Increasing range with parallel li ion batteries

[newb123]

I was wondering what is the practical limit for increasing the range of li ion cells using a parallel config

for ex. a 42 volt 10S2P pack increased to say a 10S8P....what is the practical limit of parallel cells..I usually see 4 but cant they be increased up to 10, for ex.?

 

[john61ct]

Well one limit is physical, weight and volume.

Another is layout, paralleled "sub-pack" strings cause imbalance issues past 3 or so.

But if the paralleling is done at the lowest level, before connecting in series, no problem,

technically 8P10S, rather than 4P(2P10S).

2P(4P10S) or 2P2S(4P5S)

would each be a reasonable compromise IMO, that last being 2 groups in parallel, each a pair of 4P5S "sub-packs" connected in series to get to 10S, for example making them easy to recharge with a hobby charger.

As would also be accomplished with just 2S(8P5S), but without the flexibility and redundancy of being able to use just half the bank at a time.

 

[newb123]

so a 8P10S would be a total of 80 cells?

what I did is use total 20 cells, each was still good at over 4 volts, then put 2 each in parallel..like this


++
--
++
--
++
--
++
--
++
--

then soldered the pairs together on tops and bottoms, then a thick jumper cross ways on top + to - then bottom - to + etc, snaking across the pack.
of course spot welder/nickel tabs are best but I dont have one. seems to have worked out ok, limited time spent with the iron on the cells, used lots of flux and made a square tip with the cells. Then put the BMS sense wires to each cell.

but if I went like this
+ + + + + + + +
_ _ _ _ _ _ _ _

etc.....

for a total of 80 cells in 10S config, would that work?

for the present 10S2P pack, I have a relay and code for the arduino so when one pack is under 32V, the nano switches the relay and I can use another pack (not built yet) ,but just wondering for a new pack build, how far I can push it. 80 cells is allot of capacity and range!

 

[john61ct]

Yes 80 total.

as stated 8P groups first is best, then 10 in series

paralleling 4 of your 2P strings, not as good.

No need to switch from one to the other, just use all at once with Ah counting "fuel gauge" to show status.

Beware scrap cells!

 

[newb123]

this would be what the configuration looks like?

the side drawing is just the cells from a side view in series

 

[john61ct]

I don't know how to explain it more simply.

Parallel is +to+ & -to-, to 8 of those, group is 8*mAh capacity 1S voltage.

Connect 10 of these these together, series means +to- from opposite ends, Ah capacity stays the same, 10S voltage.

Furthermost apart +/- become pack leads.

Maybe someone else with drawing ability can help, yours did nothing for me.

 

[Voltron]

These are plucked from the sphere, so sorry if I don't give proper name credit.

The first one shows basically what you're talking about... A single row of cells, paralleled together, then hooked to the next row.


The same kind of thing, but soldered, with the red lines showing where somebody thinks there should be more series connections.



Better, full current carrying, spot welded, with single rows of cells.

 

[valueseek]

John61ct
I don't think you mentioned why its a bad idea so ill try to summarize. 4(10s2p) has more range but the same charge and discharge current rating of 1(10s2p) because like john said, the sub pack parallel strings are not connected and can become imbalanced leading to one or more of the batteries turning off without you knowing it.

 

[newb123]

yeah thats basically what I was saying; 8 cells in parallel with all the (+) together and (-) together then series those going across and bottom (I just think of it like snaking through from top to bottom across the cells) careful not to connect any one pair together, your pic seems to show , 6 or 8 in series whereas mine would have 10S 8P

on one pic you referenced (pink cells) they are showing solder wire for the parallel strings, I used insulated wire, any benefit with just going with solder? seems more likely to be exposed to a short. those are some big blobs of solder on top of those cells, looks like he held the iron there quite some time.

a 2P10S is already done and being used but the range sucks. Using a printer power pack 18 V and boost CC/CV converter to charge

In my drawing I got lazy and didnt show the cross or parrallel connections.

what does this 8mAH value mean? That means 8 mAh per series string? Where did that value come from?

group is 8*mAh capacity 1S voltage

 

[john61ct]

what does this 8mAH value mean? That means 8 mAh per series string? Where did that value come from?

group is 8*mAh capacity 1S voltage

"8 times the cell mAh capacity"

So with 6000mAh cells, each 8P group is 48Ah @1S voltage, and the whole pack is rated 48 @10S voltage

 

[newb123]

To break the 48Ah into a more realistic value, would I take the average amperage used (for ex. 5Ah) and divide that into 48 to get the avg Ah im using?

 

[Voltron]

Using 5 amps isn't the same thing as amp hours... .

Re the one picture, I think they're actually uninsulated solid copper wire, not solder. But yes, that's some big blobs all over.

 

[Voltron]

You can buy RC wattmeters pretty cheap that will log how much actual watthours you're getting out of a pack...

 

[john61ct]

To break the 48Ah into a more realistic value, would I take the average amperage used (for ex. 5Ah) and divide that into 48 to get the avg Ah im using?

No idea what you're trying to say.

Ah is capacity.

What is the actual capacity in mAh of each cell?

2P you multiply by 2, so if 3200mA, the bank is 6.4Ah

for 8P it's 25.6Ah

Your amps **usage** goes up and down just like watts (A = W ÷ V)

Ah per trip would be a totalized count of energy used over that time period. (Ah = Wh ÷ avgV)

 

[newb123]

What is the actual capacity in mAh of each cell?

yes something like that, apologies for my newbie confusion

was thinking one cell has a rating for ex. 6000 mAH that means that cell can deliver 6000 mA for one hour correct?

and in parrallel, those ratings would double for each cell in parallel, correct?

so if the pack is rated at 48Ah, (10 sets of 8 in parallel, those all in series) what does this value actually represent? I could get 48 amps out continuously for one hour?

Just looking for a useful value to look at to determine the range I can guesstimate.

@voltron,

I have that wattmeter connected on my bike.

log how much actual watthours you're getting out of a pack...

so your saying, once I get down to, for ex. 28 V, I should look at the wattmetter and see the watthours used to get an estimate of what my pack is capable of?

 

[Voltron]

Exactly... I wouldn't go that low on voltage on a 10s though.

I think you can run it on the charger also and see how much you're putting in...

 

[john61ct]

How can you not know your cell capacity? I an asking for that actual information, should be printed on the cells, listed in the datasheet.

And for 10S, stop discharging at 33V resting voltage if you care about longevity, maybe 32V at high C-rates.

30V is drop-dead danger zone

 

[newb123]

I found this spec sheet after cross ref some fine print I found on the cell
R1112 SANYO (referenced to 18650 cells)
so each cell is rated at 2250 mAh, max charging current 1.5 A, max discharge current 4.30 A?

so to get the max rated amp hour, I would do this 8 x 2250 x 10?
Also what is .2CA discharge? That seems like a very fast C rate, if thats what that is

 

[Voltron]

The ratings aren't a direct indicator of how much you will get, as it changes with the discharge rate. What that is saying is it holds roughly 2.2ah, if you take the electricity out at .2 c discharge rate. It doesn't mean it you put 10 together for 22ah that you can pull 22 amps out for an hour.

Its also just the P number times the ah of a cell. So 8 in parallel would be 17.6ah, then the series gives you the voltage, so if you have 10 in series, it would be 42v at full charge, but still 17.6 ah. If you go by that chart, the the max amps you should pull (different than the amp hours) is 4.3 times 8=34.4 amps. Normal discharge would be 17.2 amps (2.15 times 8 )

But you will get different amounts of usable amp hours out of the pack at the different rates as some will get wasted on heating. Thats why the data chart is just a rough guideline.

 

[newb123]

Thanks for your reply, that clears it up a bit

I noticed that the watt hour on the meter read 120 Wh after a short ride

I was dropping about a volt per km

what does this 120 Wh mean or whats its useful interpretation?

ok looked it up. 42 Volt x 17.6 Ah = 739 wh

but how do I figure how many watts Im using per km?

also, the .2 discharge rate, isnt that a very fast discharge?

 

[john61ct]

The 0.2C rate has nothing to do with any usage or charging rating, just a (largely irrelevant) detail about how they arrived at the capacity rating.

To do a comparable CC load discharge test, you find the actual capacity at the same C-rate, so e.g. if 12Ah, draw 2.4A from 100% Full, down to 2.99V over five hours.

 

[john61ct]

10S is 36V in use (40.5 - 42.0V charging)

So 17.6Ah @ 36V would be ~630Wh total capacity.

If a short ride uses 120Wh, pack supports less than 5x that distance in total.

You really are better off sticking to Ah units in these calculations, Wh varies a lot more depending on discharge rate.

I was dropping about a volt per km

A very bad sign. Either something is causing crazy high consumption, or that pack just not support the load, voltage should not drop that fast.

 

[john61ct]

Pretty sure most packs designed for 36V are 12S,

to support 43 - 44V in use, charge at 49 - 50.4Vch max.

 

[newb123]

A very bad sign. Either something is causing crazy high consumption, or that pack just not support the load, voltage should not drop that fast.

So I should add 4 more cells (2 parrallel sets @ 4.2 each) to the 10S ? that would bring the total charge, off the charger, to 48V

I will try it and see if it increases capacity/range

but I will have to get another sense wire harnness, or just solder directly on the pins the additional sense wires, as my current one doesnt have enough wires to go out to 48V The BMS data says it supports up to 48 though

 

[john61ct]

And what problem are you trying to solve by going to a higher voltage?

My comments / questions are addressed to problems with your existing setup, nothing to do with going to a higher spec voltage?

And you can't just arbitrarily increase your voltage without being intimate with your motor and controller specs, easy to just burn one or both blindly experimenting.

Going to "48V" nominal is usually 13S with LI chemistries, 52-54.6V charging.

So from 10S means adding 3 per "P", in a 2P pack would be 6, but weren't you talking about an 8P pack for greater amps output (at the same C-rate) and also extended range?