NiMH drain down to 0V - some initial observations

[apagogeas]

Following my recent post
http://www.candlepowerforums.com/vb...-August-2012&p=4014686&highlight=#post4014686

on draining a NiMH battery down to 0V, I made a more thorough test. I have 4 Energizer NiMH 2450 rechargeable HR stamped (2 years old) with the following BI values (IRC in parenthesis) measured every 6 months (the 1st value is the initial BI before first use):

B1: 2268(1.47), 2162(1.52), 1993(1.51), 1951(1.52), 1875(1.55)
B2: 2232(1.48), 2137(1.51), 1959(1.50), 1924(1.60), 1856(1.53)
B3: 2291(1.47), 2193(1.50), 2024(1.52), 1977(1.50), 1890(1.53)
B4: 2313(1.48), 2214(1.50), 2042(1.55), 1994(1.50), 1906(1.53)

The last BI performed two days ago with a 2000mAh setting. After the BI, I discharged them in an old charger capable to drain the batteries down to 0V. I estimate it discharges around 200mAh each hour and at the end of 20 hours discharge I measured the voltages on all cells (whilst in discharge) and it was 0.03V - 0.08V. I took them out of the device, let them rest for a few hours and performed another BI with the same 2000mAh setting. The BI results are:

B1: 1989(1.61) vs 1875(1.55) -> got back 114mAh
B2: 1974(1.58) vs 1856(1.53) -> got back 118mAh
B3: 1959(1.60) vs 1890(1.53) -> got back 69mAh
B4: 1982(1.57) vs 1906(1.53) -> got back 76mAh

Overall, after that experiment I see a noticeable boost in capacity and a slight increase of internal resistance but I can't really suggest that for the resistance since I have seen such variations many times (also can be seen at the initial BIs on top). The test continues and I'll perform another drain to 0V in a few days and report back the findings. For the history these batteries have very regular use and treated nicely (except of course the 0V drain!).

UPDATE 20/9/12
The same procedure followed, drain down to 0V rest and BI with 2000mAh. The results:

B1: 2009(1.65) vs 1989(1.61) -> got back 20mAh, overall got back 134mAh
B2: 2013(1.59) vs 1974(1.58) -> got back 39mAh, overall got back 157mAh
B3: 2006(1.59) vs 1959(1.60) -> got back 47mAh, overall got back 116mAh
B4: 614(1.58) vs 1982(1.57) -> Not sure what happened here and got only 614mAh. I currently discharge @400mA to check the result again of this supposedly fully charged battery after that BI.

The voltage under load is around 1.14-1.17 on B2-B4, close to what I used to get all that time so nothing strange here. B1 on the other hand showed 1.04-1.07V during discharge which I observed that (it was around 1.07) during the initial BI before attempting to do the very first 0V drain, this one probably started to give up the last 6 months. Apart from the IRC 1.65 of B1 which further increased, I see no variation in IRC for the rest of the batteries. So, what seems very strange is the B4 capacity which I'll wait for the discharge to verify, might be a contact quirk. Quite mixed outcome really, from one end we have a definite increase of capacity (material reactivated probably) and on the other end some indication that internal resistance might increase. I also agree with the OP in checking IRC with the normal 0.9V discharge. I'll do that after I complete the 0V tests. I'll also give more time to the batteries to recover in normal use and test again.

UPDATE 20/9/12
The discharge of B4 battery returned
B4: 2009(1.58) vs 1982(1.57) -> got back 27mAh, overall got back 103mAh

UPDATE 12/10/12
So, I used these batteries normally and decided to do another normal BI.
B1: 0(2.08)
B2: 1943(1.58)
B3: 1951(1.57)
B4: 1978(1.59)

B1 really got damaged, the other batteries seem to perform ok. The capacity gained seems to go away quickly. Overall, I see no real benefit by doing a 0V drain to batteries; more surely if a battery is weak it will totally destroy it (like B1 here), stronger batteries will simply take some abuse but the benefit is minimal. So my conclusion is, very deep discharges will damage the batteries (at least the older NiMH, not sure what is the case for LSD) and will offer a marginal increase of capacity which will fade away quickly, so no real benefit going for 0V discharges, more to lose than gain by doing so. That contradicts a bit to what I have observed at my initial post (about voltage drop) on top but those batteries were new ones, so possibly what I observed was naturally what the battery would be after a few cycles, not because of the 0V drain.

UPDATE 23/10/12
I didn't throw yet B1, I put it in a cree led light and had very light usage. Obviously the battery is fully charged from the last BI on 12/10 but maha says 0 capacity because it discharges in bursts of 1A, so this battery is completely incapable to provide that current but it can of course provide lower current. I attempted to check the IRC of B1 and it climbed to 2.74! So, within 1 month the IRC gone all the way up from 1.65 to 2.74! I've never seen any such steep increase of resistance in the past within one month only. Obviously the first jump from 1.65 to 2.08 occurred due to 0V drain (damaged the cell), the other huge IRC jump (2.08 -> 2.74) occurred by the BI on 12/10 (or at least boosted the existing problem) - I always start a BI at discharged state, so BI charged the cell, attempted to discharge it but failed due to 1A bursts (thus the battery is still fully charged) and then put in another 3000+mAh in a fully charged battery! So, this heavily implies that when we use the BI function, we MUST ENSURE the battery is fully empty (normal drain to 0.9V) before attempting to rescue it, otherwise we heavily stress the battery and if it is a weak one, it dies right away.

UPDATE 11/03/13
It has been about 6 months since the test. B1 has been recycled. B2, B3 and B4 all now register HIGH in maha (all show 2.3 IRC and above). They have been charged a couple of times during that period and the observation was at every charge a rapid increase of IRC (1.8, then 2 and now 2.3 and above). I have to say that all four batteries have been treated nicely before the test so I mainly attribute the outcome to the test performed.
So from the above I understand that draining a NiMH battery down to 0V surely damages the cells - no exceptions to that. It just doesn't occur right away after a 0V drain but doing so starts an irreversible degradation process. No clue what is damaged inside, perhaps experts can explain this? I suppose this also has to do with the construction quality of the cell and the specific ones aren't the best to judge this; Eneloops might be more capable to withstand such a damage but no matter what, there is bad damage and stress to the cell regardless of make. So, avoid draining even a single cell alone down to 0V (and probably even higher than that) if you want long living batteries.

 

[stp]

Thanks for the info and waiting for follow up.

 

[NoixPecan]

a slight increase of internal resistance but I can't really suggest that for the resistance since I have seen such variations many times

All the more as you cannot really compare the ICV (impedance check voltage of the C9000) when starting at two different states of (dis)charge. It might be interesting to perform an additional break-in (after a normal discharge and a rest) to confirm any change of internal resistance.

 

[apagogeas]

Updated info on first post. I agree with your suggestion NoixPecan, I'll do that at the end of these tests.

 

[apagogeas]

Updated info on first post. My conclusion is that 0V drain heavily damages the weaker batteries and gives no real benefit to stronger ones. So it is not a good practice at all. Well, we learn something new every day. Now I have to prepare the "funeral" for battery No1 🙂

 

[stp]

Thanks for the info, and lots of respect for you.

 

[apagogeas]

Some more info on first post, this time the culprit is the BI if not used the proper way. I have read in the past here that it doesn't matter if the battery is partially charged before a break in. My observation suggest that it does matter very much, especially if we treat a weak battery! Given the BI is also used as an attempt to rescue weak batteries, the outcome contradicts the essence of BI when dealing with a partially charged cell, even worse if we can't discharge it due to 1A bursts of maha. Unless if what I observed is simply the ongoing rapid internal deformation after the 0V drain, I conclude the best practice for long living batteries and BI is to fully discharge the battery before doing the BI. Now I really have to let that battery R.I.P., I have to show some respect for the loyal 2 years of service... :thinking😱r not? :naughty:

 

[Isaiah6113]

. . . I conclude the best practice for long living batteries and BI is to fully discharge the battery before doing the BI.

apagogeas, this is very important information, thank you for your detailed testing and critical conclusion. For a noob like me this is invaluable.

I fear however, that I may have irrepairably damaged some new cells. I have recently completed BI on seven sets (4 each) of new batteries. In my ignorance, out of the package and into the C9000 for BI. For what it's worth all experienced capacity gains, however if I understand correctly these gains may be superficial due to internal damage that I can't see. :hairpull:

And that's why CPF is so important, it's a constant learning experience everytime I come here. lovecpf


Thank you,

Matthew

 

[apagogeas]

Some abuse certainly taken by that break in, new batteries can stand such abuse but surely boosts internal damage which will show up sooner than expected. I make it a habbit to always discharge new batteries before break in, mostly due to logical reasoning that since BI on its own puts 1.6x the capacity, if the battery is charged, this is by definition huge overcharging; no different to normal dumb chargers that toast the batteries inside. My observation reinforced this logical reasoning and at least for the case of that weak battery showed it is very damaging. Additionally, break ins should be used sparingly because of this abused approach i.e. once every 6 months at most. It helps rejuvenate the battery by saturating all the particles that might got lazy or some other sort of problem being fixed but at the same time it is abusive by definition. Don't worry too much, it is batteries after all.

 

[Isaiah6113]

Some abuse certainly taken by that break in, new batteries can stand such abuse but surely boosts internal damage which will show up sooner than expected. Don't worry too much, it is batteries after all.

Oh well apagogeas, one lives and one learns. Due to your efforts I have learned! Good news I hadn't run a BI on all my new batteries, so now I know, discharge first 🙂

This is information that Maha should add to their manual for the C9000. It's critical!

Best regards,

Matthew

 

[apagogeas]

Updated info on first post. Draining down to 0V damages the cells - avoid at all costs.

 

[snakebite]

i did the same for some 7ah panasonic d cells.were in a solar light that was always topped up and seldom ran any great length of time.
i decided to test these old cells which are from 1998 if i read the datecode right.well fully charged they returned 3.9ah testing with a triton at 1a rate.
recharged and did it again.accepted 4.4ah.returned 4.1.
clipped a pr2 bulb to each and let them go down.left them like this a week.
accepted 7.5 and returned 6.9!
impedence dropped from.07 to .00 on my **** smith esr meter.
so if the cells are sick already nothing to lose and everything to gain.
now the energizers have been fragile since they came out with the 2500.i imagine the 2450 being just a more honestly rated 2500.
dont recycle them yet!
see if they develope high self discharge too.

 

[apagogeas]

Actually the Energizers were performing great for their age just before the test. They just shown the normal lower capacity (info at the first post). Right after the test they all collapsed. There is nothing I can do to rescue them. The IRC got that high because of the 0V drain and this damaged them beyond recovery. What is interesting is the way this damage occurred (relatively slowly) but once started it can't be stopped. I can't check self-discharge really because they all return 0mAh by maha (due to 1A discharge burst this charger does), they are really beyond recovery but the OC voltage is at 1.29V which means they have quite good charge left after 1.5 month sitting idle (from the previous charge and use).

 

[snakebite]

now lets drill n fill.
see my post on the d cells i rehydrated.
i would try 2 or 3 drops of distilled water for a aa cell.
did your discharger discharge the cells independantly?
if in series reversing one or more is rough on those cells.

 

[apagogeas]

did your discharger discharge the cells independantly?

Yes, I estimated about 220mA drain per hour. Independent discharge. Cells never reversed charged, simply drained down to 0V.

now lets drill n fill.

I wouldn't care to try and rescue them using such techniques. What I tried to figure out here is if draining down to 0V can help to revive cells that have lost some capacity as these Energizers were experiencing around 1900mAh (lost about 350-400 mAh since brand new) just before the test. The conclusion is it doesn't really, the effect is temporary and even worse, it ruins the cell completely.