Analyzing, refurbishing/repairing a dead hybrid battery pack (2000 Honda Insight)

[eq1]

Originally Posted by IamIan

....Just keep in mind that 1 minute delay is not needed... just the time it takes the discharge current to stabilize to a given value... usually a second or so...

I was using the 'stabilize' strategy at first, but it wasn't very consistent, or easy or clear to decide when the current had stabilized... So I switched to 1 minute. It may matter whether you're using a high or low discharge rate. At first I was using 10 amps and sometimes it took a while for the level to stabilize, and/or the readout jumped around a lot nearing the top end... It wasn't that bad at only 3.5 amps...

[IamIan]

Originally Posted by eq1

So, using my example, if it were 8000 mV, 7500 mV, and 3500 mA, the result would be N mOhms? The result is 0.143, the same as before...

8,000 mV = 8V
7,500 mV = 7.5V
3,500 mA = 3.5A
change in Voltage = 0.5V @3.5A
V=IR
0.5/3.5= 0.143 Ohms

not mOhms ... the reason is that the metric prefix is just a scale it is not the unit itself ... milli is a prefix to Volt ... 1 V / 1000 = 1mV ... the division by 1000 is not the voltage term it is a multiplier.

In the original form:
V=IR
To keep the same ratio of V=IR but in milli scale ... you multiple both sides by 1000.
0.5V * 1000 = 500mV
3.5A * 1000 = 3,500mA
In the form of the V=IR equation ratio
(0.5V * 1000) = (3.5A * __ Ohms * 1000)
When you divide both sides by 3,500 mA you take away the 3.5 and the 1,000 leaving only ___ Ohms ... it is no longer mOhms.

Same effect happens any time you do algebra when one side is the product of two terms multiplied... weather the equation is F=MA or V=IR or E=MC^2 ... etc.

To avoid confusion some people always convert to base units before using ratio equations ... which ever method is easiest for you.

Originally Posted by eq1

Do you know the pair-ordering of the subpack pairs as monitored by the 10 voltage taps?

Not off the top of my head ... but it is easy enough to determine ... use a multi-meter to check for continuity between the connector that plugs into the BCM and the junction board on the side that makes contact with the stick... or if the sticks are still in place ... check for continuity from the bolt to the plug pins.

( added: As I recall there is a resistor inline, so it won't be a dead short )

Originally Posted by eq1

I was using the 'stabilize' strategy at first, but it wasn't very consistent, or easy or clear to decide when the current had stabilized... So I switched to 1 minute. It may matter whether you're using a high or low discharge rate. At first I was using 10 amps and sometimes it took a while for the level to stabilize, and/or the readout jumped around a lot nearing the top end... It wasn't that bad at only 3.5 amps...

If the battery discharging device you have does not stabilize current flow quickly your alternative method is reasonable ... just keep in mind that consistency becomes even more important ... your 1 minute results are a bit further from true ohms and are more on the side of relative ohms ... meaning don't expect your tests to give the same Ohm numbers someone else testing the same battery with a different method would get ... your method can be fine as long as the numbers you get for stick A are compared to stick B using the same method... because you aren't really looking for true Ohms ... you are using the tests to compare it to other sticks ... it is a relative comparison type of testing ... as such relative ohms and not true ohms is fine.

[eq1]

Thanks a lot for your help. I decided not to pay too much attention to my internal resistance calculations, as I don't really know what a significant value is. For example, an average of 0.14 ohms for all sticks, with 2 sticks at 0.17 and 1 at 0.15; most at 0.14, a handful at 0.13. I don't know how large a difference matters, plus, I wasn't going to change any sticks anyway... I did use them as one indicator of stick health, along with other things, like calculations for self-discharge, and mostly graphs of discharge/charge capacity over the 5 cycles...

On the pairings, Ron at HBR responded to my question about that over at Insight Central. It seemed pretty clear he was saying 1 & 2 form a pair, 3 & 4, and so on... I went with that and paired up the weakest stick with the strongest, the second weakest with the second strongest, and so on... Most of the sticks were pretty similar, though. There was really only one pair with pretty much the weakest sticks, that I didn't want together. It seemed logical to me - that if the BCM monitors pairs of sticks, you should try to make each pair the strongest possible. But who knows...

I got the pack back together and in the car. Drove about 200 miles today. Seemed flawless to me. I pushed assist and it never wavered. The lowest the SoC dash gauge fell was 16 or 17 bars. It started at 19, took some regen and popped up to 20. I'd push assist on a hill, or a freeway on-ramp, and it'd drop to 19. If starting at 19 and doing the same it might drop to 18. With multiple, half-gauge assists on the freeway, on inclines, it dropped to 17 or so. I could hold half-gauge assists for quite a long time - as long as I needed to - up freeway inclines... One thing that seemed odd, or novel, was, on some of these longer assists, the SoC gauge might drop to 17 or 16, but it would pop back up to 18 or 19 - without having observed the charge gauge charging at all in the intervening time. It was as though, under heavy load, the voltage would drop a bit and the SoC gauge would reflect that. And then once the load was removed (i.e. no more assist needed), the voltage would pop back up, and so would the SoC gauge - just like the voltage readout does when you're cycling batteries - it drops under load but then pops back up...

I did some full assists and had no problems, no big drops in the gauge, no drop outs in assist power, just a good strong pull...

So, hopefully this will last for some time. I'll try to post some of the data later, maybe the excel spreadsheet I worked with, etc.

BTW, if anyone decides to do this, here's one tip: There's 6 colored wires going from the BCM to the circuit board side of the battery pack. They run along the corner of the pack. Two are red, one long, one short. These snap into channels molded into the orange plate on that side. You need to pull them out when you take the plate off - BUT, after the week or 2 of cycling, you're likely to forget that they snap back into the orange panel and connect to 2 of the 20 PTC strip screws. Ask me how I know. They're the only 2 connections that could easily be forgotten, or rather, that one can easily mistake as being connected to something left in the car, like the DC-DC converter side of things... I got the pack all back together and in the car, only to discover that these two RED wires dangling there should have been connected before I put the circuit board back on. Lifting the battery pack into the car once is bad enough; having to take it back out and put it back in again is a real bummer (not to mention having to remove the circuit board again)...

All in all, though, there's not much you're likely to miss, just taking everything apart, not labeling or taking pictures, and just putting it all back together as seems logical as you move along...

Final note: my pack has 168,000 miles on it. I'm totally impressed with what I've seemingly achieved with this cycling process... I started the pack having charged it up as full as it would go with the car charging system. The average initial discharge was only about 1500mAh. Taking self-discharge into account (i.e. only able to do one stick at a time, so by the time you get to the 20th it's sat for awhile), as well as the average amount I've been able to discharge (~85%), I estimated that the sticks were charged to about 2000mAh when they all came out of the car. My 4th cycle discharge averaged 4607mAh (the 5th was a little different procedures so not directly comparable). And my final cycling-charge averaged 6773 mAh (I did do a final final top-up of every stick in a short interval, a balancing charge so to speak, as I don't have a grid charger to do the whole pack at once). But anyway, if I could pull the average 85% out of these sticks, I'd be pulling about 5750mAh: compare that to the 1500 I stared with. I've probably at least doubled the capacity, maybe closer to tripled, while I've probably gotten rid of some ... aspects that limited performance (I don't know, voltage depression, nickel dendrites, whatever. Maybe just the increased capacity alone is enough)... Supposedly my pack was balanced to begin with, just weak, as the P1449-78 trouble code plus subcode is the 'lowest on the totem poll', to use Ron at HBR's description. He said the other codes would be reported first if they were a problem, and since they weren't they're not... But he also said that he's had to replace most of the sticks on packs with 1449-78 code. So, I guess only time will tell if my pack's revitalized performance will actually last...

[MetroMPG]

eq1 - congrats on your success!

Originally Posted by eq1

One thing that seemed odd, or novel, was, on some of these longer assists, the SoC gauge might drop to 17 or 16, but it would pop back up to 18 or 19 - without having observed the charge gauge charging at all in the intervening time.

If I'm not mistaken, you're just describing "background charging": The car will top up the pack at a relatively low amperage while cruising, without lighting any bars on the charging side of the display.

If you're sensitive enough to your normal "fully charged" (or "pack disconnected via the big-a$$ OFF switch") lean burn window, you'll notice it's harder to enter or hold lean burn near your usual load limit when background charging is happening. Once the pack is back up to ~19/20 bars and background charging ceases, lean burn is easier to achieve/hold.

Final note: my pack has 168,000 miles on it. I'm totally impressed with what I've seemingly achieved with this cycling process... I've probably at least doubled the capacity, maybe closer to tripled

That's great!

So, I guess only time will tell if my pack's revitalized performance will actually last...

I hope you'll keep us posted. I haven't driven my car for several weeks - since I reported my results earlier in this thread. (I realize my infrequent car usage is not ideal for battery health.)

Now I'm waiting on a few parts before I install my grid charger/balancer.

[IamIan]

Originally Posted by eq1

Thanks a lot for your help. I decided not to pay too much attention to my internal resistance calculations, as I don't really know what a significant value is.

It is more important how close they are to each other than it is what the absolute value is.

I think you did the right thing ... using it as one of several battery health indicators to decide what sticks you wanted to pair up together... the only other thing would be to do the same kind of thing to decide which to replace ... if you reach the point of wanting to replace some.

Originally Posted by eq1

It seemed logical to me - that if the BCM monitors pairs of sticks, you should try to make each pair the strongest possible.

I agree.

Originally Posted by eq1

One thing that seemed odd, or novel, was, on some of these longer assists, the SoC gauge might drop to 17 or 16, but it would pop back up to 18 or 19 - without having observed the charge gauge charging at all in the intervening time.

I suspect either back ground charging ... I've seen as little as 100mA actually going into the IMA battery ... but in the low amp range the OEM dash display won't show anything at all.

The other possibility is a positive recall ... the rare but does exist event where the car corrects SoC up.

Originally Posted by eq1

So, hopefully this will last for some time.

Originally Posted by eq1

So, I guess only time will tell if my pack's revitalized performance will actually last...

Keep in mind it is treating the symptoms not the issues that caused them. without modifications to the OEM system ... it is only a matter of time before the core issues recreate the symptoms you just treated... of course you can just do it again.

[ogregev]

Originally Posted by jamesqf

More convenient still, I've found, is to get a couple of ring terminals with diameter large enough to accept the bolts that normally connect the sticks in the pack. Undo all the bolts (thus eliminating the risk of current from all the batteries), and use a couple of the metal stand-off pieces (the thick metal "washers") that are in the terminals that are outside connections to screw the ring terminals to each stick to charge it. A cordless drill with a 10 mm socket is a great help here.

One potentially life-threatening error in the above. Leaving the plates in place can keep the circuit live without the bolts. The bolts snug the plate to the stick, but it there may be enough contact to form a circuit without the bolts.

Be careful.

[ogregev]

Originally Posted by IamIan

Option #2 > Junk Yard Batteries collected and already tested for you:
Sense that time years ago ... Ron at Hybrid-Battery-Repair.com took this concept and turned it into an entire business ... You can pay him for individual subpacks he has already tested and weeded out of the questionable ones if you like ... How many of what tested quality he has fluctuates regularly.

Option #3 > Replace the Subpack with New Alternative:
Due to the issues of timely acquiring reasonably priced junk yard candidates and costs of testing and sorting the used batteries Ron then sourced his 'BetterBattery" as a new battery alternative ... as a drop in replacement ... It is not the same battery as used in the OEM ... but Ron has tested it , and he at least claims it is just as good or better than the OEM.

Ian,

There are so few sticks available anymore that HB (me) no longer sells them. They're all needed for repair jobs.

Also, BetterBattery sticks have different charge/discharge profiles than the standard sticks so they cannot be mixed in the same pack with standard sticks - it's all or nothing.

Ron (Hybrid-Battery)

[ogregev]

Originally Posted by IamIan

I suspect the Super Brain also uses some of these other methods as fail safes ... for instance it has an optional temperature sensor could potentially be used for the dT ( change in Temperature ) , and the dT/dt ( Change in Temperature per unit time ) , and the TCO ( Peak Temperature Cut Off ) ... I suspect it also has a maximum peak voltage based on selected chemistry and selected number of cells ...

The SuperBrain 989 does have a temperature probe, but you're testing six cells and you're likely to have only a single cell overheating, so it's kind of useless.

I contacted the company and they couldn't make a six-sensor adapter for any reasonable money.

[eq1]

Originally Posted by MetroMPG

If I'm not mistaken, you're just describing "background charging": The car will top up the pack at a relatively low amperage while cruising, without lighting any bars on the charging side of the display.
If you're sensitive enough to your normal "fully charged" ... lean burn window, you'll notice it's harder to enter or hold...when background charging is happening...

Hmm, that thought crossed my mind as well. Now that I think about it, there was one instance when I saw some charging at non-normal regen times, sort of around this time when I was seeing the gauge dip to 16 or 17 and then pop back up...

[IamIan]

Originally Posted by ogregev

The SuperBrain 989 does have a temperature probe, but you're testing six cells and you're likely to have only a single cell overheating, so it's kind of useless.

I contacted the company and they couldn't make a six-sensor adapter for any reasonable money.

The temperature sensor is not the only fail safe detection method.

If one really wants to ... despite what the OEM says ... the temperature sensor they use is a relatively simple device ... it's resistance changes with temperature... so weather they make one for a 6 cell test or not ... an add on circuit could be made to do the job ... sensors for each of the 6 cells ... those sensors are feed into a comparator circuit , if any one of them or all of them collectively are above some set point the resistance the SuperBrain sees is set to high ... so instead of the SuperBrain doing it , an external circuit does it.

I know the temptation people have to shorten the time it takes ... and I know it eats a lot of time ... But personally ... I see more benefits to the slower approach ... even if it does eat a lot more time ... but that's just my personal preference.

Originally Posted by ogregev

There are so few sticks available anymore that HB (me) no longer sells them. They're all needed for repair jobs.

Also, BetterBattery sticks have different charge/discharge profiles than the standard sticks so they cannot be mixed in the same pack with standard sticks - it's all or nothing.

Ron (Hybrid-Battery)

You probably know the current availability better than me... I know it is an option ... but as you say availability can be an issue.

If you don't sell them anymore than that is no longer an option for people... oh well.

You know the specs of the BetterBattery ... so if it is an all or nothing thing ... than that's what it is.