that you guys are some of the most committed, caring, assertive and above all patient people on the internet.

You should all be locked up, not normal...

As 'er indoors says, for a given value of normal.

My need is to build a pack for my partners Salsa M2 that is a permanent loaner NHS (NRS) jobbie that shares our home, and also a second pack for my lovely lady's sister who owns the same chair.

Management of the pack should be with PL8 balance charging in the first part. There is no argument that with a top balanced LiFePO4 keeping the cells in balance is a must. In several posts it has been pointed out that it does not need a full balancing charge every time (And balancing every time may stress some of the cells more, aging the pack inconsistently?).

I go back to a post about the grin technologies satiator charger. It is pricy however it is capable of putting in a charge based on metered Ah over a voltage end point. The currently beta 24-36V version is capable of 15A charge, and being designed for ebike market it is in a weatherproof shell. With careful setup it could be an answer for a weeks holiday, I can see a space under the seat where it would fit nicely, and it's charge rate can be programmed to be throttled... important when you can find yourself in a situation maybe needing a charge (admittedly reduced by extra and useable capacity) where there are reduced power sockets available (UK trains have power sockets for laptop/mobile use only, and both chair users use trains a lot).

Next, we know that lithium cells due to holding voltage for a long time during discharge will show the power chair's own fuel gauge to be a farce as it ? displays the voltage of the pack to approximate a fuel gauge. Now adding an ev power meter that measures the Ah used by the pack, we can have a better idea of how much we need to put back (possibly using a dumber charger than the PL8, but with a properly programmed charge profile). So something like:-

http://cleanpowerauto.com/store/#!/EV-Display/p/63894039/category=0

It's an expense, but this can be set so the fuel gauge can be a proportion of the installed pack. Let's say you have 100Ah capacity lithiums pack. You calibrate the fuel gauge at 100% capacity, tell the gauge it has 100Ah to use away you go..
But you COULD kill cells that way as top balanced the cells do not hit 0.0 at the same time, and later thread posts allude to using the middle 80%. So tell the gauge it has 80Ah to use, when the gauge hits 'E' you have a 20% reserve.

I wouldn't advocate forgoing individual parallel cells voltage monitoring as Less that 2.5V not good.

I have to consider certain extra complications as one chair is in the same house, the other will be much further away near London village, so not really a day trip to sort a problem out. I can test systems with my partners chair, and once robust as a unit can also put the same setup into the other chair. The closest anyone will beto getting the BM battery system over the counter.

Steves1977uk you fell for it..
BM you do have a sense of humour...

I am here to learn...

105 pages... geez...

[Burgerman]

It is pricy however it is capable of putting in a charge based on metered Ah

Does it know how much is put back every time you decelerate through regeneration and how much is lost through efficiency losses as heat/resistance? And how does it determine which cell group has the highest self discharge? Meaning that it will drop lower over a few days, worse over a few weeks?

As long as you only ever put back 90% of the Ah you use, that should be safe for occasional use. But I cant help thinking you are wildly over complicating everything. And any voltage based fuel gauge cant work other than at either end of the charge..

As for humour I know I am funny. My replies make me laugh all the time. Most people never get that. Very dry. People often think I am serious!

[Dogsbody]

Does it know how much is put back every time you decelerate through regeneration and how much is lost through efficiency losses as heat/resistance? And how does it determine which cell group has the highest self discharge? Meaning that it will drop lower over a few days, worse over a few weeks?

The satiator is a charger, pretty smart without the ability to balance the string, so normal operation of the chair will be using its standard controller.

Assuming hat the graphing capabilities of the PL8 enable you to see the point when the first cell hits Vmax (3.55V say) the satiator can be set to stop the pack charge at this point, or meter it back in on Ah useage.

And any voltage based fuel gauge cant work other than at either end of the charge..

The one linked uses the same setup as a clamp ammeter, measuring flow of electrons. Managing the expectations of the intended user means adding a certain amount of complexity from our side in the engineering of the system.

Better to say "here you go. Charge with this when the fuel gauge is low (PL8). Charge with this if no other option. If the fuel gauge says empty, There's 20% left. If alarms start sounding stop and charge. This will not happen as soon as you think. Any problems call me." Rather than "I can do you a battery, but if I do to maximise the usage of it you need to go to that website your sister has been complaining to you about because I've been reading it obsessively, read a 105 page string to get some understanding of what is happening, then I'll build you a battery based on those principles, you'll need to spend £1000 on cells and £200 on a charger, and read the manuals to know what's happening... oh and you need to carry this extra PSU if you want to charge from any AC outlet. And your battery meter built into the chair is no good anymore, so if it goes into the red you'll need to spend another £1000 on cells"

Yes complication is there. No it's not necessary, but I need to make a user friendly system. Providing the system is good and robust she can manage on a day to day basis then I can sweep in for more major issues.

The system can be cheap and simple, but not robust,
The system can be robust and cheap, but not simple,
The system can be simple and robust, but not cheap.
The system cannot be simple and robust and cheap.

[Burgerman]

Assuming hat the graphing capabilities of the PL8 enable you to see the point when the first cell hits Vmax (3.55V say) the satiator can be set to stop the pack charge at this point, or meter it back in on Ah useage.

No. All cells will stay at around 3.3xx volts from around 50% to 99.5% full. The volts dont change rapidly as it charges, or may even drop as charge increases at times, or vary by a few mV between cells at the same state of charge. Then at the very last instant, one cell at a time has its voltage shoot sky high as it reaches full. Think about it as filling a tin of petrol at constant speed and when full it suddenly overflows.... How can a dumb charger possibly know when thats going to happen. It may be all the cells that are 0.05v or so high or a few are, or one cell is .4v higher than the pack. So 3.3x becomes 3.8x volts!

All you can safely do is to add 80Ah back measured when you take 100Ah out measured. Or balance it as you charge . And you cannot do that below 95 to 98% charged because at that point all the cells read the same regardless of state of charge.

[Dogsbody]

So meter back in to the pack is the only way to do that providing the charger can do it. I believe that charger can, it's just a balance of cost over benefit, though convenience is the primary benefit with that charger.

Every choice is a compromise.

So the SOC/Ah/fuel gauge is a solid concept? I believe you mentioned something similar earlier in the thread, the one linked above will run off the pack voltage so as such will be a load across the whole pack, and not contribute to unbalancing the pack?

Many thanks,