Reverse charge protection?

Gather App
Candle Power Forums

Help Support Candle Power:

G

gwbaltzell

Enlightened
Joined
Dec 8, 2003
Messages
578
City & State/Province
StL MO
A question that's a result of the How many NiMhs thread. A very long time ago (> 10 yrs) I saw a suggestion to put a diode across each cell to prevent (actually reduce) reverse charge from happening. While I have thought about this from time to time, and the problems of doing it, has anyone actually tried it.
Known problems:
1. increased cost
2. each must have a fuse or something to limit current to reduce fire risk if diode shorts
3. each diode must be large enough to carry the full load
4. must have a very low Vf (Schottky barrier)
 
Yes.

I worked on a defribulator (the gizzie that shocks your heart back into life). It used 980 Volts from a very large cap bank to do the deed. 48 volts of gel cells drove the inverter, each of the six volters had a diode across it. It wasn't so much to protect the cell, but to ensure that a flat one wouldn't kill the string (after all, it's a life and death deal).

The problem is the cell is dead flat, even reversed by a bit, before the diode does anything. For NiMH it will be too late. In the above case (six volt batteries), a single cell would be reverse charging before the diode kicks in anyway.

A large part of service was the battery pack. The charge termination was very bad (voltage drifted a lot), lots of ambulance companies left them on full time charge, some had crews that would 'test' it from time to time expecting guaranteed number of full shocks and call for service if they didn't get it.

Doug Owen
 
Yes, I can see using diodes to insure that current can still be drawn from the rest of the battery.
Sorry, I was thinking about NiCds and I didn't say this. The Vf of any diode would be too high to do any good for NiMh or gel cells. But I've read that NiCd have some tolerance to reverse charge but don't find any real data on this.

While searching to see if anyone currently implemented this I did come across a discharge limit I'd not seen before. One of the Asian makers recommends the common 1.0 V per cell limit for packs up to six cells but for 7 to 20 cells recommends 1.2 V * (number of cells - 1). Following this would seem to increase the likelyhood of reverse charging. Of course this same maker likes the quasi-constant current method of charging (a resistor) because the constant current method is "costly"!

George
 
Hi there gw,

You could check the other thread where we were talking about
a protection circuit that monitors every cell individually.
That would catch one long before it reversed on you.

Take care,
Al
 
Thanks MrAl. I have been following that thread with interest and agree that not only monitoring but charging of each cell is ideal. This is what is done often on satellites. Just wondering if a slightly lower tech. approach could be worth while? Yes, I know its possible this approach could wind up costing more, I just think some people would be more comfortable with it. Just can't find any definite info. if NiCds would actually tolerate this limited reverse voltage or for how long.

George
 
[ QUOTE ]
gwbaltzell said:
But I've read that NiCd have some tolerance to reverse charge but don't find any real data on this.

George

[/ QUOTE ]
I remember reading [reliable source] long ago that discharging a single NiMh or NiCd cell down to 0V is basically a non-issue. The undesirable, non-reversible stuff happens at a reverse charge of 0.2V. Unfortunately, even a schottky would have a higher Vf and thus not offer much protection.
 
Well, I suppose you could use two diodes per cell but the voltage drop in normal operation would be too high. Knew there had to be a reason I never actually tried this. /ubbthreads/images/graemlins/rolleyes.gif
Looks like monitoring each cell is really the only solution.

Thanks all!
George
 
[ QUOTE ]
gwbaltzell said:
Well, I suppose you could use two diodes per cell but the voltage drop in normal operation would be too high

[/ QUOTE ]
???? Maybe I misunderstood the original post. I thought the idea was to put the diodes in parallel with each cell. In this case, the Vf of the diodes doesn't come into play in normal operation since they normally are not conducting any current.
 
Sorry, should have known that would be confusing. Adding a diode in series with the chain and one in parallel with the combination might counter one part of the problem but create another problem.
<font class="small">Code:</font><hr /><pre> |
|
|
__|__
___
|____
| |
_|_ |
^ |
| _|_
__|__ ^
___ |
|____|
| repeat 2 diodes and cell</pre><hr />
 
[ QUOTE ]
gwbaltzell said:
Sorry, should have known that would be confusing. Adding a diode in series with the chain and one in parallel with the combination might counter one part of the problem but create another problem.
<font class="small">Code:</font><hr /><pre> |
|
|
__|__
___
|____
| |
_|_ |
^ |
| _|_
__|__ ^
___ |
|____|
| repeat 2 diodes and cell</pre><hr />

[/ QUOTE ]

And you thought I was confused before?? /ubbthreads/images/graemlins/thinking.gif
/ubbthreads/images/graemlins/grin.gif
 
Hi all,

You could use a zener with a specified rating to prevent undervoltage, then use a normal diode to prevent reverse charging.. But I'd expect that the losses involved in that setup wouldn't be worth it.

pb
 
Sorry, the original question was just a diode in parallel with the cell. This two diode thing was just an idea that is really not worth it. The forward voltage of the one in series would kind of counter the Vf of the one in parallel. But the loss in normal operation would be too great and of course make it impossible to recharge! I should not have brought it up.
 
Hello again,

I guess there is just no *really* low tech answer, but
then again if was going to go through all the trouble
of connected an individual wire to each and every cell
so that each cell could be monitored individually i sure
wouldnt want to have to use real heavy wire and i would
want that extra effort really worthwhile.
Sensing voltage only requires light gauge wire, which
helps a little, and definitely protects the cells from
going too low and thus reversing.
It is always possible to construct a somewhat simpler
circuit if you're willing to supply an extra small
battery to run the circuit itself. Some of the comparators
out there use such little power a battery in this kind
of app would last a year before needing replacement.
Maybe one of those 3v button cells could power the circuit,
or a small 12v type.
The circuit itself could be mounted on a small pc board,
with solder terminals for the individual battery wires.

Take care,
Al
 
Part of the 'best solution' includes how the battery holder works. It's just not practical to put taps in tube type holders, and not at all convenient in flat packs where there are two or more in series.

Given an open world, I suggest the best deal is separate leads from each end of each cell. Bring all to a single connector (say a d-sub) along with the necessary light leads. Now you can use a header to put them in series for use, and a different one to connect the charger to individual cells (which can even be charged in parallel if you wish).

Doug Owen
 
You must log in or register to reply here.

Latest posts

Back
Top