NI-MH in parallel or "fully" parallel?

[malow]

i will do a pack with 12 NI-MH AA (LSD), and was wondering, witch is the best configuration?

"fully parallel" is better to balance the the cells variation in power? and to better avoid reverse charging?

all cells will be charged in separated independent charging, so they "should" be always in the same state of charge

 

[45/70]

I've done this on a smaller scale, with various NiMH cell sizes. I discharge 4 "D" NiMH cells in parallel on my CBA II when conditioning them, for example. True, there is no way you can have a reverse charge "event", provided you don't put these paralleled "banks" in series, and use them that way.

EDIT: malow added diagrams while I was composing this post. See post#4

Dave

 

[Mr Happy]

I don't see any short circuits?

However, I think the second picture is preferable. Intuitively it seems like it should provide more balancing between the cells.

Do make sure that you charge the cells separately and do not attempt to charge them in either pack formation.

 

[45/70]

My apologies, malow.

I don't see any short circuits?

No, there aren't.:) malow had posted the diagrams while I was composing my post. I didn't realize he was referring to a series/parallel pack. Between my multi tasking (which I've never been especially good at) and having to leave for a meeting, 10 minutes before I actually did, I could not properly respond to his question. From my perspective at the time, his diagrams did appear to be "massive series/parallel short circuits", as that wasn't what I was thinking of! Poor choice of words.:sigh:

My only experience similar to your idea, malow, is using 2AA to "D" adapters for use in various 4 "D" devices, as in 4s2p. In this configuration, I have always felt that the risk of running down and reverse charging 2 cells (2p), is less than a single cell. Technically speaking though, the risk is just the same, since the 2p adapters are treated the same as a single cell, regardless how many cells are in parallel. As I said in my previous post, if you run paralleled cells in series, you still run the risk of reverse charging, it will just reverse charge all the cells in that group, as one.

I will edit my previous post. :candle:

Dave

 

[TakeTheActive]

i will do a pack with 12 NI-MH AA (LSD)...

What would be the proper naming convention for the packs above?

Personally, I would work from the outside in (groups-to-cells), thus:

• 3P4S - Top (3 Parallel Groups of 4 Serial Cells)
  .
• 4S3P - Bottom (4 Serial Groups of 3 Parallel Cells)

...witch is the best configuration?

I believe the bottom configuration, 4S3P, would be better for the cells because the weakest cell in each group of 3 would be supported by 2 other cells and thus the WHOLE group would have to go negative together.

In the top configuration, 3P4S, each cell is pretty much on it's own. But, you'd still need a pretty big imbalance such as (only comparing 2 series strings):

• 1.4 + 1.4 + 1.4 + 0.0 = 4.2
• 1.1 + 1.0 + 1.1 + 1.0 = 4.2

or

• 1.2 + 1.2 + 1.2 + 0.0 = 3.6
• 0.9 + 0.9 + 0.9 + 0.9 = 3.6

 

[65535]

Technically the second setup would help keep the batteries in a better balance with each other.

If using a peak charger you don't particularly want to parallel NiMH or NiCd cells, lithium like it fine with a CC/CV charge, but peak detection chargers tend to get false readings and can damage parallel cells.

It would be best to use a bay charger with individual charging, or charge in series only, whether charging each string of 4 individually or in a single 12S setup.

If you charge to 1.4V or 1.5V cutoff and don't use peak detection it wouldn't matter though.

 

[malow]

tnks for the answers guys... solved my doubts :twothumbs

 

[Mr Happy]

If you charge to 1.4V or 1.5V cutoff and don't use peak detection it wouldn't matter though.

I think it can matter depending on where you set the peak.

The trouble with parallel NiMH charging is that when cells get near to full charge they get warm and their internal resistance drops with increasing temperature (that's why RC racers like to use packs "hot off the charger" -- they can deliver more amps that way). What can happen is the cell or string that warms up first takes more charging current than the others and warms up faster. This causes it to take even more current and so on. A positive feedback loop results that can cause a big imbalance in the current distribution. This problem can happen even if minus delta-V termination is not being used.

The overall result is that when using low charging currents there might be uneven charge distribution between the cells, and when using high charging currents there might be a bad overcharge or a meltdown.

 

[malow]

another thing... no problem doing this right?

sorry for many questions... i know the logic but don't have the technical knowledge... :(

 

[Mr Happy]

another thing... no problem doing this right?

I guess it doesn't make a lot of sense. Are you implying that some of the cells are of different sizes?

Good design normally keeps everything symmetrical.

 

[malow]

all cells are AA. all batteries are the same capacity.

i guess only batteries with different capacity in serial configuration is bad right?

 

[MrAl]

Hi,


If you are going to parallel strings that are in series that means you have a lot more
voltage to work with than when dealing with single cells. This means that series
diodes are a possible way to make this pack work very well with no problems
that are sometimes associated with parallel banks.
The idea is to (from the drawings) make three banks of four cells in series, then
connect one good quality Schottky diode in series with each bank (that requires
three diodes). If you connect the anodes to the positive terminals, then you would
connect all three cathodes together and that would become your new positive
terminal for the entire bank.
The negative terminals of the three banks are connected together and that becomes
the negative terminal of the bank.

To charge, depending on the charger you can use three more diodes (almost any
kind that can handle the current) connecting the cathodes to the positive terminals
of the three banks, and connect all of the anodes together. The positive charge node
is then the three anodes.

What the diodes do is prevent ANY cell from getting charged directly from any other
cell and so the pack works very nicely. The drawback is that the diodes introduce a
voltage drop of about 0.4v, but with the higher voltage of four cells in series it is not
as much of an issue as with one cell alone.