Recelling my Giant Lafree battery

[tangent]

I am investigating options for replacing the NiMH cells in my Giant Lafree batteries.

I have worked out that I can fit 4 12V sub-C packs (e.g. like those available here component-shop) inside the case in a 2s2p arrangement. Has anyone tried this?

If I do this with fresh identical packs, would I need to incorporate diodes to prevent driving a current from one of the parallel packs through the other?

Should have read more of this excellent forum before posting this! From what I have read, it looks like I might be better off with 8 6V sub-C packs, with pairs of 2 C packs in parallel and then all 4 pairs in series. Would I get away without the diodes if I did this?

 

[PaulC]

Alternative to recelling:
EbikesNW: Parts and Accessories

Regards
Paul

 

[rog_london]

I am investigating options for replacing the NiMH cells in my Giant Lafree batteries.

I have worked out that I can fit 4 12V sub-C packs (e.g. like those available here component-shop) inside the case in a 2s2p arrangement. Has anyone tried this?

If I do this with fresh identical packs, would I need to incorporate diodes to prevent driving a current from one of the parallel packs through the other?

Should have read more of this excellent forum before posting this! From what I have read, it looks like I might be better off with 8 6V sub-C packs, with pairs of 2 C packs in parallel and then all 4 pairs in series. Would I get away without the diodes if I did this?

I believe you always need diodes in any battery technology other than lead acid. They all have tiny differences in voltage between cells and low internal resistance which promotes self-discharge back and forth between parallel banks if you don't have diodes.

Of course diodes are implicit in BMS systems.

Lead acid batteries don't need diodes because in all cases the charging voltage is significantly higher than the discharge or standing voltage, so self-discharge is not a problem in parallel banks. It's also, I suspect, a key to the fact that lead acid batteries have a long shelf life between charges - the other technologies probably always self-discharge internally to some extent between different parts of the same cell, although I understand some methods of construction make this much less of a problem.

Rog.

 

[Andrew harvey]

Try Sakura batteries;
Tel: 020 8965 4567

Unit 1, 10-24, Standard Rd, London, NW10 6EU.

I've been told they may be able to help.

 

[NRG]

I believe you always need diodes in any battery technology other than lead acid. They all have tiny differences in voltage between cells and low internal resistance which promotes self-discharge back and forth between parallel banks if you don't have diodes.

Of course diodes are implicit in BMS systems.

Lead acid batteries don't need diodes because in all cases the charging voltage is significantly higher than the discharge or standing voltage, so self-discharge is not a problem in parallel banks. It's also, I suspect, a key to the fact that lead acid batteries have a long shelf life between charges - the other technologies probably always self-discharge internally to some extent between different parts of the same cell, although I understand some methods of construction make this much less of a problem.

Rog.

I read this the other day and its been playing on my mind! I've been looking at TI application ccts for battery monitoring and balancing but there are no diodes used or or mention of them in the diagrams I've looked at. Diodes would introduce an unnecessary voltage drop and would not be an ideal way to balance cells...I also read a lot on ES and there's some very bright guys on there and one of the recommendations, if using parallel packs, is to connect up at the cell level so the pack can (to a limited degree) perform some self balancing. No mention of diodes. I'm curious.....

 

[rog_london]

I read this the other day and its been playing on my mind! I've been looking at TI application ccts for battery monitoring and balancing but there are no diodes used or or mention of them in the diagrams I've looked at. Diodes would introduce an unnecessary voltage drop and would not be an ideal way to balance cells...I also read a lot on ES and there's some very bright guys on there and one of the recommendations, if using parallel packs, is to connect up at the cell level so the pack can (to a limited degree) perform some self balancing. No mention of diodes. I'm curious.....

It's not to do with balancing. It's a problem with connecting different battery packs in parallel and leaving them to stand for a while. Due to minor differences in voltage between (even apparently identical) cells you get current flow between them if you don't have blocking diodes. Note that of course this relates to *parallel* connected batteries. Most e-bikes just have one series chain of cells, and then of course it doesn't apply.

You need one diode per battery for discharging (usually a heavy duty job in this sort of application) and one diode per battery (lighter duty usually) for charging. It's best on the discharge side to either use low forward voltage diodes or one of the special circuits which I think uses a MOSFET to give a very low forward drop.

I seem to remember last year one of the members was producing such a circuit for sale - but I can't remember any more than that.

You really only need the diodes when the batteries are not in use, and you could simply disconnect the batteries from each other the rest of the time and dispense with the diodes.

One other thing - in practice it's difficult to charge parallel-connected batteries in most technologies evenly, as their terminal voltages don't behave sensibly - if I remember correctly, NiCds and NiMhs (for example) actually drop their terminal voltage slightly when they're full (apart from getting pretty warm if the charging current is at all substantial) and so the first battery to get there would 'hog' the charger.

If you have a BMS setup (as in all Li-Ion and associated technology batteries) then probably you don't need to do anything else on the charging side. A BMS will turn off its own battery when it's full, and that resolves the problem. In such cases with parallel batteries you might still need diodes on the discharge side though, as a BMS usually does its work only on the charging cycle (though there are exceptions) and it's not involved in the heavy-duty discharge side at all - the bike's controller is what usually cuts off as the overall voltage drops.

Rog.

 

[flecc]

I seem to remember last year one of the members was producing such a circuit for sale - but I can't remember any more than that.

Yes, Tiberius (Nick) has made up some "ideal diodes". All the details are on the link below:

Ideal Diodes in the Classified Section
.

 

[NRG]

It's not to do with balancing. It's a problem with connecting different battery packs in parallel and leaving them to stand for a while. Due to minor differences in voltage between (even apparently identical) cells you get current flow between them if you don't have blocking diodes. Note that of course this relates to *parallel* connected batteries. Most e-bikes just have one series chain of cells, and then of course it doesn't apply.

Hi Rog, I would think most E-Bikes would have a number of series connected cells connected in parallel to get the Ah rating up.

However, I fully understand what you mean if the need is for paralleling dissimilar chemistry batteries and Nick's ideal diodes would be used in that instance. However, with batteries of the same chemistry and rating diodes aren't needed. I've checked this for myself and IME this works out in practice with RC Lipo type cells very well.

You need one diode per battery for discharging (usually a heavy duty job in this sort of application) and one diode per battery (lighter duty usually) for charging. It's best on the discharge side to either use low forward current diodes or one of the special circuits which I think uses a MOSFET to give a very low forward drop.

I seem to remember last year one of the members was producing such a circuit for sale - but I can't remember any more than that.

Yes these are Nicks diodes as Flecc mentions using a Mosfet with its very low Source to Drain impedance. Certainly a normal diode would not be used in this case (or shouldn't!)

You really only need the diodes when the batteries are not in use, and you could simply disconnect the batteries from each other the rest of the time and dispense with the diodes.

Yes, and this is what I do with my packs but I've also left them interconnected at the cell level and found there is an element of self-balancing, just as the guys on ES said.

One other thing - in practice it's difficult to charge parallel-connected batteries in most technologies evenly, as their terminal voltages don't behave sensibly - if I remember correctly, NiCds and NiMhs (for example) actually drop their terminal voltage slightly when they're full (apart from getting pretty warm if the charging current is at all substantial) and so the first battery to get there would 'hog' the charger.

Indeed, and with Lipo the cell in a series of cells that gets to 4.2v first will 'starve' the other cells leading to an unbalanced pack over a number of cycles if not corrected. A bleed balancer can be used in combination with the charger (may even be an integrated unit) to pull back the wayward cell whilst the others catch up.

If you have a BMS setup (as in all Li-Ion and associated technology batteries) then probably you don't need to do anything else on the charging side. A BMS will turn off its own battery when it's full, and that resolves the problem. In such cases with parallel batteries you might still need diodes on the discharge side though, as a BMS usually does its work only on the charging cycle (though there are exceptions) and it's not involved in the heavy-duty discharge side at all - the bike's controller is what usually cuts off as the overall voltage drops.

Rog.

Agreed but I think only really applicable to different chemistries with different discharge characteristics. Good discussion though, thanks for replying.

 

[flecc]

I would think most E-Bikes would have a number of series connected cells connected in parallel to get the Ah rating up.

Until recently none did, 26 volt batteries using 7 series cells and 37 volt using 10 series cells, each cell being the battery nominal capacity. Even now I only know of three using series/parallel, the first two just the higher capacity option:

Wisper 37 volt 14 Ah, 2P x 10S cells of 7 Ah

eZee 37 volt 14 Ah, 2P x 10S cells of 7 Ah

Panasonic 26 volt 10 Ah, 2P x 7S cells of 5 Ah
.

 

[NRG]

Thanks Flecc, interesting I hadn't realised single cell lithium or Nimh batteries had been available in large Ah capacities...

 

[flecc]

Thanks Flecc, interesting I hadn't realised single cell lithium or Nimh batteries had been available in large Ah capacities...

Yes, they had 10 Ah cells right from the start with cobalt cathodes, later manganese ones. Here's the original eZee lithium manganese battery from 2006 with 10x 3.7 volt 10 Ah cells:

 

[tangent]

I Have found a UK supplier who will make me up a 9Ah 24V NiMH D cell pack in the 7/6/7 configuration required for the Lafree battery case. Cost is £125 including delivery. I will report back on my experience.

I am also going to play around with some 4600 mAh sub-C 12V packs in parallel and serial arrangements, without diodes. According to the toy car/plane lot there is no problem using similar packs in parallel without diodes. However, I have been warned that I should not try to charge these in parallel either with or without diodes - it is dangerous due to the -dV drop off when NiMH cells become fully charged. Essentially, this causes more current to flow through the already charged battery than the parallel uncharged one. As the battery is fully charged and no further chemical reactions can take place to absorb the energy, the energy just comes out as heat. There are a number of examples on the web of what happens when parallel charging is attempted. Sometimes you get away with it, sometimes the cells melt or catch fire.

e.g. Exploding NiMH Batteries

 

[rog_london]

However, I have been warned that I should not try to charge these in parallel either with or without diodes - it is dangerous due to the -dV drop off when NiMH cells become fully charged. Essentially, this causes more current to flow through the already charged battery than the parallel uncharged one. As the battery is fully charged and no further chemical reactions can take place to absorb the energy, the energy just comes out as heat. There are a number of examples on the web of what happens when parallel charging is attempted. Sometimes you get away with it, sometimes the cells melt or catch fire.

e.g. Exploding NiMH Batteries

As I said.....

The slight drop-off in voltage is hard to detect reliably in any case, and it's common for a temperature sensor to be buried in each battery to trigger a BMS or other simpler cut-off mechanism appropriately.

You'll notice also his very pertinent observation about the need to use diodes on parallel-connected packs! One shorted cell, and away she goes....

Rog.

 

[NRG]

Isn't this a particular problem with NiMH though and not a problem with SLA or Lithium?

 

[rog_london]

Isn't this a particular problem with NiMH though and not a problem with SLA or Lithium?

It IS a problem with lithium batteries unless you have a BMS - which usually you do. Lithium batteries are difficult to manage without cell-level supervision. There have been quite a few cases of lithium batteries developing faults which caused alarming high temperatures and occasionally fires - maybe not on current e-bike technologies, but certainly on the smaller consumable models used on laptop computers etc.

As I've said before, it's not usually a requirement with lead acid. Their different charge/discharge characteristics make them excellent at self-balancing in a parallel configuration. You could still get a shorted cell, but generally they don't get abused in that way (i.e. destructive high temperature) because as long as the charger has the appropriate top-out voltage they don't get hot.

I have only seen short-circuited cells in lead acid batteries due to extreme age or mechanical damage (vibration, being dropped). The sulphation which eventually 'wears out' lead acid cells can eventually lead to short circuits. In older technology lead acid batteries (with free acid and the need for distilled water) excessive charging would cause plate material to flake off and build up in the bottom of the battery eventually leading to short circuits.

The larger lead acid batteries are currently carefully manufactured to avoid this likelihood, as the short-circuit current could run into hundreds of amps and cause spectacular meltdown, hot acid everywhere - etc.

So - no diodes needed for SLA or allied technology.

Rog.

 

[tangent]

As I said.....

The slight drop-off in voltage is hard to detect reliably in any case, and it's common for a temperature sensor to be buried in each battery to trigger a BMS or other simpler cut-off mechanism appropriately.

You'll notice also his very pertinent observation about the need to use diodes on parallel-connected packs! One shorted cell, and away she goes....

Rog.

The impression I get from the racing car people is that diodes are not essential for discharging, however, they probably don't leave parallel connected packs lying around for too long. I guess they avoid anything that adds to weight, like diodes/heat sinks. Personally I think it is well worth having some fuses in there at least to guard against shorted cells.

 

[rog_london]

The impression I get from the racing car people is that diodes are not essential for discharging, however, they probably don't leave parallel connected packs lying around for too long. I guess they avoid anything that adds to weight, like diodes/heat sinks. Personally I think it is well worth having some fuses in there at least to guard against shorted cells.

That sounds perfectly reasonable. A low-tech solution is very often the best. As with all these things, there often isn't a definitive answer and the best solution needs to take into account your application.

In the case of e-bikes, there's a strong likelihood that the battery setup will be hanging around for perhaps some days (especially in winter) and without blocking diodes you're likely to find that all the batteries are totally flat the next time you come to use it unless you remember to dismount them when you dismount yourself!

The self-discharge 'thing' is a back-and-forth operation - the one with the lower voltage will swap around with time, and the end result is generally that the lot go flat. It can take a few days or even weeks for that to happen - it's not at all predictable.

I always think about them as behaving like an autumn leaf which falls off a tree - it doesn't fall straight to earth, but does a shimmy from side to side on its way down. That's sort of what the parallel batteries do - but rather more slowly. Well, it works for me, but I'm not sure it's an entirely logical explanation!

Rog.

 

[neil mcfarlane]

re-celling la free

Hi

This is my first post though I've sought advice by email from flecc a few times. Thought I'd try to bother him less, though he seems so prolific on these boards I wouldn't be surprised if he replies to this...

I've looked through a lot of these threads while waiting for my account to be activated, and I was wondering what the current situation was re re-celling a lafree lite battery? (a lot of the posts here go back a few years). Flecc mentioned on his site that powerbike no longer do Ni Mh batteries. Are Ni Mh batteries all pretty much the same? I found these from a supplier in China:

electric bike Nimh rechargeable battery pack products, buy electric bike Nimh rechargeable battery pack products from alibaba.com

and was quoted $125 CIF UK (I don't know what CIF means). The minimum order is 5, but I don't mind getting 5, maybe a couple for myself and I expect I'd eventually find people here who would take the other 3 off my hands.

I also saw a post from Tangent who said he'd report back on his experience with a UK company, but couldn't find any follow-up posts about that.

I'd appreciate up-to-date advice on this. If I do eventually have a go at re-celling, I can imagine posting photos here later saying things like "what do i do with that red wire at the top of the photo?" etc as my knowledge of electronics just about extends to changing a fuse.

 

[flecc]

As long as the pack is a 7-6-7 cell formation oblong shape using D cells Neil, you can follow the instructions on my Lafree website:

Recelling the NiMh battery

This introductory page also mentions the availability of suitable singe packs from Batteryspace in the USA, although that might attract import duty and VAT charges. There's also a mention of The Component Shop who will make up a single pack for you at about £125:

Battery replacement options
.

 

[jac]

hi this pack is on ebay germany might be worth trying to ask if they will send to uk 110550688177




jim