Understanding P1449

P1449 isn't exclusively about the PTC strips -- it can be either about the strips or about battery module degradation. You need to pull blink codes to figure out which one. Once you have -- I'm sure you're getting P1449 @ 78 -- Ctrl-F for the right version of that blink code after P1449 in http://gershon.ucoz.com/HONDA/HONDA_DTC.pdf.

To pull blink codes, short pin 9 of the OBD2 port to the ground bracket around the port with a paperclip, then keyON and watch the IMA light blink.

1500 watts is 30% of the stated capacity of the battery when it was new. I am sure it's pushing 40 to 50 percent of the capacity of a 22 year old battery.

Click to expand...

You're thinking 1500mAh, not 1500W, maybe. The battery can and does supply about 10kW at max assist (~80A @ 130V).

As for what's causing it -- when you test your pack to rebuild it, you discharge it and recharge it. That process brings cells in line and gets the pack happy for a while. After rebuilding my pack, I'm still getting the occasional IMA light after dumping the accelerator on the highway for >30s.

The "Anatomy of an IMA light" thread here is very helpful for describing what causes the IMA light -- long story short, the BCM detects a reversed cell and interprets that as the bottom of the pack; it then triggers a recal to find the top of the pack and then current-counts to figure out what the capacity of the pack is.

When I next pull the pack, I'm going to grid-charge and re-test the potentials I get: it will show me how the "top" of the battery capacity has changed since I cycled and reinstalled my pack, and expose which cells that have become weak.

What I honestly recommend for weeding out bad cells is to dump even more current out of the battery than you currently are. I worked with ~1900W; aim for >5kW if you can. Your goal is to recreate the scenario the BCM is watching for. That will let you weed out cells that will fall flat under high current as in the "Anatomy of an IMA light" thread. The Battery Module's ON/OFF switch is rated to break this kind of current, but definitely have the BCM and MCM disconnected when you do this.

I don't think power output has much to do with any of the trouble codes, or at least with any of the P1449s. There's a lot of leeway in that regard, throttling, compensation, whatever. The most typical P1449 is the 78, and that has to do with cell charge-imbalance - the pack can't be charged at least 10% of the nominal capacity (i.e 650mAh of 6500mAh) without hitting some kind of 'this is f'ed up' threshold. It has to be a pretty serious/bad threshold/metric because once the pack is found to be effectively empty and the IMA tries to charge, it relaxes its normal criteria and allows you to charge more than usual. So, if you can't charge a meager 10% under those relaxed conditions - things are really bad.

My guess is fast self discharge of a single cell is the most common cause. A single cell fairly quickly sinks to a low charge state relative to others, it's then driven low/used at bottom all the time - while other cells don't discharge low, the single cell develops a high voltage profile, much higher than other cells, while the other cells develop a lower one from not being discharged low, voltage imbalance sets in, charge balance gets worse, you're hitting empty ('neg recal') way prematurely and then the pack can't be re-charged much before probably a tap voltage level or slope that signifies full is detected, P1449-78 triggers.

If you want to figure it out, if indeed you're getting the P1449-78, then drive the car/use assist close to where the pack normally goes empty, stop in auto-stop, hop out and measure tap voltages, then try to measure them again just as the BAT gauge falls to 1 bar (i.e. when the pack is considered empty). The tap/stick pair with the offending cell will have a noticeably larger voltage drop than the others, like other taps will be around say 14.9V at time 1, and maybe 14.7V at time 2, while one will be at say 14.9V at time 1 and 14.3V at time 2. That's the tap/stick pair that likely has a fast self discharge cell. You can then pull those sticks and discharge each cell on the bench to confirm/find out my how much the charge imbalance from cell to cell differs. etc etc.

^ I strongly recommend you do the tap measurement thing I mentioned above before you pull the pack. It can really cut down on the labor, most likely telling you which stick pair needs to be pulled. That way you don't have to pull all the sticks and the job becomes... pretty easy.

edit: hmm, on the other hand, do you have an OBDIIC&C? A 'calpod' clutch switch? If you don't, I guess it can be kind of hard to know just when you're gonna hit neg recal, and you don't have any way to reset stuff if you do. Kind of hard to 'watch the BAT gauge' and know that, 'Yeah, my pack's gonna neg recal (go empty) soon'. I forget I have those things/use them in the process... You don't have to be exactly at empty when you take your tap measurements, but the pack needs to be pretty close. You're basically just trying to catch the stick pair with the cell that's truly going empty, catching the voltage drop of that cell in the time 1 and 2 measurements. Its voltage will drop noticeably even if you don't actually hit neg recal, too.

An alternative might be to get the pack pretty close to empty then stick a light bulb discharger on the pack, with the biggest lightbulb you have (as in wattage). Measure taps at the start of the discharge, then some time later, depends how close you are to empty. Close to empty will likely be at least one tap below ... maybe 14.5V. Kind of hard to say with varying pack conditions/balance etc... Neg recal for me in similar situations always happens when a tap drops to 14.3V at about a 1 amp load... Since the NiMH reaction takes place on a more or less flat plateau, most of your charged cells are hanging around 1.20V or more, or >14.4V for a stick-pair. Basically, this makes it pretty easy to catch that voltage drop, that happens when a cell goes truly empty.

Az-Insight said:

...Trying to decide if it's worth the bother to rebuild it, or just getting new replacement from Bumblebee?

Click to expand...

A couple thoughts: In general I'd say no, it's not worth it. I think if anyone were thinking about buying a replacement pack, I'd suggest they think about getting the mudder lithium thing. Stock management + aftermarket packs seems pretty hit or miss. I think some people have had quite a bit of success, but many others see the pack failing in a matter of years. Given what the IMA offers, which really isn't much, "a matter of years" doesn't seem worth a couple thousand dollars. Full IMA bypass, with pack removed, also seems like a competitive option. Personally, if I weren't able to make my pack work (well) seemingly indefinitely, which I've been able to do, I'd most likely be doing the full bypass.

Az-Insight said:

I'm going to thoroughly check the battery sticks again and then deep cycle and recharge the pack several more times and see if this helps.

Click to expand...

I don't know what your 'thorough check' of the battery sticks is, and I don't have the patience to inquire and review. I'll say this though: checking/testing 'sticks' is more a matter of checking/testing cells - it only takes a single wayward cell to ruin the whole pack. You can often/usually? spot sticks with a wayward cell, like if you cycle the whole stick on the bench, but not always, and not typically uneven self discharge at all. So, if you do nothing at cell-level the whole operation will be just a matter of luck (usually bad luck).

The easiest way to check/test initially is to discharge each stick to about 6.75V-6.90V at about 6 amps, then discharge each cell individually at say 1 amp. Personally, I have a hobby charger that can monitor cell pairs and also reduce current when voltage approaches a user set threshold. So, for instance, I'll discharge a stick at 6.5 amps, be able to see in realtime how the cell pairs are tracking during the discharge, current will reduce once voltage hits say 6.9V and discharge will stop at 3 amps. Then I'll attach the charger/discharger to each cell and discharge each to 1V, starting at 3 amps and tapering to 1 amp. This sort of operation will show you which cell has been charged the least, likely which one is worst, and generally how charge-balanced the cells are/have been able to achieve.

Etc etc. There's a lot more to it all though, it's a lot of work if you want to really be sure you're doing what you need to be doing to have a pack that works - not only works period but works well. Overall it takes very little to have a working, functional pack, so if the pack can't work in the car, like getting IMA lights/not able to charge enough to work, you have a really bad problem...

Other than this, I've found the most important thing to do is discharge every cell completely. This usually ends up 'resetting' them, balancing them as much as they can be. It's more like matching them rather than balancing. Over time and usage cells develop weird electro chemical behavior that full, low current discharges seem to fix. If you don't get them all discharged they'll continue to misbehave... I guess the operation I described above gets you close to this, you (one) might consider dropping current way lower though, discharge each cell not to 1V at 1 amp, but probably more like... 0.5V at 100mA or less, the lesser the better (I have a cell-level discharge rig with I think it has 33 ohm resistors, so at 1V that'd be 33mA discharge current, for instance).

@Az-Insight, got it, so you've already cycled the pack. Your original load testing was similar enough to mine current-wise, so I think the difference between us must be one of opinion RE: what qualifies a bad stick.

Here's what my sticks looked like in the last half of my rebuild. Note that I measured each stick (not just each tap) by probing the stick bolt contacts below the plastic guard on the right, and touching the left probe to the battery tap's crimps through the back of the BCM tap connector. I would switch stick-side, then tap side, as needed, and voice-record my readouts.

At 12A, any one stick that sees a sudden 1V drop ("For example ...") has a bad cell, at least while your pack is in the normal charge ranges (think >125V under 12A load). Pitch it. The problem quickly becomes one of getting replacement sticks; HybridRevolt sells "usable rejects" which

output at least 5.5 amp hour (5500mAh) under an 8 amp load and also measure at least 4 amp hour (4000mAh) under 80A

Click to expand...

As you can see from the spreadsheet, I replaced any sticks that looked bad in waves before retesting so I could test all cells the same way (ended up discarding ~25% of those I bought, e.g. my replacement for stick T on 10/30 still had a failing cell at a 130V pack). I now only have 3 original sticks.

I just picked up @Jedz's original Insight and spare HCH packs, and will be adding fresh test data from that rebuild so you have something to compare against.