NiMH and voltage depression

ChrisA

Enlightened
Joined
Feb 10, 2002
Messages
369
Location
Germany
i did a search on that topic, but found nothing. so please forgive me if this is common knowledge.

well, most of us use nimh cells from day to day to power their flashlights, digital cameras, mp3-players and so on... and most of us, at some point looong ago 😉, have fallen for the low prices these cells are going for over the net. at least that's how i got these nice aaa's:

nimhvd1.jpg


as you can see, these are supposed to hold 700mAh of charge and were made in china. i got 8 of them and paid ~10 USD including shipping. after a couple of charge/discharge-cycles none of them made it past 500mAh on my charger, so i figured that this was a 10 dollars lesson learned. so, after another unsatisfactory discharge i pulled them and was decided on trashing them.

out of curiosity i checked one cell with a dmm: open circuit voltage was ~1,1V, slowly declining to 0,9V if put under load with a 50 mOhm resistor. that's the point, the charger will stop the discharge cycle and starts recharging the cell. so far so good. but if you let the cell rest for a couple of minutes, it would recover to 1,1V
shrug.gif
.

that's where voltage depression comes into play. the cells still contained a fair amount of charge, but would not release it fast enough. this phenomenon is caused by Ni-use in rechargeables. NiCad's suffered double from that problem and when NiMHs were developed, we only got rid of one part of it. it has to do with the way the Ni is disposed in the cell and the way it will form crystals that won't break up as easily (btw - crystal forming will benefit from slow charging, so fast chargers are better suited for NiMH chemistry, if they reliably detect a full charge).

anyway, the charger would not let me discharge the cells completely, since the load it put on the cells would cause a voltage depression, after which the charger sensed an empty cell and started recharging. so i put the supposedly empty cell in an arc aaa and it would start out bright, but within 30 seconds it's output would decrease until completely dead in about two minutes. but turning it off and on again would work over and over - not a good way to deplete the cell...
rolleye11.gif
so, what to do ?

individual NiCad cells can be discharged to 0V without ill effects. that's what the rc-guys do, if their battery-packs show signs of aging. just connect both poles with a resistor and let the cell sit for a couple of hours (that's for individual cells only, so for battery packs they used a device called "nagelbrett").

NiMHs will be harmed by deep-discharging them, so a simple resistor is not suitable to get rid of the lazy battery effect. diodes to the rescue:

nimhvd2.jpg


that's a 2.7 Ohm resistor and a 1N4001 diode in series, connected to both poles of the cell (and a nice example of a bad soldering job - but it works nevertheless 😉 ). the 1N4001 has a breakdown voltage of ~0.7V. it will let current pass through, unless the voltage sags under 0.7V. cost of materials is about one dollar and assembly may take three minutes (if your skills with a soldering iron are as bad as mine
mpr.gif
).

i let it sit for a day:

nimhvd3.jpg


that brought open cicuit voltage down to 0.88V and following charge/discharge-cycle got 0.62Ah from that cell
party.gif
. i will repeat the process a couple of times to see wether it will gain even more capacity. anyway, that's close enough to 700mAh for the price i paid.

so, before you trash old or supposedly crappy cells, it may be worth to give it a try... (i doubt it will work on suspiciously cheap 2.800mAh Tamiyas though 😉 )

chris
 
Last edited:
Well, voltage depression isn't a certain voltage, it's when you discharge part of the cells capacity and recharge, repeatedly to the same point! Ni-MH do suffer from this but not as bad as Ni-CD, it takes longer to happen and it can be solved with a complete cycle or two.

RC guys don't discharge Ni-CD to zero volts. Why would they want to lose power or control in mid-flight? Besides, try putting a zero volt cell in a smart charger, it'll tell you, no cell!!
 
Well, voltage depression isn't a certain voltage, it's when you discharge part of the cells capacity and recharge, repeatedly to the same point!

okay, here's how i understand "voltage depression": a cell using Ni chemistry forms large crystals during charge (which happens if cells are overcharged, not cycled constantly and stored over a longer period of time) which will lead to a "reduced" capacity during discharge. that's because the larger crystals won't break up easily, so that the voltage of the cell drops, although there's still plenty of reactive material left...

if you do a cell analysis, a charger like the bc-900 will discharge the cell to ~0.9V and then starts recharging. if the cell is "damaged" due to the large ni-crystals it will sag under 0.9V even if there is still reactive material in the cell and capacity is "lost".

RC guys don't discharge Ni-CD to zero volts. Why would they want to lose power or control in mid-flight?

hm, you got me wrong... a buddy of mine is into rc-boats. he "revives" low voltage nicad packs by discharging each cell down to 0V (not during use of course). because he does not want to disassemble the pack, he uses a thing called "nagelbrett"... the shrink wrapper is punctured by nails, so that each cell can be discharged completely.

Besides, try putting a zero volt cell in a smart charger, it'll tell you, no cell!!

he uses a special charger (like the schulze charger tom mentions from time to time) that will recognize the battery pack even if completely depleted.

chris
 
No, voltage depression or memory effect is from repeated partial charges. I have a Schultze and it's very picky about what it likes! It will do a non-timed, non-termination charge, if that's what he means.

I neglected some brand new AA Ni-MH and some read zero volts. I tried all tricks, some survived and some didn't!!
 
I neglected some brand new AA Ni-MH and some read zero volts.

i read that post and it was very informative, thank you. that's what the diodes are for - they will prevent the NiMH cell from discharging beyond 0.7V, since this will likely cause damage (IIRC there's a sanyo twicell article about that). the 0V-method works for NiCds only.

chris
 
Back in the day,
We had to discharge 144V NiCad battery packs used in portable X-Ray units and we discharged the packs to 0.000 volts. Attached a large bank of lights to them and let them run 24 hours past the point of no glow on the filament. Then we took the packs apart and attached shorting bars to each cell and let them sit for 24 hours dead shorted. Recharged the packs and repeated it every 6 months. We replaced the packs every 5 years and never had any problems.
 
The diode is a nice idea!

Voltage depression is how Duracell describes the memory effekt.
http://www.sanyo.com/batteries/pdfs/twicellT_E.pdf
http://www.duracell.com/oem/rechargeable/Nickel/voltdep.asp
http://www.panasonic.com/industrial/battery/oem/images/pdf/Panasonic_NiMH_Overview.pdf

I´am not convinced in discharging NiMH down to an open voltage <1,1V.
As you already said the voltages drops under load. There is NO benefit in discharging them under 1V.
http://www.ife.ee.ethz.ch/~zinniker/batak/ideal/index.html

But yes it´s a interessing question...
I want to know more about that.



@ BentHeadTX

I own a portabel screwdriver since 2001
The unit use NiCd Sub-c. unbelieveable how strong they are.

Do you already have expierence with NiMH in similar applications?
 
Last edited:
@bentheadTX - 144V packs ? wow, that must have been an impressive battery pack ! would be interesting to know the loads, these were discharged at... 🙂

the packs my buddy uses for his rc-boat are only ~7V but current draw will peak at ~50A
ooo.gif
. that's why he still uses nicads, since nimhs aren't capable to deliver currents that high.

@lamperich - thanks for the very informative links ! i just read the duracell paper and this is what i found:

"The voltage drop occurs because only a portion
of the active materials in the cell is discharged and
recharged during shallow or partial discharging. The
active materials that have not been cycled change in
physical characteristics and increase in resistance.
Subsequent full discharge/charge cycling will restore the
active materials to their original state."

that's basically what i said above. anyway, the duracell paper goes on:

"The extent of voltage depression and capacity
loss depends on depth of discharge and can be avoided by
discharging the battery to an appropriate cutoff
voltage. Voltage depression is most apparent when the
discharge is terminated at higher cutoff voltages, such as
1.2 volts per cell. A smaller voltage depression and
capacity loss occurs if the discharge is cut off between
1.15 volts to 1.10 volts per cell. Discharging to 1.0
volts per cell should not result in significant voltage
depression or capacity loss during subsequent discharges."

well, that's the interesting part. according to duracell, it's sufficient to cycle the cell a couple of times to get back to full capacity. that's what i tried with the cell i mentioned above.

the chargers i mostly use (akkutrainer 1 and 3) will discharge a cell down to 0.9V and then start recharging. four full charge/discharge cycles brought the e-cell close to 500mAh, but that's all i got. after discharging it through the resistor + diode the same cell got 620mAh. that's quite an increase and i guess it's due to some reactive material being "reactivated" by the slow discharge to 0.7V
au.gif
icon5.gif


chris
 
Last edited:
"""it will sag under 0.9V even if there is still reactive material in the cell and capacity is "lost""".

yup, it sure will, the speed of discharge at the low end will be slow.
and dropping it a bit further for real, minus the huge drop that hits when using these "smart" chargers, is missing cycling the end.
good stuff.

Silver reported some manufacture saying it was Ok to drop to .4v slowly. i would think that would achieve the best attempt at cycling them, when they arent speedy at the bottom end.

i like to drop further than spec during a Cycling discharge, if i am going to crank them right back up again. IF i am having to specifically cycle to get it alive.
your method looks easy, now why didnt they bother to put a "feature" for doing a fuller deeper discharge in the "smart" chargers.

i have 2 old RS charge discharger things, that held 4 batts, might be a good place to put in that method.

i still dislike AAAs for recharging, because they are so puny and practially useless, and they get hurt easily. any improvement to them would be valuable.
 
Last edited:
your method looks easy,...

it's not "my" method 😉 - i just added a diode to the resistor to prevent deep discharge. otherwise this technique is used on nicads for quite some time with good results.

...now why didnt they bother to put a "feature" for doing a fuller deeper discharge in the "smart" chargers.

i guess it just takes too long... 🙂 the cell i tested took 24h to drop down to 0.88V open circuit voltage, since current through the resistor will drop to a couple of mA when the V "cut off" of the diode is reached.

i still dislike AAAs for recharging, because they are so puny and practially useless, and they get hurt easily. any improvement to them would be valuable.

you're absolutely right. it's really easy to overcharge them. that's why i prefer a fast charger and monitor the time - at ~1C they give a decent cut off signal, otherwise i pull them after 1:20h. the accutrainer 1 will run into timer cut off with 600+ mAh AAA cells, because the rate is too low and it doesn't sense the charge peak. that way, i ruined two nice sets of sanyos 🙁

chris
 
ChrisA said:
the chargers i mostly use (akkutrainer 1 and 3) will discharge a cell down to 0.9V and then start recharging. four full charge/discharge cycles brought the e-cell close to 500mAh, but that's all i got. after discharging it through the resistor + diode the same cell got 620mAh. that's quite an increase and i guess it's due to some reactive material being "reactivated" by the slow discharge to 0.7V
au.gif
icon5.gif


chris

that is a very interesting observation

The AT1 usually discharge to a lower level ;-)
http://www.digicamfotos.de/index3.htm?http://forum.penum.de/showthread.php?id=25821
 
>>>
A smaller voltage depression and
capacity loss occurs if the discharge is cut off between
1.15 volts to 1.10 volts per cell.
>>>

I wonder how manys devices really do this? I know that a Fluke 189 DMM doesn't!
 
The AT1 usually discharge to a lower level

in case of my at1 (i got the older model from conrad, about 2 years ago) the cell will recover to a higher voltage, as soon as you pull it from the charger. the above e-cell recovered to 1.1V. and that's exactly what i try to understand.

the open circuit voltage of a depleteted cell should be ~0.9V. so i guess a cell that is "force" discharged to 0.9V (or even to a lower voltage) isn't really empty, since it will recover over time. that seems to indicate, that there is still reactive material in the cell which might be bound into structures that need longer periods of time to break up... otherwise i don't understand, why there's such a jump in capacity
help.gif


chris
 
ChrisA:

The same point came up another thread and SilverFox ran tests discharging cells all the way down to .5V. The cells always recovered to over 1V.
 
i dont think of it as a increase in capacity, because i think that the capacity is THERE, its just that the chemicals are slow to "spring" back.
the reasons we get a TESTED increase in capacity, is because were always stopping at some set voltage, when testing.
were not really completly depeating the cell to deadness, to get the SAME test, we use specific machines or stop points.

if the battery is still pushing juice out Quickly (speed) at the low end of the discharge (about 1v), then it will display that it has more capacity, but did the ACTUAL capacity change, or was the cell just vibrant enough at the low end, to KEEP the voltage up there, by keeping the flow going.

IMO if you cycle it THERE also (at the low end), then it is speedier there.
SO
if you were discharging a 5mm led with 3 of them, it will keep plugging away, because the 5mm led will only be pulling 2ma. Eventually you will get all possible capacity out of it. (mabey kill it, but you get it all)

but general use of things is that your trying to get a lot of current (for its total) out of them, the faster they discharge at the end , the more capacity the average device will give you, before it cuts out due to lack of "voltage".

we put rechargables in things that we need juice quickly, not because a alkaline would do the job 🙂 so general use of them a speedier discharge at the neer end of its discharge cycle, can be a fair ammount of time longer.
 
Last edited:
ChrisA said:
in case of my at1 (i got the older model from conrad, about 2 years ago) the cell will recover to a higher voltage, as soon as you pull it from the charger. the above e-cell recovered to 1.1V. and that's exactly what i try to understand.

the open circuit voltage of a depleteted cell should be ~0.9V. so i guess a cell that is "force" discharged to 0.9V (or even to a lower voltage) isn't really empty, since it will recover over time. that seems to indicate, that there is still reactive material in the cell which might be bound into structures that need longer periods of time to break up... otherwise i don't understand, why there's such a jump in capacity
help.gif


chris

iam not convinced in discharging a NiMH to an open voltage of 0,9V!
but it´s a very interesting question.


Can you trust the capacity showen by the BC900 or AT3?

Can you measure Ri ?
http://www.elv.de/shopping/ArtikelD...Referenz=471-35&Stufe=2&Gruppe=ME-BA&Menue=Ja
if link doesn´t work search for "Rim 1000" on www.elv.de
 
@vidpro - so it's not capacity that changes, but voltage during discharge (called "Spannungslage" - i don't know the correct english term). hm, that makes sense
icon14.gif
.

chris
 
@chris


---
As with Ni-Cd batteries, the charge voltage of nickel-metal
hydride batteries may show signs of swinging
(pseudo -.V) when they have been kept standing
for a long time or when they have discharged
excessively, etc. The initial delay timer is needed
to prevent charge from stopping (to prevent
malfunctioning) due to this pseudo -.V.
---
http://panasonic.com/industrial/battery/oem/images/pdf/Panasonic_NiMH_ChargeMethods.pdf

The question is: What is "excessively discharged"?

i don´t know it.


If you only have a charger like the Akkumanager10 or 20 or AT1 you should not be surprised if you also get a short runtimes like Quickbeam get with his first NiMH Runtimeplot in a Streamlight 4AA.
http://www.candlepowerforums.com/vb/showthread.php?t=103445

Quickbeam didn´t store the old runtimeplot. but i did.



in Digicam you get pretty soon pretty short runtimes.
BUT
This coud be pretty dangerours for example in a Fenix 2AA.
Remebers this regulator also works with only 1 battery nearly as bright as with 2!

THAT`S exactly why i don´t like NiMH´s anymore.
The risk that we user abusing them without knowing it is pretty high.

also read this posting made by bitcharger.
You can find there an ugly "spannungslage" http://forum.penum.de/showthread.php?id=24293&eintrag=60 🙂
 
Last edited:
ChrisA said:
@vidpro - so it's not capacity that changes, but voltage during discharge (called "Spannungslage" - i don't know the correct english term). hm, that makes sense
icon14.gif
.

chris

ya thats it, the Spannangslage effect. i knew sombody would finnaly name it, so we can put a scope on it and test it 🙂

i have an idea that it works, because i will use light bulbs for discharge, which dont cut out at 1v , but they sure do dimm.
and i still get better results with my light bulbs than with the machine, BUT the machine is easier, and i only deep discharge when needed.

i never thought about using a diode that stops at a specific voltage, which would be more protective.

the problems that will come from explaining the Spannangslage effect, is some nutcase with a SERIES rc pack, will discharge the PACK reverse charging some of the cells in a pack, and say, Hey that was stupid 🙂

well mabey it is stupid, and that is why were HERE, and why others can test it, on slow moving cells, and see what they come up with. if a cell is only putting out 550ma, ya wouldnt have anything to LOSE from trying, then we get more data.
 
Last edited:
Back
Top