24v 12Ah NiMH pack

[Spence /1/]

Alright, so I am making a weak electric bike using a 24v 12A motor that has an actual output of about 220W. I want to use NiMH batteries as they are cheap, have a good capacity and are perfect for this project. I would only discharge the pack at 1C (using 12A constantly).

What I am wondering is how I go about charging this pack and regulate the voltage.

The pack consists of 88 AA cells where there are 4 rows of 22 cells to give an initially charged pack of 30.8V that after 5% DOD will be at a total voltage of 26.4V until it hits about 90% DOD where the total pack voltage will hit 22V, which is when I would stop discharging the pack (this data is from a 1C discharge curve).

 

[amberwolf]

you'll need a nimh-specific charger, that can hook up to the thermal sensor in the nimh pack.

nimh cells that can actually output 12a are not the cheap common kind. aa cells wont do what you want. even f cells may not.

you cannot directly paralell nimh without risk of fire. if you don't research how nimh works in detail befor you build your pack, you won't have one for long


there are varous threads on es that you can find about nimh and what you can and cant do with it, and safe ways tou use it.

 

[Spence /1/]

a 3000 mAh aa NiMH cell can't put out 3A of current continuously?

how do you wire NiMH cells in parallel then?

 

[amberwolf]

you cannot directly paralell nimh without risk of fire. if you don't research how nimh works in detail befor you build your pack, you won't have one for long


there are varous threads on es that you can find about nimh and what you can and cant do with it, and safe ways tou use it.

sorry broken hand keeping me from usual lengthy replies.

 

[Sunder]

I built a 48v nimh pack using 40x 10ah 'D' cells. The bike worked fine when on the stand, but once I put load on it, serious voltage sag. They were supposed to be able to deliver 1c continuous and 2c for up to 30 seconds... but I think the dozens of tiny links between each cell was just causing too much resistance.

Not saying you can't do it, but it might be a bigger challenge than you think.

 

[Ykick]

NiMh sucks IMO. I ran 13Ah F cells years ago and they're barely capable of 1C output rate. Ask for 2C or more and you'll be lucky to get 50% rated capacity from 'em. Paralleling individual cells will only be asking for more problems.

But the thing that really bugged me is/was self-discharge. These things lose 10% charge in 24-48 hours and more as the days drag on. Unlike Lipo, they don't seem to mind hitting 0 Volts and they store quite well for long periods of time in nearly full discharged condition. But, they will need a few cycles to "wake up" after periods of inactivity.

Charging curve is weird and you do need to implement some type of temperature sensing. Energy density in real world application isn't much better than SLA. Maybe 50% weight savings compared to similar capacity Lead and just about as saggy.

In this day and age you can do a lot better than NiMh for an eBike...

 

[dogman dan]

It's just a bad idea. if nothing else, because you have to make strings of 1p out of the cells, and then charge each 1p string seperately. You can't paralell them and then safely charge.

So it would be a pain even if you did get it to work under load.

Same thing goes for nicads, but even the 5c F cells have been dropped now, because lithium is so much better.

 

[bobale]

I want to use NiMH batteries as they are cheap, have a good capacity and are perfect for this project.

So are LiPo. Why not save yourself a bother, and build a pack from Turnigy 4S hardpacks? It'll be much simpler, lighter, and really not that much expensive.

 

[El_Steak]

Same here, if you really want to use NiMH, stick with high discharge D or F cells in a 1p20s config. 4p AA are a bad idea, you'll have to charge the 4 strings separately so you'll need 4 NIMH chargers. NEVER charge them in parallel or you'll get fireworks. Get a weak cell in there and it will reverse charge and go bang. (there's a video from Justin on ES showing just how dangerous NiMH/NiCad can be when it goes bad - those things will punch holes in walls (or you) when they go ballistic).

Look at Lipo or Life, I'm sure you can get a pack close in price with no sag, light weight, easy charging and long life cycle. If you're looking for turn key, a Ping 24V 10ah pack with BMS and charger will only cost you 200$. If you're more of a DIY guy, look at the turnigy Lipo.

I used a 48V 8.5ah NiMH pack for a while a few years ago. It was made up of high quality SAFT D cells and worked ok on a 15A controller. Still, the cells would get hot enough to melt their plastic wrappers and they ended up leaking and dying after 2 seasons. Not worth it in this day and age IMHO.

 

[El_Steak]

The links to the justin videos I mentionned are here:

http://endless-sphere.com/forums/viewtopic.php?f=14&t=23105&p=339784&hilit=+justin+battery+video#p336200

Video #2 has a good explanation on the balancing / cell reversal issues you might have
Video #3 has some nice pictures of NiMH pack destruction

And a very informative PDF version of the presentation:

http://www.ebikes.ca/images/Lead_Free_Talk.pdf

 

[Ypedal]

This goes back to 2007
http://ypedal.com/Chaos2.htm

but yeah, ++ to everything said above.. forget AA cells.. dont believe the 3ah claims, it's rubbish, if you can make due with single string of D or F cells, then fine but do not bother with AA stuff..

Spring holders will not work either, it's been tried before, the cells need to be spot welded tabbed together..

You would be better off with a pair of 12v 7ah SLA's.

 

[El_Steak]

I dug out some old pictures of my NiMH days.

This first attempt was with PVC tubes and caps compressing the cells together. Complete failure, a lot of resistance in the connections made the whole pack heat up real quick. Looked like a bomb

This was my second design, here the cells where soldered end-to-end RC style. Also looked like a bomb
This worked pretty well, but by then the cells were damaged and it only lasted for a season before I had to retire it. Moved to lithium and never looked back.

 

[999zip999]

Love those El-steak how do they get thru the airport ? Looks harmless just a bike battery. How could this be a problem ?

 

[El_Steak]

Love those El-steak how do they get thru the airport ? Looks harmless just a bike battery. How could this be a problem ?

Those had such a poor capacity that I had to charge at work during the day. I remember getting a few worried looks from my co-workers when I first brought them in my office to charge

 

[chilledoutuk]

I am sorry but i don't agree that NIMH is a horrible battery technology for evbikes.

Its about using the right ones for the job a lot of the D cell NIMH batterys were not designed for high drain use and thus will
perform badly.

Most sub-c cells which are purposely designed for high drain situations like RC cars.

I am currently using 3x 12v VapexTech 4600mah NIMH in series with 9c kit with 27a controller.

http://www.vapextech.com/pdf/down/VP4600SC%20SPEC.pdf

I paid 100 pounds for three of these packs which compared to 73 pounds for 3 yuasa SLA batterys i though was much bettery vaule

My pack weighs 2.1kg compared to what would have been around 12KG for the SLA cells.

The cells are performing very well and when i tested the range i got 8.5 miles without pedaling at all.

The cells get the slightest warmth in them after a run if used from when the pack is cold in fact i have been having problems keeping the cells warm enough to get
good performance out of them. I wrapped them in a towel and put them in my backpack once and they were much more powerful and afterwards they were warm to touch.

I think they key to using large nickle based battery packs is to regularly balance charge the cells after a quick charge and run a LVC of around 1v per a cell to avoid cell reversals.

However if you think you have over discharged a pack and maybe reversed a cell a 0.1c balancing charge until all the cells are nice and warm (usually 15 hours) should recover any reversed cells and bring the pack back into balance If there has been signs of the cell venting then it may have lost capacity permanently and probably will likely reverse again although i haven't had this problem before.

I love high quality lipo and a123 dont get me wrong if you can afford High Quality lipo or even better A123 cells then that is the route to go. But too many of my china lipo batterys just dont last. I have a kokam pack that i have had for over 3 years and its never shown any signs of problems and in perfect balance yet all the china lipo packs i have had have swelled up and look like there about to explode.

I think the problem with parrelling nimh cells is you need to have a different charging method than Delta V. I think if you did some sort of CVCC buy first testing to see what voltage they cells get to when they become full and then basing the final voltage on that would work better and be much safer.

 

[amberwolf]

I think the problem with parrelling nimh cells is you need to have a different charging method than Delta V. I think if you did some sort of CVCC buy first testing to see what voltage they cells get to when they become full and then basing the final voltage on that would work better and be much safer.

if paarleled, cell/string that reaches full first DROPS in v and heats up to dissipate that energy

then all chargig current PLUS current from all other cells/strings paralel to it flows INTO IT, heating it further and overcharging it

can leak, burst, explode, or catch fire or multiple of those


so CANNOT direcvtly parallel nimh

 

[chilledoutuk]

I think your mistaking a limitation in NIMH for what really is a limitation in how we are used to charging NIMH.

Peak charging was developed to get every last bit of capacity out of a cell and took advantage of the fact that Nickle based cells can take being overcharged without bursting into
flames.

If you were to take a cell test them to see what voltage they peaked at and then parallel a bunch of the same cells and then charge to say 0.01v less than that voltage with a cvcc method then there would be no voltage drop as current would taper off before overcharging could make the voltage drop and one cell could suck all the power.

I currently charge my NIMH cells in parallel using zero peak detection with a trickle charge afterwards to top up all the cells and balance them nicely.
I would prefer to charge them using cccv but i dont have the equipment to do that at the moment.

Any high quality Lithium tech battery is great but the cheaper ones in my opinion are just a waste of valuable lithium and i would be very surprised if they could ever last as long as a Kokam or similar quality pack. A123 well there amazing but also expensive.

I wish sanyo made larger cell eneloop cells as it could be an interesting alternative to lipo with a really good cycle life.

 

[amberwolf]

I think your mistaking a limitation in NIMH for what really is a limitation in how we are used to charging NIMH.

Peak charging was developed to get every last bit of capacity out of a cell and took advantage of the fact that Nickle based cells can take being overcharged without bursting into
flames.

true; but since cells degrade unequally over time you'll have to test every cell (or string) periodicsally so you knowthe lowest ca[pascity one, and use that as the limit. if you don't, it is possible for a cell thast reaches a voltage it can no longer output much current to begin draining the paralell cells. if there is enough capacity difference you might begin heating that cell enough to damage it (since it may already be hot from discharge). if it isin the core of the pack where it cant be cooled, fire risk still exists.

the same is true of charging, if that cell gets full before others, it's ri drops and more current gies thru it than others, and it heats more. if capacity diff is high enough, and it heats long enough....


when all cells are matche it is ok, but as they age it coiud be a problem,

the longer the series string and the more parallekled groups, the easier the problem is to create,

alsp if a cell ever grows a shoirting dendrite in it, then it doesnt matter how msany matchings youve done. its rarer in nimh thn nicd but still happens.

using series diodes at the end of every string will fix all of these potential issues except still wont allow typical nimh chargers to be used. (unless you use the pre-diode poiunt as charger point, and charge each string separaytely)

 

[Spence /1/]

Couldn't one make a BMS that could charge these cells to "their full potential" and discharge each of them at an even rate?

I was just wondering, also these NiMH cells are from china.... so maybe ill just get something american... Thinking of A123, but I wouldn't mind using lipo either what are some trusted battery manufacturers? (I don't have a sh*t load of cash )

 

[El_Steak]

I think that for your modest requirements (220W), the 24V10Ah Ping pack I mentionned on the first page is the best bet. It's 2ah short from your 24V12ah pack, but in reality I'm sure you wouldn't get anything near 12AH out of those AA NiMH cells, even less out of SLAs.

If you have 70$ more, you can fork for the Ping 24V15ah pack which will give you a lot of range at 220w and will give you a better max continuous discharge rate should you ever upgrade to a more powerful setup.

Those packs include everything you need, battery, BMS and charger in a turn-key solution.

The other option is LiPo, where you can built pretty much any pack configuration you fancy. You'll have to devise your own wiring, cell monitoring and charging solution for those however. Lots of information on ES on that subject, I suggest you read a lot to get comfortable with the ups and downs of LiPo before you go that way. The most popular source of Lipo with ES members is HobbyKing.

 

[Spence /1/]

I want to use NiMH batteries as they are cheap, have a good capacity and are perfect for this project.

So are LiPo. Why not save yourself a bother, and build a pack from Turnigy 4S hardpacks? It'll be much simpler, lighter, and really not that much expensive.

These look really good, Now when I hook two rows of two together to give me 2 cells in series a row and 2 parallel, How do I go about charging them? and what BMS could I use?

 

[bobale]

In the e-bike world that is called 8S2P. Charging that pack would be very simple. There is ample supply of 8S balancing chargers for ocassional balancing, and for everyday use you could buy Meanwell S-350-36 PSU and charge that pack in under an hour .

 

[Spence /1/]

its 8S because of the 8 individual 3.7V cells in the packs right?

 

[Ypedal]

8S means 8 cells in SERIES.
2P means 2 cells in PARALLEL

A pack like this has a total of 16 individual cells,

 

[El_Steak]

its 8S because of the 8 individual 3.7V cells in the packs right?

I don't know if you're limited to 24V with your controller/motor setup, but 8s would give you about 30V nominal and 34V fully charged.

You could get closer to 24V with a really simple 6s2p pack. This would give you 22.2V nominal and 25V fully charged.

The best deal would be the turnigy 6s 5ah 20c pack. 2 of those would give you a 22.2V, 10ah pack for 90$. Those are great packs, with a max discharge rate of 100A !!! and true to their capacity ratings (often more than 5Ah each). You can get a dirt cheap 50W charger like that and a dirt cheap power supply like that and you're all set to go for just over 100$ (plus shipping).

I wouldn't worry too much about the slight voltage diff from 24V to 22.2V since lipo will hold its voltage much better than NiMH under load, so you'd get the same if not better performance. The 3.7V nominal is under-rated for lipo, as you'll get most of your watts between 4V and 3.8V. 3.7V is really close to empty, not nominal.

Bonus, this will only weight a bit over 1.5KG (3.5 pounds) !!!

Just make sure you never go under 3.5V on your packs or you risk them getting puffy and then being good for the garbage bin. There are low voltage cutoff boards sold by ES members that can protect you from that for cheap. Just search for Lipo LVC. I think Method, Geoff and Ggoodrum sell them.