2003 Civic Hybrid: refurbishing/cycling the hybrid battery pack
 

---------------------------

Admin note: for background/more information about "reviving/refurbishing" Honda hybrid battery packs, see also: http://ecomodder.com/forum/showthrea...ems-13610.html

---------------------------
A word of caution:
Working on a battery pack from a hybrid car can be extremely dangerous if you do not know what you are doing. They produce very high voltage and current and can easily kill you if not handled properly. Do not work on a hybrid battery unless you are very well versed in electricity and how to work around it safely.

Background:
I purchased a 2003 Honda Civic Hybrid in fall of 2009. It has a manual transmission and had 68,000 miles when I bought it. I am the second owner. The first sign of battery trouble came about a year later with 85,000 miles when the car set IMA trouble code P1449, which can indicate battery deterioration or battery overheating. I cleared the code and it stayed off for several months then eventually came back. It gradually began to recur more and more frequently and the useable capacity of the battery pack declined. Eventually the pack could provide almost no assist before the P1449 code was set. To keep the IMA light from coming on so I could pass inspection I wired a switch in parallel with the clutch safety switch, which disables electric assist and regenerative braking, and added a relay that would bypass the switch when I stepped on the brake so I would still have regenerative braking. This kept the battery topped off and essentially masked the problem so the IMA light would not set. This also left the task of moving the car up to the 1.3 liter, 85 horsepower engine, which is manageable but not much fun to drive. I made the decision to disassemble the battery pack and cycle the 20 subpacks with a specialized battery charger/analyzer. The unit of choice for this task is the MRC Superbrain 989, which is designed primarily with radio controlled vehicle batteries in mind.

I was about to tear down and start refurbishing my battery pack this weekend. I unpacked the Super Brain I just bought, plugged it in and found that the power supply is dead out of the box. Not exactly confidence inspiring to say the least. Can't reach MRC or the place I bought it from until Monday. So I lost an entire weekend plus of time working on the pack.

Not necessarily?

Doesn't the power supply just convert 110v AC to 12v DC?

If so, you could power the super brain from a spare battery supported by a 12v charger.

It's actually 15v, but it can run off a 12v battery. However, the only battery charger I have is a trickle charger, so obviously it won't support the amperage needed to cycle the subpacks.

Jumper cables to the car? :)

No sense in burning all that fuel unnecessarily, I'll just wait until I can get a working power supply. I just can't believe they would have such poor quality control to ship a defective unit.

Update on my pack cycling progress. I got my Superbrain issues squared away, the vendor I purchased mine from was very helpful and promptly sent out a new power supply. I got the battery out of the car and disassembled on Friday night and started cycling subpacks Saturday morning. So far I've run six subpacks and the seventh is in process right now. The subpacks have returned capacities ranging from 6,252-6,806mah. A few questions for anyone who has any insight (no pun intended...ok maybe somewhat intended LOL)

A: Some of the subpacks had excellent capacity after just one cycling. One was 6,747mah, another 6,806mah. Is it still worth running those subpacks through some additional cycles? Since my weakest one so far was 6,252mah, and that after four cyclings, would any additional capacity I can coax out of the stronger subpacks be essentially useless?

B. Has anyone else found that the Superbrain is not capable of sustaining the 10 amp discharge rate recommended for cycling? On mine I can start a discharge cycle at 10 amps but the internal temperature regulating software seems to reduce it to about 5.3 amps after a few minutes. The internal fan on the Superbrain works and I have a fan blowing on the unit as well.

Yup. The pack capacity is limited by the weakest subpack.

I didn't see this, but I had a second cooling fan duct taped right on the vent area of the unit helping to draw air through it to aid the internal fan, not just blowing at it.

Another update: I'm at the halfway point, and I think I may have found a big part of my pack's problem. Subpack J (I used letters instead of numbers) returned only about 800 mah on its initial discharge, while most subpacks were around 3,000 mah, and one as high as 4,000 mah. Obviously with over 3,000 mah of capacity separating the strongest and weakest subpacks there was very little useful capacity. That weak one showed a capacity of only 4,349 mah on it's first charge, but subsequent cyclings brought that to 5,955 mah, 6,001 mah, and 6,069 mah. It's still the weakest subpack so far but well within established tolerances for a useable subpack. I'm giving it one more cycle to see if I can pull some more out of it. So far the results I've seen make me very optimistic about how the pack will perform once I'm finished.

Formula - I'm going to split your posts into a new thread with a link back to the original, generic one. I think it's worth separating your process/results as a distinct project that future Civic hybrid owners can look at.

I think you're right! That seems like a dud.

Your earlier subpack reports had me scratching my chin since they seemed like pretty good numbers.

Thanks Metro, I will also flesh out the original post with more detail and background.

Good idea - thanks. Keep us posted!

Another update:

Getting near the end, the 17th subpack is on the charger right now. The results have been very good so far, the lowest capacity returned after cycling was 6069 MaH, 93% of it's rated capacity. Two subpacks were very weak, with initial discharge capacities of 870 and 593. Both took only about 4300 MaH on their first charge but repeated cycling brought both of them over 6000. Those two subpacks would seem to be the cause of the pack's problems. I'm really excited to see how the pack will perform now.

A few questions that occurred to me during the process:

What are people's opinions on whether the position of a subpack in the pack will affect it's performance? Someone suggested this in another thread, I didn't think it would make much difference but I did notice that the two weakest subpacks are located dead center in the pack, which might subject them to more heat. Of the other three subpacks located in the center of the battery, two had good initial discharge and ultimate capacity, one was fair. Does anyone have any data from cycling another pack that might back this theory up?

Second question: Once I'm done with this process, some subpacks will have been sitting idle for almost two weeks, I was thinking it would be a good idea to top off each subpack before reassembly. Using the Superbrain to do this, would it make more sense to use the charge function or the trickle charge function? Any recommended settings?

And one more thing I was curious about: Has anyone noticed that sometimes individual cells in a subpack sometimes make noises during the charging cycle? A few times I have noticed cells making a sort of popping/hissing noise during charging, usually towards the end of the charge cycle. I didn't notice any cells getting exceptionally hot, I'm just curious if this is normal behavior.

Glad to hear you're nearly done!

My picture doesn't match yours, as far as location vs. capacity:



http://ecomodder.com/forum/attachmen...1&d=1326828189

In my pack there appears to be a trend from best to worst discharge capacity based on row (on average, best at the top, worst at the bottom). But I wouldn't conclude this is a design fault without confirming with data from other peoples' packs.

from Post #91 in http://ecomodder.com/forum/showthrea...tml#post285012

In the grand scheme of things, I don't think it matters. I'd do it quickly, otherwise you'll be perpetually "topping off" your sticks before reinstalling them. (They self-discharge fairly quickly.)

I didn't hear any of mine doing that! (Not that they didn't - I had multiple fans running all the time.) My gut tells me to be wary of those subpacks. If you don't get any replies from the experts here, I'd definitely search or post that question on insightcentral.

Once the pack is back in the car, heed Mr Small's advice: put the pack on a regular grid-charge trickle regimen. And periodically exercise the pack vigorously (full assist to "empty", full regen to "full") to help keep the subpacks from getting lazy & weak.

Looking forward to the next update.

I would talk to Eli over at insight central. He is selling completely new battery packs under the name Bumblebee batteries I think. He requires a core, so you could probably trade him your bad stick for a good one for a reasonable price. For what it is worth, I used two 90V 300mA LED power sources as an overnight charger to recover battery performance. It has worked quite well, although my battery was not nearly as deteriorated as yours.

What color are your sticks? Is this a 2nd, 3rd, 4th gen HCH? What city and state are? I have 2 used HCH sticks that have been cycled on a super brain 989 Il be willing to deal if needed. These are the paired sticks from a 2nd gen HCH.

Yup, thats the bases. Since the voltage taps are per pair match the best stick with the worse stick and so on, otherwise if one "tap" (TAPS OUT) :eek: under aload it will trigger a negative recal.

Others use a 350ma charge to balance the cells, however it seems cycleing helps to bring back those that are low/damaged.

No more updates?

I am guessing the 17th pack was a real doozy? :eek:

I had a guy tell me that there is a plug for the "TAPs' and it wouldnt be too difficult to rig up a plug to back feed it through a charger to avoid removing the cells from the pack.

Ok, here's an update: I put the pack back in the car. Zero change. Provided barely any assist at all, went to forced charging, set P1449 within a few miles. Thought twice about pushing the car off a bridge at that point. Obviously there are bigger problems than reduced capacity. I have zero ambition to screw with this thing again anytime soon, if I take it back out to work on it some more it won't be 'til the Spring, if at all.

Bummer.

However, the "forced charging" part is normal after doing that procedure on the pack: The computer's memory of the state of charge (SOC) no longer agrees with the actual state of charge. So when it detects the "extra" juice on board, it needs to recalibrate the pack (by forced charging to try to "top it up" and reset then the battery SOC gauge).

I guess there's something else going on with yours, but that part at least was "normal".

I should clarify that part. It did the same thing it always does if the 12V battery has been disconnected. After starting the car it showed 4 bars of charging for about 2-3 minutes, after which the charge/assist gauge went back to the center. I did this with the car sitting in the driveway. Once I started driving the car it provided weak assist at first, then stopped providing assist and went into forced charging.

I had the good fortune to get my pack cycled three times a week after my IMA first came on. Thats full topping and discharging. I believe balancing is the key issue. My pack would lose voltage then gain voltage before finally slowly dropping as the pack discharged. I believe the unbalanced state of my pack was the main cause. Over time it deteriorated to greatly reduced capacity, 1/3 useable capacity from full before regen, longer very inefficient force regens, and reduced assist. Balancing the pack just from charging all the cells and discharging to zero without replacing sticks restored its health.

My discharge rate looked like a normal gradual hill. It's been great for 4 months now without a grid charge balancing. Driving from the first grid charge i had a ton of capacity. Once that wore off, my normal capacity improved a lot.

Now i can tell that my battery is running at reduced capacity from a good battery. However, It is behaving itself well. I'm using a good range of my useable capacity, and my regen efficiency is good. (my car never does the force regen where it deletes all your bars and regens to full.)

Long story short, you want the minimum difference of capacity from in sticks from highest to lowest. Also grid charging the whole pack regularly will negate force regens, improving mpg. Since you seem to be a knowledgeable diy guy, i'd suggest building a grid charger for less than $100 in parts.

Ok, just skimming this thread. If your just cycling to improve each stick, and then putting them all back in the pack, it will still be unballanced.

Secondly, if it's a 2003, don't you still have a warranty on the battery? Don't know about civics, but 1st gen Insight packs are warranteed 10 years, 1xx,000 something or other miles.

Balance is not only the capacity, but also the IR of each stick, it is the IR which indicates the balance of all sticks not the capacity.:turtle:

Update:

I took the pack out of the car again for more testing. This time I did a self discharge test of all the subpacks, charging them with the superbrain, recording their capacity, letting them sit for a week and discharging them and recording mAH returned. I was expecting to find a few subpacks with very high self discharge rates but this didn't prove to be the case. The capacity returned on discharge for each subpack after a one week rest was between 77%-84%, not an exceptionally wide spread to my eyes. Before I put the pack back in the car I charged each subpack again just to be certain they were all full and drove the car. It set P1449 before I even made it out of second gear.

So now that I've established that all subpacks have good capacity (>6,000mAH) and similar rates of self discharge, the only other potential issue I'm aware of would be that some subpacks bench test well but cannot handle the high discharge rates they are subjected to in the car. I read on another forum about another testing procedure which involves using a high current battery load tester to test the subpacks under loads similar to actual driving, and monitoring the voltage of each cell in the subpack to see which ones drop out. Is anyone here familiar with this test, or has anyone done this and had success? Here is the load tester that was used for this test:

500 Amp Carbon Pile Load Tester

Bump

Anyone?

P1449 has several subcodes you need to find out which it is.

You need to get the equipment detailed in that thread on IC and test your sticks.

You could have a few dud sticks (if you are lucky) or you might find that the IR of all your sticks has increased and they can simply not cope with high IMA currents anymore, in which case all would need replacing. Nothing you can do about high IR.

If you confirm high IR on all sticks then you could consider my IMAC&C device for the HCH1 which has a battery protect mode stopping or reducing the ima requests to levels the battery can cope with. It's well documented on IC.

The best option would be to fix the pack of course.