Problem with weak NiMh battery

[zgurl]

Hi everyone,

I wanted to share a bad experience with a weak NiMh cell. If somebody could confirm my undertanding of what happen and give me some clue to avoid it happen in the future, I would be very grateful .

I had just nearly finished my first bike light (with 3 P4 Crees, roughly based on nightrider ideas) using a 6xAA NiMh and I was testing the electronic.
I was very surprised because after a short time (5 minutes or so), the batteries ware getting very hot (so that the plastic holder start to melt down ). I measured the voltage and I found Vin=4V while ligh was on and Vin=7.5V when it was off.

:thinking: What was the problem?????

I do not have high skills in electronic (well... I am not sure I have basic skills either ) and I did not know this voltage was supposed to be (so) different with or without charge.

The problem is actually that this voltage drop implied a high current out of the batteries (nearly 3A) and certainly was the reason for the battery heating! I swicth the NimH batteries with a lead 6V battery and everything was OK, the on/off voltage was about the same and the current drained out was less 2A and nobody get hot. But I definitely don't want to carry that heavy lead battery on my bike! So I have to understand where the problem was with NiMh...

Then I tried the experience once again checking more precisely everything: voltage, current, heating... It appears very quickly that one of the batteries was getting warmer than the others... and when i checked its voltage (while light was still on), it was getting lower and lower. I went on the experience and saw the total voltage suddenly falling down (and I imagine the current rose up). I stoped the experience before my battery holder was melted once again

My conclusion of this is the following:
One of the cell had a problem (or I forgot to charge it as the other ones?) and was nearly empty, so that it could not sustain the current driven by the board, so that its voltage dropped, so that it was more and more empty, so that voltage droped all the more, etc...

OK, once I have identified the battery, I can replace it. But when using 6 NiMh, I am never certain that they absolutely identical (whit same capacity, same history, etc) and it will necessarily happen that one is weaker thant the others and while it will reach the end of its capacity, I suspect I will have the same type of problem.

If I am right, can I imagine a sort of "protection" for NiMh (something similar to LiIon, avoiding high current)?

PS: This bad experience draw my attention on the possible danger of using batteries... and I can imagine the same type of problem with Lithium cells, but then the problem could be more dramatic than only a melted holder

 

[VidPro]

you did not indicate the curcuit your using to control anything?

if you got stuff getting that hot with a POOR battery, or bad connections to it or anything, whats to say that wont happen again?
so the first thing i would say is dont try and suck any ni-mhy down in only 1 hour, if you intend to use this thing daily, you should have enough JUICE in there to run a few times before charging it. like mabey using C cells. or your suposed to be running 3 cells in parellel with 3 other cells, as discused below.

the Connections on Cheap battery holders are useless for high power, while having the batteries ACCESABLE via a plastic holder is great, most of them suck horribly but can be "fixed" if it has areas where the electrical conduction is made with metals that are just pressed together (rivets) you can solder the metals to eachother.
the reason i mention that, is a Bad connection will heat itself up and the battery and melt the holder.
also the wiring in a holder can be limited. for me most of the holders are a hunk of plastic and some springs, something to start with, but never good enough.

in a system like this, having a "Breaker" in line with the batteries would be a good idea, like a 3 or 5 amp resettable breaker, they are Easy to get hold of. it is preferable to put the breaker as close to the battery Pack as possilble, so if other wires short, its covered.
inline fuses, and inline blade fuses could be used, depending on need. they even have mini inline blade fuses now.

i really think you should check your Connections to everything, cause it might not have been a battery.
and your "LOAD" seems aufully high, are you sure of your battery source voltage, VRSES the voltage your pumping into the leds? mabey you were supposed to run 2x parellel and only 3 in series, giving you a max voltage of 4.2. it SOUNDS like your jamming 6X series batteries into a poor 3.8V leds.

A weak cell will heat quickly when it is drained, is being used in Series, and gets a "reverse charge" from the rest of the cells, it gets damaged a lot, if it heats from reverse charge. that could indeed have been your problem. but why is your load so high.

anyways that is all i can think of and is all guesses, till you show some pics, or schematic drawing of what you got set-up.

are your LED wired in Series or parellel?
are your batteries in series or parellel?

 

[VidPro]

the build your referring to uses a George80 N-FLEX board, that board can take as much as 16V input, and will current control as many as 4 crees in Series. it has current possibilities of 250-500-750-1000 if i remember right.
but it does not do any "boosting" meaning the input voltage needs to be higher than the output voltage.
but its more efficient if the input voltages and output voltages are reasonably similar.
so if you stuff 14V in, you would probably want to wire your 3 leds in series.

proper application of that curcuit, and because it does not "boost" would mean that the max power that should be drawn from the batteries would be the 1Amp or 1000ma that the board does.

so as soon as you get over the 1 amp, IF you were using that curcuit, you should pull the power and wonder what is wrong.

if your not using that curcuit that guy used, then what you ARE using will be more important than what the guy did for his.
meaning you wouldnt require that curcuit, to assemble a similar looking light, but you would need to indicate what you did do, for anyone to help.

 

[zgurl]

Thanks for your answers,

you did not indicate the curcuit your using to control anything?

I used a Fatman from Taskled, I planned to use a pot to dim the light but the one I bough did not fit inside the aluminium tube, so the wires are connected to nothing (yellow wires on the pic below).

if you got stuff getting that hot with a POOR battery, or bad connections to it or anything, whats to say that wont happen again?
so the first thing i would say is dont try and suck any ni-mhy down in only 1 hour, if you intend to use this thing daily, you should have enough JUICE in there to run a few times before charging it. like mabey using C cells. or your suposed to be running 3 cells in parellel with 3 other cells, as discused below.

Do you mean that 2 amps during one hour is too big a current for a NiMh Cell? (if I want to be able to use more than once)

the Connections on Cheap battery holders are useless for high power, while having the batteries ACCESABLE via a plastic holder is great, most of them suck horribly but can be "fixed" if it has areas where the electrical conduction is made with metals that are just pressed together (rivets) you can solder the metals to eachother.
the reason i mention that, is a Bad connection will heat itself up and the battery and melt the holder.
also the wiring in a holder can be limited. for me most of the holders are a hunk of plastic and some springs, something to start with, but never good enough.

Yes you are right.

Here is a picture of the melted holder (my first posted pic!)

in a system like this, having a "Breaker" in line with the batteries would be a good idea, like a 3 or 5 amp resettable breaker, they are Easy to get hold of. it is preferable to put the breaker as close to the battery Pack as possilble, so if other wires short, its covered.
inline fuses, and inline blade fuses could be used, depending on need. they even have mini inline blade fuses now.

Good idea to use a fuse to avoid such problem :thumbsup:

i really think you should check your Connections to everything, cause it might not have been a battery.
and your "LOAD" seems aufully high, are you sure of your battery source voltage, VRSES the voltage your pumping into the leds? mabey you were supposed to run 2x parellel and only 3 in series, giving you a max voltage of 4.2. it SOUNDS like your jamming 6X series batteries into a poor 3.8V leds.

Actually, I use 6 cells in series.
3 Crees in series:

Fatman driver (the yellow wires are connected to nothing)

And finnally the overall circuit:

A weak cell will heat quickly when it is drained, is being used in Series, and gets a "reverse charge" from the rest of the cells, it gets damaged a lot, if it hetats from reverse charge. that could indeed have been your problem.

That's what I guess also

anyways that is all i can think of and is all guesses, till you show some pics, or schematic drawing of what you got set-up.

are your LED wired in Series or parellel?
are your batteries in series or parellel?

 

[VidPro]

Do you mean that 2 amps during one hour is too big a current for a NiMh Cell? (if I want to be able to use more than once)

well its not "to much" by specs, but if you want longevity of the cell, and some runtime, its rediculous to draw the cell down in only 1 hour, when you could have an adequite capacity and then be discharging the battery at a rate it could easily handle.

with that "system" you can potentially run it at 1/2 power, then boost as needed, then the capacity/runtime is sufficient.

Fatman is a Boost regulator (step up), so input voltage must be less than the output voltage to ensure Fatman remains in regulation
with to much input voltage for the LED voltage , it will start doing a direct drive. BUT your 8.4V max (6xnimhy) Should be below the voltage of the 3 Crees wired in series. sooo, all that seems quite correct? looks like a good method.

if you used the N-flex Instead, and some correct voltage of battery for the VF of the leds, you would have a "cutoff" of sorts for the ni-mhy batteries, because when it falls out of "regulation" it would draw LESS current, as the voltage dropped on the leds.
that way when your battery voltage went below some ammount , it would dim down, and not "reverse charge" any cells in the pack.
By making the Vin (battery voltage) and the Vout (led Vf) close, adding in say 1-1.5V for the curcuit , when any Ni-mhy cell was completely depleated and headed for reverse charge, the light would be going very slow/low.

hopefully OTHERS skilled in electronics can Chime in, because curcuits are often beyond my skill levels. i still dont understand how you got the 3Amps, unless there is a high level of inefficency in the boost, and the voltage was drooping badly. knowing georges curcuits, it wasnt about to regulate "less" untill the battery is zip another reason for the n-flex and matched input voltages

 

[VidPro]

that IS a riveted battery holder, they CAN be fixed, just carefully (without overheating) solder the metal Connection points TO the metal spring connectors. then solder your connection wires STRAIGHT to the connectors.

then prefer not to use the 9V adapter for high currents, it gets loose over time, it can still be used as a connection point, but it will get sloppy and hot with high power devices.

 

[VidPro]

hey i dont know how to wire up a fatman, but your pics show a switch disconnecting the LEDs, from what i read, some "BOOST" converters can be "damaged" if the "load" is removed. so what is the switch for?

great pics by the way

 

[Power Me Up]

hey i dont know how to wire up a fatman, but your pics show a switch disconnecting the LEDs, from what i read, some "BOOST" converters can be "damaged" if the "load" is removed. so what is the switch for?

It looks to me like the switch is between the power source and the regulator circuit...

 

[VidPro]

ahh see now that makes sence, so the 2 yellow lines were for the led.

 

[Power Me Up]

ahh see now that makes sence, so the 2 yellow lines were for the led.

Actually, I would have said that the orange and grey wires were the ones going to the LEDs. I'd say that the yellow wire with aligator clips is there to allow a multimeter to be inserted to measure current going to the LEDs.

It wouldn't make sense for the yellow leads to be the ones for the LEDs since they are polarity sensitive...

I wasn't sure what the yellow leads were for until I re-read zgurls post - they were originally for a pot that was to be used to dim the light, but it didn't fit.

 

[VidPro]

http://www.taskled.com/leds/fattop.jpg
that oughta help

 

[zgurl]

Actually, I would have said that the orange and grey wires were the ones going to the LEDs. I'd say that the yellow wire with aligator clips is there to allow a multimeter to be inserted to measure current going to the LEDs.

It wouldn't make sense for the yellow leads to be the ones for the LEDs since they are polarity sensitive...

I wasn't sure what the yellow leads were for until I re-read zgurls post - they were originally for a pot that was to be used to dim the light, but it didn't fit.

You got it PowerMeUp!

I am sorry guys, I should have given those color explanation myself. But that was a good reverse engineering exercise .

 

[zgurl]

that IS a riveted battery holder, they CAN be fixed, just carefully (without overheating) solder the metal Connection points TO the metal spring connectors. then solder your connection wires STRAIGHT to the connectors.

then prefer not to use the 9V adapter for high currents, it gets loose over time, it can still be used as a connection point, but it will get sloppy and hot with high power devices.

Thanks for the advice VidPro, I guess this little trick can enhance my chance of getting the proper light!
Would you advice to solder wires directly instead of using 9V adaptator?

 

[VidPro]

yes, that is what i do, when i have to use cheap radioshack holders. course good ones weigh a lot more, so ya do what ya gotta do