Honda Insight Forum banner

Stick level reconditioning....conflicting information

5K views 26 replies 7 participants last post by  sasmyth 
#1 ·
Hello All,
I wanted to try and get some information on reconditioning the IMA battery at the stick level.
I continue to hear several different theories, experiences, and other ideas as to the effectiveness of stick level reconditioning.

I know that there is information already around here, but this is the problem. I continue to find conflicting information,
not only from separate users, but even the same insight central posters over the years, where they go back and forth on their
stances. I understand that the answers are difficult to pin down, and each is a case by case basis, and there isn't necessarily a definitive answer on this whole deal.

I have been looking through the forum on this topic for a few months at least, trying to determine what to do. I already have a grid charger, and while this works well, I have been having self-discharge issues with my battery, where if left for a couple of days, it drops to about 75% capacity. While grid charging seems to be working alright, it just isn't enough, and due to my living arrangements grid charging is not always an option, or able to be done due to time constraints.

I was unable to obtain the famed superbrain charger, however, I was able to get a 300w balancing charger with good reviews from amazon for 100 bucks.

Based on what I have read here, the balancing charger may help fix some sticks that are only slightly weak, but more importantly identify any truly bad sticks/cells. These sticks will need to be replaced, and then the pack should be discharged and grid charged to complete the process. Many say that this whole process is a fools errand and does not show considerable improvement over a simple discharge and grid charge.

I would appreciate any help.

highmiles
 
See less See more
#4 · (Edited)
Here is the charger in question. I tried to find the tenergy one you like so much, but it isnt in stock anywhere. This is the closest I could get . It seems
to be pretty sweet, you can pick and choose all the adjustment for charge current, discharge current, temp monitoring, individual cell voltage monitoring (terminates charging when the delta V between cells is out of user defined range????). It also graphs the charging cycle and can store 10 graphs as well.

https://www.amazon.com/gp/product/B01D4POP6C/ref=oh_aui_detailpage_o02_s00?ie=UTF8&psc=1


S Keith,
Ill also get the stick voltages for reference when I have the chance, Im sure you'll say that this means nothing without knowing those first, you seem to really like those...
 
#3 · (Edited)
Keith's ideas for my battery worked out for me, it's been about 3 months since i rebuilt my pack and it hasn't even shown me any signs of major deterioration or need for a grid charge yet.

when i put the pack back together i placed the strongest sticks in pairs with the weakest until i came to the middle of my list.

the procedure required a 50a battery load pile tester, which are relatively cheap.

success seems to rely on your sticks and how well they can handle the load once you test them. a charge, discharge and recharge cycle might bring them back to life to a degree but i didn't need to do any real reconditioning in my case since it was a bad cell that caused it all, but i think the pairings were beneficial.
 
#6 ·
....
the procedure required a 50a battery load pile tester, which are relatively cheap.

...but i didn't need to do any real reconditioning in my case since it was a bad cell that caused it all, but i think the pairings were beneficial.

I got ahold of a carbon pile tester last summer in preparation for this.
I also think I may have a couple of sticks with bad cells, it would explain the strange behavior my battery has been exhibiting, and some of it seems characteristic of a bad cell taking down a stick, taking down the whole battery's performance IMO. I dont know a whole lot though, just speculation at this point.
 
#7 · (Edited)
You can go a long way testing sticks - finding the really bad ones - at high rate discharge, especially if you're able to monitor cell voltages. Reconditioning is a whole 'nuther ball game, though. So, you got the charger, do a 20 amp discharge from full on each stick, log the discharge (preferably on cells), note the amp-hour capacity output. You should be able to spot the worst offenders pretty easily... Here's a graph I just posted of 19.5 amp discharges for all the sticks in my former pack. I'm monitoring cell pairs during this discharge. Look at the differences, spot the 'worst offenders' (there's one particular offender that is indeed the worst): http://www.insightcentral.net/forum...bat-gauge-dash-soc-bar-gauge.html#post1237129

Also, before you put anything back together, test for self discharge at the cell level: charge all the sticks to the same level, then measure cell voltages at least once daily over at least a week. Cell voltage change should be about the same for all the cells. If there's any outlying low values, consider the stick suspect...
 
#8 ·
If you've been doing grid charges/discharges, your sticks are likely as "reconditioned" as they're going to get. You have little to gain besides identifying the bad sticks. and that can be done without this level of effort as Lyger indicated.

If you're committed to this path, I recommend you return the charger.

I recommend you get this charger:

https://hobbyking.com/en_us/turnigy-reaktor-30a-1000w-balance-charger.html

It's everything the SB989 was and more.

I recommend you not use the carbon pile tester. As Jime has mentioned before, they are erratic. Getting them to hold a consistent current is a pain. I have one, and I don't use it anymore for that reason. I use the crap out of it for testing 12V batteries though... who cares if the current is ±10% in a pass/fail situation?

I recommend a BT-100 from Amazon for $20. It will put about 60A of load on a stick. IMHO, that's good enough, and it's pretty consistent.

If your goal is to identify and replace bad sticks, this exercise has value, but you need to commit to probably a couple of weeks of down time.

Good luck,

Steve
 
#9 · (Edited)
Thanks for the reply, I understand that no magic can revive a cooked battery, the idea is just to locate and replace the bad sticks. I diddled with the charger, but amazon should take it back just fine, however the charger you linked just isnt in the cards on my budget (college student). Would it be possible to attain positive results using what I've already purchased?

Im pretty committed as well, the battery has already been removed and bypassed.

Also, I said I had a carbon pile tester, but I was wrong, I actually have this guy https://www.harborfreight.com/100-amp-612v-battery-load-tester-61747.html

Its probably not appropriate for the task im guessing.


:confused:
 
#10 ·
The charger I linked is only ~$60 more with shipping. If $60 is a make or break type of dollar value for you, stick replacement is going to be a challenge - if you find any reliable ones.

The difference is the charger you have is NOT a 300W unit. It's a 300W charge unit, and a 30W discharge unit. I didn't see that it has any regenerative capability (it might). he Reaktor can discharge internally at 10A (80W). If you discharge into a 12V battery with sufficient capacity, you can at 30A (up to 1000W); however, you will be limited by a 12V power source as there are other current/voltage limits associated with those power ratings.

Actually, the tester you have is equivalent to the BT-100. It should give about 60A load on a stick, which is very useful from a diagnostic perspective, and it's pretty consistent, though you'll want to give it a minute or two of rest with cooling between each stick for consistency.

The charger you have will suffice provided you rely on the load tester to confirm the sticks do not have excessive IR. The charger will only be able to discharge at 4A, so if you intend to cycle, you're looking at about 3.5-4 hours per cycle, which means you'll likely only be able to complete one per day depending on the number of cycles you need. You may find that after 2-3 cycles, you will discover that you aren't getting any capacity improvements due to your previous pack level reconditioning with grid charges/discharges.

This is Lyger's thread where he used a grid charger and the equivalent load tester to sort through things:

http://www.insightcentral.net/forum...4-my-luck-common-theme-sitting-batteries.html

You can do that and supplement it with cycling and direct capacity measurement to improve the results.

Good luck,

Steve
 
#11 ·
The charger I linked is only ~$60 more with shipping. If $60 is a make or break type of dollar value for you, stick replacement is going to be a challenge - if you find any reliable ones.
It's not so much that the 60 dollars is a make or break, I just see it as potentially 60 more dollars in the hole that could be lost to a gamble, instead of an important book or food for a month.

I dont mind replacing the sticks, I expect to have to do that, Im just afraid of being careless and pouring all my money into a battery pit. Just trying to do this without breaking the bank.Theres a seller on ebay selling tested IMA sticks that are good for $35/stick.

Based on my IMA's behavior and resting voltage I believe I have one to three bad sticks and maybe one or two with some self discharge. Overall very fixable, but its really hampering the performance of my pack. again just an educated guess
:)
 
#12 ·
Unless that seller is a trusted member of this community, I wouldn't take it for granted that they are good.

Do you have a link?

1-3 sticks? How did you arrive at this guess? Did you measure voltage taps and note that you had 3 outliers? Is this resting voltage of the pack or sticks? If sticks, then you may indeed be correct.
 
#14 ·
Okay. That's Matt with Hybrid Revolt. My only issue is that if he's not willing to use them in his own reconditioned packs, are they truly suitable for use?

4Ah at 80A load is hard for me to believe, but I'll take everything at face value until I have other data.

Resting pack voltage tells you nothing and everything - you just don't know how to interpret it. :) 1V is the difference between a working pack and the same pack in a completely non-working state.

Sticks are rated for 6500mAh at 1.3A, but their low IR allows them to deliver very nearly that capacity at 20A.

Settings:
Charge at 6.5A
Discharge at 4A
20 minute wait between each cycle
3 cycles min/8 cycles max depending on need.
NiMH sensitivity 4mV/cell
If it has safeties, I would set maximum capacity of 9500mAh input, 50°C if it has temperature sensing.
 
#15 ·
I bought 10 of those 'tested' ones before, same specs, a long time ago when they were only like $120 or so for 10. One was a dud, the others were, well, not acceptable to my standards today, but they could probably make a dysfunctional pack work. The only way they could do 4Ah at 80 amps is if voltage were allowed to be really low - lower than what's usable in the car... As I recall they averaged about 20mΩ IR... You can calculate the voltage drop at 80 amps: say a stick is full and it's 8.4V, you hit it with an 80 amp load, voltage drop will be about 80A x 0.02Ω=1.6V, so 7V for the stick. 7 x 20 sticks=140V for the pack. But it's a little different than that in the real-world, as the full 8.4V value is short-lived; the real starting voltage is more like 7.8V, so 7.8V minus 1.6V=6.2V, and 6.2V x 20=124V. So maybe these sticks can eek-out a few seconds at near full assist, when warm, primed, etc., that is, if they really pulled 4Ah at 80A on the bench...

Some of Steve's setting suggestions seem too aggressive to me (50C temp cutoff, 4mV delta V, 9500mAh capacity cutoff).
 
#17 ·
I routinely watch cells in operation at 45-50°C.

SB989 default -dV is 5mv. Typical recommendations are 5-20mV/cell. 4mV is Reaktor default and is more conservative than most recommendations.

9500mAh is just a safety. I routinely see charges into the high 8s low 9s for really strong subpacks.

Temp and cap limits are only for if there's no -dV trigger.

I had forgotten I had measured the tap voltages a week or two ago. all of the sticks, save 2-4 outliers, hung around 7.5 to 7.6 volts, with the packs resting voltage hanging around 160 after sitting a while. The outliers ranged from like 7.3ish to like 6.9 v. These numbers on top of the resting pack voltage is where I got the guess of a few sticks from.

This is all just off the top of my head, some of it may not be quite accurate. I can check again before I start cycling, note their positions, do voltage measurements of each cell, things like that to determine definitively which sticks and to get a baseline.

I dont think the battery is irredeemable, it just seems like those 2 or 3 sticks may be hampering the whole pack.

thoughts now that I have some numbers?
"off the top of my head" isn't the best quality data.

In most cases, all outliers are bad. Some may recover to the point that periodic grid charging renders them usable.

You can get a lot more insight into what's going on with the sticks by carefully piercing the shrink with your voltmeter leads and check the individual cell voltages. There should be very little variation between them.
 
#16 · (Edited)
I had forgotten I had measured the tap voltages a week or two ago. all of the sticks, save 2-4 outliers, hung around 7.5 to 7.6 volts, with the packs resting voltage hanging around 160 after sitting a while. The outliers ranged from like 7.3ish to like 6.9 v. These numbers on top of the resting pack voltage is where I got the guess of a few sticks from.

This is all just off the top of my head, some of it may not be quite accurate. I can check again before I start cycling, note their positions, do voltage measurements of each cell, things like that to determine definitively which sticks and to get a baseline.

I dont think the battery is irredeemable, it just seems like those 2 or 3 sticks may be hampering the whole pack.

thoughts now that I have some numbers?
 
#19 ·
Again, 4mV/cell is conservative by most recommendations. You know this. I've run with as little as 2mV, and false -dV detection can be a problem.

Temperature increase is actually a major contributing factor to the -dV phenomena. You should know this.

If you rely on -dV to any degree, you are overcharging. Period. You're splitting hairs when you talk overcharging if you ignore this.

Most sources say NiMH charging is 66% efficient when you figure across the entire SoC range. My figures are closer to 77%. If you're able to get 6500 following a waiting period out of only 7000 input, then you have the most magical cells on the planet, or you have bad data. I have correlation on 35 chargers on my 77% number.

When I'm talking 45-50°C, I'm talking OPERATION, i.e., IN-CAR driving. Granted, the Prius runs warmer than the Honda, but only by about 5°C. The statement was intended to establish safe vs. non-safe. In bench testing operations, most don't exceed 45°C.

MOST of the newer chargers have the habit of terminating all operations when a limit is hit (like the reaktors). I'm assuming that he isn't going to be standing around watching them for hours at a time, and he's going to want to get useable data rather than have it lost when the charger terminates (like the Reaktor). I'm also assuming his charger behaves like most everything but the IMAX B6.

My recommendations for safeties are:
1) above where you would ever want them to be, but not so high that one would expect damage.
2) yet still high enough to not impact normal testing.
 
#20 ·
Again, 4mV/cell is conservative by most recommendations. You know this. I've run with as little as 2mV, and false -dV detection can be a problem.
Don't you remember this post/info?: http://www.insightcentral.net/forum...attery-stick-reconditioning-9.html#post923161

Beyond what Eli has to say there, I've found, over hundreds of cell and stick charges, that there's nothing to be gained from letting the charge go on until 4mV delta V per cell... On the contrary, there's a higher risk of venting... False delta V can be an issue, true, especially with crusty sticks. I don't even use delta V, so, perhaps you would have more to say about 'cycling' - since you actually do it...

Temperature increase is actually a major contributing factor to the -dV phenomena. You should know this.
'Bonafide delta V' isn't temperature based; you can have delta V without temperature change. Yet, of course, you do see a drop in voltage with increases in temperature, and increases in temperature happen in earnest when cells near full charge. So it's hard to untangle the two, or easy to mistake one for the other...

Most sources say NiMH charging is 66% efficient when you figure across the entire SoC range. My figures are closer to 77%....
For Insight cells I've found it's easily over 90% (Coulombic efficiency) if you exclude the top ~3% of charge, while I've found no benefit and the risk of harm if you continue to charge - especially at full current - that additional ~3%. Charge to full if you drop the current, otherwise, don't try to slam current into it just for that additional ~3%, or in some futile attempt to balance the cells...

Automatic cycles with delta V detection is a brutal way to 'recondition' sticks. Some sticks do need gentle cycling to get the juices flowing, like sticks that have sat unused. Otherwise it's mostly about balancing the cells, and using automatic cycling with delta V to balance cells is a pretty lame strategy... One would be better off building a cheap and easy cell-level discharge rig, discharge the cells slow, fully, deeply; then charge the whole stick to full; and maybe do a couple extra cycles with more liberal settings. That way you don't have to worry much about grossly overcharging any one cell... One can typically start a charge on a stick with some cells near empty and others around 60%: the 60% cells get waaay overcharged with a set-and-forget method - automatic cycles with delta V and otherwise too aggressive settings... It's a pretty brainless way to go about coaxing more life out of sticks, like bludgeoning them with a hammer, repeatedly, yelling, "work, damn you, work!"

Anyway, always happy to add to the growing trove of "conflicting information."
 
#21 · (Edited)
Long Discharged Sticks

Well given the thread title, I guess this is a good place for another small bit of confusion. Someone mentioned, not sure - couldn't re-find the words, that sticks become unbalanced badly as the self discharge. Said person also seemed to prefer a low current charge to "wake" up the weaker cells. I never gave this a great deal of thought until in read that comment this evening.

Just for laughs, It occurred to me that I could simply check the cell voltages BEFORE I started my usual 3 reconditioning cycles (prior to high current testing). Welllll........on the first stick I tried I see that it is badly unbalanced, in fact two cells are so low in voltage as to be considered dead from a performance view, not a surprise since it has been sitting for a long time.

Voltages were 1.27, 1.19, .54, 1.16, .55 and 1.27 volts.

Started me wondering about best way to initially charge such a stick. I'm sure that those cells will register 7V + as soon as I hit the charge button, but is there a better plan???

I usually charge at 2A, but now I'm wondering if a lower current would do a better recovery job.
 
#22 ·
As I mentioned in the post above yours - cell discharge rig, or just place a small resistor on each cell, clamp them there... i.e. discharge the cells individually then charge... Of course, you don't need to discharge the already-self discharged cells...
 
#23 · (Edited)
Woops, sorry got bogged down in the quicksand and lost it right there in front of my eyes:(

You also mention that for long stored sticks you would wake them up with gentle cycling. Two questions come quickly to mind: a. what charge/discharge current are you indicating, and b. got any data or just a kinda experiential based feel that one needs to be gentle?

I really don't mean to press too hard, I know how hard it is pin down the process.
 
#24 · (Edited)
If you're asking about what I say above about 'gentle cycling to get the juices flowing', that's mainly based on experience. The most important aspect is to be gentle at the top, like not doing a full delta V-based charge on the first charge. It's super easy to vent cells that have sat, even before cell temps get very warm at all...

Just about the worse thing one could do is pull a stick off the shelf and set the cycling device to 3 automatic cycles with a 0.9V/cell discharge cutoff and a delta V of 4mV (the 'old-school, boilerplate' recommendations). That's a recipe for failure: You start with a very likely imbalanced stick - like the one you mentioned above, with uneven cell voltages - immediately hit it with say 6.5A current, and then the charge doesn't stop until some ill-defined cutoff/s that can be the result of various not-so-good conditions, like a 24mV voltage drop caused by one cell overheating, maybe all cells overheating a bit, etc. I mean, usually we only have one temp probe anyway - it's just a roll of the die if that one cell happens to be the hottest and triggers the temp cutoff, if you have one set...

Much better to first discharge cells individually (resistors on cells), do something like a <1.3A charge for around 500-1000mAh, increase current to say 6.5A until a total of around 6000mAh has passed (or perhaps voltage reaches about 8.7V), then drop current to around 500-600mA and let it go until about 7600mAh total has been input. Sounds terrible, I know. My Reaktor though basically automates most of this. It does a super-low current pulse charge until voltage reaches about 1.3V/cell, then it will ramp up to the user set current. Then it will go until a user defined voltage threshold is reached and taper current...

Alternatively, don't charge to full on the first charge, maybe take it up to no more than 1.46V per cell at 6.5A. Maybe use a current lower than 6.5A for the charge and not go totally full. Etc... Use a moderate voltage threshold for discharge, too. I usually do the discharge until first cell pair reaches 2V, so roughly first cell at 1V and the discharge stops. If you don't have cell-level capability, keep the threshold a bit higher, like around 1.1V per cell so 6.6V total. 0.9V per cell will likely drive one cell too low, with the possibility for high current reversal, too. Also, note that my voltage cutoffs are based on measurements at the cells, with little lead resistance; if your voltage values are on the main leads, then you have to adjust your cutoffs to account for lead resistance. Typically it won't make too much difference, though (unless it's a Superbrain)...

In a nut shell, on the first cycle of sticks that have sat for a while, it's best to keep away from the 'tails', the extreme top and bottom...

After this stuff you can do whatever kind of cycles you want, more or less...

I guess most of this applies to any sticks, really, not just ones that have sat, mainly because even sticks recently used in the car tend to get so imbalanced (that's likely the reason the pack is failing in the first place). But for ones that have sat, the risk of venting, etc. is worse...
 
#25 ·
Thank you eq1 for your thorough response. I read it carefully and intend to read it again.

Unfortunately, I'm still on the dark side with the Rodney Dangerfield chargers, but I have modified the lead setups considerably, eliminating the alligator clips and the plastic side plugs. Also, as discussed some time ago, I do not rely on the meters but do my measurements at both the cell level and the stick level using other meters. Since the SB meters aren't accurate, I compensate with the settings to get the starts and stops I want. I also set the battery size to control charge amount, rather than relying on delta V. Anyway, I think I've gotten around most of the issues, but I do need to watch the cell meters a bit better.

Already, I do do an initial low current charge of 1000-1500mAh. Kinda seems right to me too. I am perhaps a bit heavy handed with the successive cycles. Really tough to know about possible venting since I try not to disturb the heat shrink and have not yet watched the sticks late in the charge cycle. Thanks for the warning.

Again, thanks for your effort in the detailed write-up. I need to read it another time or two.
 
#26 ·
As always with a discussion forum, you need to make judgements about the veracity of the data being given, and the credibility of those giving it.

I don't have an iron in the fire (my car has no IMA or battery pack) but if you watch the forum long enough, you will see people who ride battery hobby horses like they're in an Olympic event, and others who make up their own unique theories - free of any constraints like peer referencing or even scientific common sense.

It's not of much help to you, but I suggest that you PM those who have a history of achieving battery outcomes in a commercial sense, and who (via PM) may be prepared to share some of their actual, demonstrated wisdom with you.

And a contentious final comment: the people who are really achieving seldom have time to expound long-winded battery theories here: they're too busy doing it for real!
 
#27 ·
This is a 3 year old thread but thought I might tag my advice to the new owner of my 2001 CVT regarding doing stick level balancing of the pack. I couldn't find a more relevant thread and was reluctant to start "yet another conflicting" one. I'm hoping that the new owner can follow my advice as well as using this thread to find other's with complementary (rather than conflicting) advice.

My car had a fresh (reconditioned) pack installed around 120K by the previous owner under Honda's graceful extended warranty coverage not long before I bought it. It did not begin to show IMA problems until about 180k miles at which point I began to get occasional IMA light/codes. I began to reset with a 12V power cycle and learned to avoid triggering the IMA, usually through using the IMA too heavily when it was hot... After a year and 10k miles or so it began to get worse and with the help of another Insight enthusiast we pulled both of our battery packs and did several charge/discharge cycles on each stick with a 2 week rest and full discharge to measure self-discharge. My sticks ranged from 4500 - 6300 mAh with only about 4 under 5000 mAh. The multiple charge/discharge cycles did seem to improve the range of each stick and the lower capacity ones (<5000mAh) were also the ones with the highest (percentage) self-discharge. I reassembled the pack and viola no more IMA lights... for a while. Meanwhile my friend who had rebalanced his (which were in better condition overall) decided to replace with a Bumblebee so we swapped his pack into my car, pulled my sticks form my pack and sent the empty "core" to Bumblebee (which is all they want anyway, they were not re-using old sticks). This pack worked well for another year, and I did a complete "rebalance" and "evaluation" of my original sticks whose condition was roughly the same (by measure) as when I'd worked with them before. My intention at that point was to pull the "better" pack, rebalance/test the other cells with my particular method, and put together a more balanced set from the two packs.

Meanwhile I was "overcome by events", bought a 2011 Chevy Volt that was due for a battery replacement and it became my hobby and daily driver replacing my Insight for the most part. Lack of regular use of the Insight only aggravated the IMA condition. I also began to use the spare NiMH sticks from the Insight for other projects like strapping up 5 to drive my 36v electric bicycle and a couple of water pump and fan projects with solar charging.

What I delivered to the new owner was the car with an IMA light that tends to come on nearly every time I drive it unless it is both cool and I'm excruciatingly careful not to push the boost or regen very hard. I included 8 extra sticks, several which I had tested just before I delivered last weekend.

The new owner has a 2002 MT with a grid-charger and is going to be installing a second harness and exercising/grid-charging the CVT before going to the effort of pulling/testing/balancing the pack.

For his reference here is my method for exercising/testing the sticks: I used a GOOL RC balance charger set with their NiMH profile and a 3A (.5C) charge and 1A (< .2C) discharge. My protocol was to start with 2 full charge/discharge cycles, followed by a full charge, a 2 week rest, and a full discharge to measure the self-discharge. I recorded the last full charge, and the last 2 discharges to characterize the condition of each stick. My plan (and recommendation) was to remove the 8 worst sticks from the 28. Worst would be defined by the lowest full-discharge measure before resting (first) and then the lowest discharge (after the rest).

My "theory" is that after the charge/discharge cycles, the last discharge is the amount of useful power the stick can hold when "fully charged" and that the rest defines the amount of self-discharge (from internal nickel whisker formation?). With regular grid-charging, the self-discharge may not be as important by itself. Hopefully there will be 20 good-enough sticks between the 28 that I sent along.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top