Honda Insight Forum banner

1 - 20 of 409 Posts

·
Registered
Joined
·
232 Posts
Discussion Starter #1 (Edited)
I have read a little on the forums about cell reconditioning. Recently my IMA light was coming on pretty much every day for about two weeks in a row. I didn't have the time to do anything about it, and just had to keep using the car. So, i may have totally killed the IMA pack, however, I hope to be able to rejuvenate it by doing a stick level reconditioning. I have built a trickle charger, but I have the opportunity now to not drive the car for a few weeks, so i think the time is right to tear the pack apart.

Now the problem.. I can't find a Super Brain 989 in stock anywhere.. What is the best alternative? I am trying to find something pronto, so I can kick of the reconditioning process. The best alternative I can find seems to be the DuraTrax ICE.

Assuming I go for the DuraTrax, I see it will have lower discharge capabilities, something like 10 Amp. I read some old threads about them melting down at that level, so I would guess 6.5 Amp might be better and that puts a discharge at 1 hour. So, that's about a 1 hour charge + 1 hour discharge per stick. 4 cycles = 8 hours per stick. Perfect to set one in the morning, and one in the evening. Two sticks per day and it will be done in 10 days.

Am I thinking this through correctly? Missing anything? How will I identify sticks that are totally shot? Do these chargers display a calculated Ah rating?

I'll keep reading myself to see if I can find guidelines to know what internal resistance levels are no good, I saw several methods to determine internal resistance, but nothing yet to know what good or bad numbers are.

Appreciate the help, and yes I am sure someone else has posted this stuff before... But this forum is a pain to search, and sometimes it's just nice to get the problem out there in your own words..
 

·
Registered
Joined
·
6,840 Posts
I glanced over the Duratrax specs and it looked like it'd work. The only thing that caught my eye was the discharge cutoff voltage range, 0.8V being the low end. Normally that'd be fine, as we have typically done 0.9V. But lately there's been speculation and some testing exploring really deep discharges, like down to 0.3V (check out Mike D.s grid charger thread)... So, I personally would not buy the Duratrax if it wouldn't let me discharge below 0.8V... Also, note that you need to buy a separate power supply for the Duratrax.

There's a Super Brain 989 at this site, but it seems expensive at $150:
MRC Super Brain 989 Delta Peak Charger - Airsoft Battery Charger - AirSplat.com - AirSoft Gun Warehouse

Sounds like you're on the right track. I'd keep looking for other chargers if $150 seems exorbitant (I got mine for $75 about a year ago)...

I was checking out the PowerLab chargers a while back. More expensive, but more features:
Cellpro PowerLab 8 (EC5 version)

As far as charge and discharge rates are concerned, 5 amp charge and 10 amp discharge would be fine - the Duratrax said it was limited on the charge side, I think it said 7 or 8 amps. Some people say discharge as fast as possible, others don't seem to follow that. Personally, I'd probably do a 6.5 amp charge and discharge (the last time I did 10 amp discharge, 7 amp charge)...

On identifying sticks that are shot, you'll graph the results of your successive cycles and be able to see which sticks performed the best and which the worst; which gained and by how much, which didn't. I'd do 3 cycles exactly the same, then look at the results before moving forward: some sticks might be fine and won't need anything else, while others will need more... You can calculate IR from data you note while doing the cycling, but it's mostly good for comparison across sticks rather than absolute values. Keeping your test methods exactly the same is most important. You can also calculate self-discharge if you wait a while. All-in-all, the graphs seem the most tell-tale in determining which sticks are the weakest; the IR and self-discharge calculations can help interpret things a bit.

Here's a couple threads I used when doing my pack:
Analyzing, refurbishing/repairing a dead hybrid battery pack (2000 Honda Insight) - Fuel Economy, Hypermiling, EcoModding News and Forum - EcoModder.com
http://www.insightcentral.net/forums/modifications-technical-issues/19283-spare-ima-pack-rebuilding-testing-ima-battery.html
 

·
Registered
Joined
·
2,015 Posts
I have to say that the PowerLab 8 is great, though expensive.

I recommend at least 3 cycles on each stick. This will be about 9 hours per stick.

The reality is you'll do some additional testing afterwards. You can easily spend 2 weeks or more on this project.
 

·
Registered
Joined
·
232 Posts
Discussion Starter #4
Thanks for the feedback. Went ahead and ordered the duratrax.. We'll see. I have the time to spend on it now as long as I start soon. As far as deep discharge, yeah that's a disadvantage, but it should get the job done even if its a little less effective.

After looking at results do people just replace the weakest stick, or is there a rule of thumb for a performance threshold below which sticks should be replaced?
 

·
Registered
Joined
·
6,840 Posts
I don't know if there's a rule of thumb here, in terms of super brain methods. There might be if one were doing real internal resistance calculations, though... The general strategy would have you replacing the weakest sticks, basically. You want them all to match as much as possible, yet I think it's pretty difficult to go too far along these lines with the super brain cycling methods. My understanding is that the pack will be limited to the performance/capacity of the weakest sticks - so you just replace the weakest sticks. How weak is too weak depends on your usage...

My weakest stick was only putting out about 3900mAh after the cycling, but I never replaced it or similar ones. I probably have about 5 sticks that are dragging the pack down. But still, I can pull enough out of the pack that it's pretty useable. Actually, I get just about that -- 0.5 X 3900 -- out of the pack before a negative recal (the 0.5 reflects the roughly 25% to 75% state of charge window that the car operates within, i.e. 50% of the overall capacity). So normally that'd be 0.5 X 6500, or 3250mAh, and I'm only getting about 2000mAh... There's a little more to it, too; depending on the failure modes going on, like what's wrong with your pack, sticks, cells, the performance of the pack can vary quite a bit within the limited capacity you do have. So maybe you can squeak out say 2000 mAh, but half of that would be only under modest levels of assist, such as 20-30 amps instead of 40+ amps...

So, I don't know. I'd say, at the end of your cycling regime, replace any sticks that are putting out less than 4500mAh at 10 amps... And/or pick out the sticks with the most lackluster curves (after graphing results). For example, among my 20 curves, 5 or so clearly were at the bottom, sort of clustered together, whereas 12 or so were clustered together higher up, and 2 or 3 were at the top...

Here's 4 graphs of 4 packs. Mine's at the lower right:
 

·
Premium Member
Joined
·
1,069 Posts
.....On identifying sticks that are shot, you'll graph the results of your successive cycles and be able to see which sticks performed the best and which the worst; which gained and by how much, which didn't. I'd do 3 cycles exactly the same, then look at the results before moving forward: some sticks might be fine and won't need anything else, while others will need more......
To add to what eq1, and elaborate a little.....

I made the mistake of going through several spare sticks, looking for one to replace a very weak stick capacity wise in my pack.

So naturally, I charged and discharged each stick looking for one that had similar capacity as most of those in the pack.

Found one, swapped it into the pack and viola, ..... had a weaker pack than before!?!?

What happened????

I neglected to measure the self discharge of the new stick, ..... and to put it simply..... it's terrible. By far the worst one in the pack. My pack is pretty much unusable at this point, because this particular stick discharges so much in just 24 hours, that MIMA stops Assist at 154 volts or so, at least the last time I checked it. This stick has completely un-balanced the pack.

So with that in mind, I would recommend the following steps to refurbish a pack.

1) Perform charging and measure the Ah.
2) Perform discharge and again measure Ah.
3) Measure self discharge (very important, just as much as those above).

Jim.
 

·
Registered
Joined
·
232 Posts
Discussion Starter #7
Great tips. I think self discharge is a big issue in my pack currently because of the frequency of the IMA lights even with forced charging. I'll report my findings in a couple weeks I suppose.
 

·
Administrator
Joined
·
11,753 Posts
All good advice.

As we know the three main IMA battery issues are.

1) Reduced Ah capacity. (permanent or temporary). Leads to poor assist/regen duration.

2) Increased internal resistance. Leads to limited assist/regen current capability, increased self heating, poor efficiency.

3) Varying self discharge rate. Leads to battery imbalance, poor assist/regen duration and poor efficiency.

All the above can also set IMA codes.

So when we asses sticks for reuse my plan would be.

1) Cycle them all first at say 5-10A to determine maximum capacity and to try and restore lost capacity.

2) During cycling if the charger has that option note the battery IR internal resistance reading.
(The Robotronic Overloader II has the IR function I don't know about the others)

3) After cycling, fully charge sticks then leave for a week (24hrs minimum) then discharge and calculate the self discharge rate.

4) If you have a high current charge/discharge capability ~50A then test sticks and note any that can't supply/accept at least 50A
(Most people won't have this capability)

With the paramaters 1-3 above we can now make reasonable choices about which sticks to use to rebuild a pack.
(This assumes you have some spare tested sticks) If you don't then it's all largely academic. :(

In order of importance I put. (Other may not agree)

1) Internal resistance. (Select the best low IR matched sticks, you want full assist/regen and low self heating) If you have a single wildly low IR stick it may not be best to select that. It's compromise between low IR and matched IR.
2) Ah Capacity. (Select the sticks with most capacity, you want good assist/regen duration)
3) Self discharge rate. (Select the sticks with the lowest self discharge rate, you want them to stay in-balance if possible)

Why?

The imbalance caused by varying self discharge if not too acute can be dealt with using a cheap grid charger using a weekly balance charge. That's why I give it a lower priority. It's place in the list could change if you have some really rouge self discharge sticks.

With any rebuilt pack IMO a cheap grid charger is a pre-requisite to keep it going for a reasonable time.

I hope that helps. :)
 

·
Registered
Joined
·
6,840 Posts
Great tips. I think self discharge is a big issue in my pack currently because of the frequency of the IMA lights even with forced charging. I'll report my findings in a couple weeks I suppose.
I don't think you mentioned what codes you're getting. What codes are you getting? Those can help indicate what's wrong with the pack... Not sure which code tells you what though. I had a 1449-78 and was told that that was the 'lowest on the totem pole' and that the battery should be balanced because, otherwise, I would have gotten another code first, the one that has to do with imbalance. Not sure how true that is, though... My reading is that they can all be inter-related, one problem leading to others...
 

·
Premium Member
Joined
·
2,890 Posts
As we know the three main IMA battery issues are.
Agreed 99%.

My only personal tweaks ( just my personal preferences ).

  • I would set the Self Discharge Minimum wait period to a week ... Usually I do a 1 month ... and going through 20 sticks can take easily take a week to do the initial test , so the first one tested will be ready for it's self discharge test by the time the 20th one is done with the other tests.
  • I usually recommend starting off first with a discharge test for the pack coming out of a car ... it gives a snap shot idea of where the SoC balance was between the sticks in the car ... if one starts off with a charge that initial SoC balance data is lost.
  • Consistency is #1 priority for good comparisons:
    ie ... I would recommend against comparing the capacity or IR from a 5A test with one from a 10A test ... or ... comparing Self Discharge rate determined from a 1 week wait vs a 3 week wait... etc... I think it matters less what the specifics of the test are , be it 5A or 50A ... be it 1 week or 2 months ... I think it is more important for the comparison to only be comparing between similar ( ideally the same ) test conditions / methods.
  • Lastly ... I think matched is nearly ( if not as ) important as how strong or weak for any of the 3 testing criteria sticks are ... ie a pack of 20 sticks all the same ... will over all function better than a pack with only 1 stick showing significantly better or worse capacity ,resistance , or self discharge from the other 19.
  • Although perhaps not necessary , anyone who wants to get into this more than once or twice for a 20 stick set ... I would recommend a PC link data device ... if the charger itself doesn't do this , set up some other PC link device to record the Voltage and Amps over time of the cycle ... with those two a spread sheet can easily calculate the Ah or Wh of capacity , the IR , etc ... and it makes saving and comparing the data much easier.
 

·
Registered
Joined
·
272 Posts
I agree. It is really hard sometimes to keep all the parameters straight and this summation helps especially for us who have not been down at the disassembled battery level and/or have the equipment to do the tests.
 

·
Super Moderator
Joined
·
8,209 Posts
What is the best strategy for keeping the voltage taps happy? There is a tap for each two sticks in series, if my reading of the schematic is correct, so there is probably an optimum way to pair off the sticks for the best and most trouble free performance (i.e. to get the fewest recals and codes).

I have a spare pack which I try to learn from, but when I refurbished the sticks on my Superbrain, I just put them back into the original positions, less the two which needed replacement on a capacity basis.
 

·
Registered
Joined
·
272 Posts
You may not like this comment, but you and I are beginning to think alike "like we are connected" (you've seen that Allstate commercial, haven't you? I don't have long hair, in fact I don't have any on top:)). Seriously that question on the best pairing considering the voltage taps fleetingly cross my mind a couple of days ago while reading about everyone's experiments with stick conditioning.

I think maybe the answer will be we don't know how Honda uses the voltage readings in their SW algorithms so we can't take advantage of the placement.
 

·
Registered
Joined
·
6,840 Posts
I'm not sure. I paired the best and better with the weakest and weak. This strategy assumes the average performance will be better, the voltage higher (or just better). On the other hand, it's sort of an attempt to trick the BCM: whether it's fooled, I don't know, plus, it may be better not to try to trick, it might be better to make sure the BCM sees the true condition of each cell as best as possible and that it manages IMA based on that... Basically, if a pair looks better than it is under some metrics, then the bad stick might get pushed more than it should be...

[edit] Stumbled across this thread looking for stuff. Thought I'd mention that I've experimented a bit with pairs. My last pack I paired weakest with strongest, so that all the pairs came out about the same. What this does is 'drive' the weaker stick harder, and the stronger stick lighter. On one reading this should be good - because the weaker stick had been getting under-charged in the car, based on a voltage cutoff indicating 'full' in the car (the weaker stick will hit the high voltage cutoff before it's actually charged). So if you pair it with a stronger one, the weak stick voltage will go higher than normal, charging more, the strong stick lower, and the total voltage where it should be. Yet I'm not sure what being fully charged actually means when it comes to old and weak sticks, now; could be that - and my hunch is that - higher-than-normal-voltage charging in a weak stick is actually overcharging...

And that's my other 'reading': strong-weak pairs drive the weak stick into the grave. I had great and predictable results from my last pack at first, with strong-weak pairs, but over about a half year of hard use, performance is weakening. I'm concerned that the weak sticks in the pairs have been deteriorating rapidly. Could be due to the strong-weak pairing, could be due to just older, weaker sticks being driven hard. I've been pretty brutal on this pack, much more so than I ever was with earlier ones (the original and a semi-reconditioned one)...

In a nut shell, creating strong-weak pairs hides the weak stick and the pack will perform better, no doubt. But that performance MAY be shorter-lived. If you create most-similar-pairs, then the pack will perform as good as the weakest pair can stand. The BCM will manage the pack, manage performance, so that the weakest pair never falls outside the car's performance thresholds (such as low and high voltage cutoffs)... I'm really not sure what I'd do, what I will do, next time. It's a shame to stunt the pack performance to the performance of the weakest sticks... To some degree it seems unnecessary...
 

·
Registered
Joined
·
20 Posts
I'm thinking about getting a PowerLab 8. Is it hard to setup? That is, does anyone know what the settings are for the powerlab to charge?




I have to say that the PowerLab 8 is great, though expensive.

I recommend at least 3 cycles on each stick. This will be about 9 hours per stick.

The reality is you'll do some additional testing afterwards. You can easily spend 2 weeks or more on this project.
 

·
Premium Member
Joined
·
2,890 Posts
I'm thinking about getting a PowerLab 8. Is it hard to setup? That is, does anyone know what the settings are for the powerlab to charge?
Out of the Box it has a "default" NiMH option that you can select ... would just have to set a few things like number of cells ... if that is enough for you fine , your done ... off you go test your batteries.

Basic setup is easy ... just keep in mind the PL8 is one of those devices with lots of options and controls ... you don't have to use them all ... but they are there if you want them.

Just be aware that the PL8 does let you take more control of the charge and discharge process ... you don't have to use defaults ... you can alter the algorithms used ... if you want to do that for testing different things or slightly different ways , the PL8 lets you do a fair bit of customizing... you don't have to ... you can just use the defaults ... but it does have more options.

All that having been said ... that ability to tweak and customize things ... although it adds some complexity , is one of the selling features of the PL8 ... if all you want is super easy and such , I'd say to consider other products first ... why spend money for features you won't use ... decide what it is you actually want to do ... look at the available options ... and pick the product that meets your usage needs , for a reasonable amount of money... for me that was the PL8 ... and I've been very happy with it ... because I liked and wanted the features it offered , I wanted the additional tweaking complexity of the controls , etc ... that was me ... use the tool that best meets your own needs.
 

·
Super Moderator
Joined
·
8,209 Posts
Ian, does the PL8 allow one to set up slope detection shutoff on NiMH discharge? Couldn't find that addressed in manual.

Also, the manual seems to imply that the PL8 measures internal resistance only for Lithium based packs, and not NiMH. Is this true?
 

·
Registered
Joined
·
20 Posts
Thanks for the info. I'm interested in learning about this. I have some RC planes that I can use this charger on the Lipo batteries as well as for my Civic and Insight.

I also been looking at grid charging and was wondering if a charger with this much versatility could be used. That is, with voltage multipliers. ie 6 x 30v gives 180v. Again just doing some thinking.

Thanks again


Out of the Box it has a "default" NiMH option that you can select ... would just have to set a few things like number of cells ... if that is enough for you fine , your done ... off you go test your batteries.

Basic setup is easy ... just keep in mind the PL8 is one of those devices with lots of options and controls ... you don't have to use them all ... but they are there if you want them.

Just be aware that the PL8 does let you take more control of the charge and discharge process ... you don't have to use defaults ... you can alter the algorithms used ... if you want to do that for testing different things or slightly different ways , the PL8 lets you do a fair bit of customizing... you don't have to ... you can just use the defaults ... but it does have more options.

All that having been said ... that ability to tweak and customize things ... although it adds some complexity , is one of the selling features of the PL8 ... if all you want is super easy and such , I'd say to consider other products first ... why spend money for features you won't use ... decide what it is you actually want to do ... look at the available options ... and pick the product that meets your usage needs , for a reasonable amount of money... for me that was the PL8 ... and I've been very happy with it ... because I liked and wanted the features it offered , I wanted the additional tweaking complexity of the controls , etc ... that was me ... use the tool that best meets your own needs.
 

·
Premium Member
Joined
·
2,890 Posts
Ian, does the PL8 allow one to set up slope detection shutoff on NiMH discharge? Couldn't find that addressed in manual.

Also, the manual seems to imply that the PL8 measures internal resistance only for Lithium based packs, and not NiMH. Is this true?
The PL8 default NiMH has a negative or fall back slope detection as one of the shutoff conditions ... but the details about that slope detection and shut off like other aspects of the PL8 are adjustable if the user wishes to manually see the difference between various adjustments to it... so yes it has it ... and yes it lets you alter it in many ways if you like... how many mV of dV ... how long to wait to test for it ... a scale modifier like 1/2 of setting for __ and then full setting after ___ .... etc.

It is also not the only end of charge detection method available for NiMH ... you can set the charge to stop based on other conditions / events as well... Ah capacity , time, etc...

It does not have at this time an OEM , + Slope rise dV shut off ... so as the OEM software is now ... you can't program it to terminate charging when the + rising slope is at some given angle or something like that alone ... You'd have to have it terminate based on something else ... at least for now ... they change and add more to the software based on how popular a customer options is ... right now people aren't asking for that particular ( + Up slope angle) , feature so it is not likely to get added any time soon , by the OEM.

- - - - -

As for Internal Resistance.... there are two methods available with a PL8:
  • The built in Resistance test.
    • This is actually a impedance test for the AC impedance of the cell being tested ... calibrated for Lithium batteries to give a fair estimate of what that cells DC resistance would be ... Sense the PL8 controls the current flow to and from the battery with PWM , the PL8 can see the rise and fall of terminal voltage at the speed of the PWM , frequency ... the issue is that this AC dV with dI , is that the dV for impedance is not only caused by resistance ... the Inductance of the wires and such can also give influence ... because of this the calibrated algorithm built in is designed to use this function for only some battery chemistry types that the OEM thinks it has been tweaked enough to be reasonably accurate.
    • This is available automatically any time one of the OEM approved chemistry types and testing methods is indicated... and it will display an estimate in Ohms ... real time adjusting at regular intervals over the entire battery test... when the type of test ... either battery type or or testing conditions don't fit the OEM approved list ... it is not shown.
    • The PL8 doesn't do chemical tests to confirm what kind of battery you have it connected to ... it tries to use built in algorithms to look for things like voltage characteristics and such ... but you can always lie to it ... things like 3 NiMH cells in series has a nominal Voltage of 3.6V same as a Li-Iion single cell ... if the PL8 program detects the NiMH lie , it will give and error ... if it doesn't it gives you the results assuming it was a Li-Ion cell ... sense the program is design to compensate the impedance to resistance based partially on battery , keep in mind it's estimates of other chemistry types may be less accurate... be VERY careful about lying about the battery chemistry.
  • If you connect the PL8 to a PC for data logging ... you can see the change in DC Voltage when a Change in current happens ... either from charging or discharging ... you can use this to estimate the ohms via ... V=IR ... it won't give you a regularly updating ohms number over the entire test ... just the start and stop points ... with the start point after a reasonable rest for the terminal voltage to level off , being the more accurate of the two ... but you can do this option for any chemistry , or any type of test.
 
1 - 20 of 409 Posts
Top