Re-celled battery won't work ?

[chazpope]

Hi

I re-celled my 24V NiMh battery with 20 of these 10000mAh cells - (someone here found them on ebay)


However the new battery won't work on the bike - it shows full voltage but when I turn the motor on the voltage drops immediately so you get a flutter effect and it's not usable.

When discharged with a car bulb (5 A) it goes down to below 20V in just about 10 minutes, also when charging (< 2A accoriding to charger) it takes only about 2h to cut off - so I get about 1Ah (not 10Ah) from it.

I repeated the discharge/charge exercise twice and it made no difference.


Any ideas ?


p.s. I have done this before and that battery still works very well - but I used cells from maplins

 

[Deleted member 4366]

It could be that the cells are out of balance. How does you system balance them?

 

[chazpope]

no idea about 'balance' - does NiMh need any ? I have 2 chargers of slightly different design and as far as I can see they only have a temperature probe and that's that.

 

[flecc]

The cells need to be quite evenly balanced in terms of voltage in the first instance before assembly. That's why it's best to overbuy and discard any that don't measure up to standard. The charger and battery have no balancing mechanism, just cutoff by temperature at end of charge.

Don't discharge using a bulb ever, if the voltage is taken too low, one or more cells can be destroyed if out of balance with the rest.

It sounds as though the charger cutoff is triggering far too early, so the thermistor in the battery may have failed. Recheck whether you've wired the thermistor back in correctly and there's no break in the connections to it.

What bike is it by the way?

 

[chazpope]

so what can I do now - how do I ballance the cells ? these were new cells they should have been the same

I did measure the voltage of the 4 x 5 packs before connecting them it was the same within 1%

it is a powacycle cambridge, I am sure the thermistor works as when the battery is hot the crarger flashes as it should. does it cutoff just by temperature - I would have thought it monitors the voltage drop too

 

[Deleted member 4366]

There's not much difference in voltage of a Nihm cell between being 3/4 charged and 1/4 charged. To eliminate the possibility that the cells are out of balance, you could put a load on the battery and them measure the voltage on each cell. If they're more or less equal, then the cells are not capable of providing the current you want for whatever reason. If the voltages are all over the place under load, you need to get some charge into the lower ones.

 

[flecc]

The cells being new doesn't necessarily mean they are balanced and poor performing new cells are not unusual. One reason why manufacturers were so keen to abandon NiMh was the balance problem, the best batteries only made with cells from single production batches and then individually matched to make up packs. All this was time consuming and costly, and not doing it meant too many failure returns.

The cutoff I referred to was when charging, with NiMh that is by temperature when the battery is full. This is because the voltage rise at end of charge is too gradual to get a clear cut off point, but the temperature gradient is quite steep at the end so enabling a precise cutoff point. These chargers usually have a secondary safety cutoff by voltage rise as well.

If the charger you are using is the Powacycle one, you should have gained quite a fair charge after two hours, at least half charge. The fact that the battery is getting hot when the charge cuts off at about two hours indicates the charger and thermistor are probably ok, the problem possibly being one or more failed cells.

When you think it's fully charged, measure the individual cell voltages to check they are all nearly the same. Then switch on and operate the bike, reopen the battery and check the cells voltages again. Also check for any hot cells straight after the charge.

You can't really balance any cell that falls short on voltage after charging, just chargng that one until it matches won't enable it to retain that match in use. That's why overbuying is best so that any duds can be discarded, especially with unknown types like these ebay ones.

N.B. Crossed with d8veh post.

 

[chazpope]

Shouldn't it be charging for 5h at least ? I use the powacycle charger - it is rated at 1.8A - so for 10Ah it shouldn't be as short as 2h

I will do some of these tests tonight, it's not going to be easy cause you can't just measure voltages when the cells are packed in a metal tube. I think the first thing I will do is separate the thermistor from the battery and see what happens

 

[flecc]

Yes, but after two hours the half charge should run the bike for a long time, not refuse to run.

The Powacycle charger probably won't even start charging with the thermistor out of circuit, their 2007 one wouldn't and most others won't either on safety grounds.

The Powacycle battery cell layout wasn't in metal tubes, here's a Powacycle one that I used to recell another battery make, which is why I thought you could measure the cells individually:

 

[chazpope]

I see - Click here to see it - you must have seen another version

 

[flecc]

I see - Click here to see it - you must have seen another version

Yes, different bike model the Cambridge, with a tall slim battery box. I used to use the Salisbury/Windsor model's battery (£99 at the time) to scavenge the cellpack.

 

[piotrmacheta]

Another thing worth mentioning is that NiMH batteries need a few full cycles before they give full capacity but with cheap ebay batteries you cannot be sure of the quality. I once bought a subC racing pack declared as 6.8Ah which turned out to be only 2Ah - I got my money back via Paypal. Also I have some high quality 10Ahr Ansmann D cells and even they aren't very good with current delivery and their capacity at more than a few amps drops significantly so they only give 6.5Ah in real conditions use. So unfortunately these issues coupled with the cell balance issues may mean it doesn't work very well. The way I did it was to make up the pack and charge each cell individually then discharge the pack (but not fully) and recharge as a whole pack.

 

[chazpope]

for now without breaking up the battery I found that :
- charger charges with a constant 1.8 A
- 4 groups of 5 batteries behave consistently voltage-wise during chraging
- after just under 2h the whole battery heats up a little and charging stops
so it behaves like a smaller battery - I will do more tests tonight.

question remains why it would not maintain voltage under load even when full, it kind of drops to 22 v very quickly and I haven't yet found out is it one cell doing it or all. Is there such a thing as 'current rating' i.e where it only delivers up to x current ? If you discharge a fresh cell at say 0.5C - what should the voltage be

 

[flecc]

That sudden drop in voltage and the fact that it's heating up noticeably in only 2 hours both indicate one or more faulty cells is letting the whole pack down, a common NiMh problem especially with recelling. It shouldn't show enough heat to cut out until at least 4 hours from empty.

However, it may already be half charged after your charging, so the 2 hours will be correct then, once again indicating a cell or more collapsing under load when it fails to deliver.

Charge rates and voltages then are a red herring, I'm sure this is a dud cells problem. I've also seen the same problems as piotrmacheta above, including the dud Ansmann supposedly quality cells.

 

[Deleted member 4366]

According to the advert. the cells are rated at 3C, which means 30 amps. You need to get the battery under load, while you measure the individual cells to see, which ones are faulty.

 

[chazpope]

Ok so these nicely done heat-shrunk battery packs - I broke them up to investigate cell by cell.

To start with - with the temperature probe not touching the cells the charger still switches off (turns green) - probably because it is voltage drop sensing. This happens after about 2 hours anyway, thereafter by 10 mins so it's not timed. So I cannot force a full charge even if I wanted to.

Second - after the 2 hour charging if I apply a 4.5 Amp load (.45C) - after 10 mins some (3-4) cells lose all voltage and go into reverse.

Also initially some cells did get hot while charging. I discharged them idividually and they do not go hot anymore.

Not sure what to do next - I can't charge cells one by one to test and they won't charge for long enough as a whole, also some cells seem to be *****d. I suppose I can try chaging them with my lacrosse charger with some contraption, not sure if it goes as far as 10ah though. Should I buy some more or try to reject the whole lot - (this may be impossible - Hong Kong)

I must say I had none of this crap with the maplins cells - 3 years ago I just put them in and now after 1000 charges they still work as 80% new.

I want to find a solution though as there is no other option

 

[flecc]

As I thought, some cells below par, very common. It may be worth the risk of buying to replace the 3 or 4 cells that you see are poor, but to be sure it's best to buy two over since one or two of those replacements may be a bit down. It's a pain, but necessary with NiMh from unknown sources. You were obviously lucky with the Maplin cells which probably through good fortune came from the same production batch and therefore had better cell to cell consistency.

 

[Deleted member 4366]

As I thought, some cells below par, very common. It may be worth the risk of buying to replace the 3 or 4 cells that you see are poor, but to be sure it's best to buy two over since one or two of those replacements may be a bit down. It's a pain, but necessary with NiMh from unknown sources. You were obviously lucky with the Maplin cells which probably through good fortune came from the same production batch and therefore had better cell to cell consistency.

I think, before chucking them, I'd have a go at charging them individually. You've nothing to lose. You could use any nimh charger and just run some wires from the contacts to each cell. Then you'll have the whole pack fully charged. If after that it still gives problems, then it's time to think about chucking something.

 

[chazpope]

believe me I would not chuck 50 quid's worth of cella just like that I was wishfully thinking about 'money back option'

I have fiddled a bit more and yes there are maybe 5 in the pack that behave differently. During charging these show a higher voltage (by about 0.15V) than 'healthy' cells. This I guess causes the charger to reach 33.0V and cut off prematurely after about 1.5h. Then, during discharge - these cell seem to hold much less charge and get depleted very quickly. I will try talking to shop about this - see where I will get.

 

[chazpope]

And where are they now ?

What I did was charge and test them individually on my Lacrosse charger at 1.8 A (same as bike) and compare with the old cells I was replacing. I found the 'new' ones were way way below on performance - about 3Ah capacity at a considerable lower voltage (for the 'best' in the pack). So much for manuf claims of 10Ah and 30Amp currents.

I guess after seeing that their weight was about half the weight of similar cells I should have known not to buy them. But I had my money back in the end - no complaint there.

Question now is - where do I get 'good ones' from ?