Ok, on a *recent* attempt to fullfill my end of the warrenty contract my vehicle was tampered with at the dealership. When I recieved it back their were NO codes, NO check engine light, NO ima light and EVERYTHING seemed to be working like it usually. Regular assist, Regual charge, normal millage. It was claimed the tow truck driver disconnected the battery (yeah right) With 1500km left on the warrnety i realized I may have seriously been screwed once again. 3 days of driving and like i said above, everything was working. From all the posts a few months back about scan guage i had returned my regualr OBD2 scanner and picked one up and was using it as an info/code scanner. With PC hardware, to test for failure you do a stress test. Run prime 95, Memtest 86+ ETC so what would be a "Stress test" for the IMA system to show fault in a deteriorated pack? What I did was drove my insight under full accelleration SHIFTING early (almost to the point of stall) so that full assist would be achieved for the maximum amount of time, with the least amount of speed. I did this on a fairly large hill with an open straight away above. Now here is where it gets interesting. The insight accelerated like normal all the way through the gears, but right near the peak of the hill something strange happened. The scanguage show a drop from the steady 13.9V 12v system monitor to 12.1V, then, seconds later the Assist guage CUT in half, and then completly disappeared and went offline. The voltage then jumped back up to 13.9V and I experienced what seemed to be a "Normal" recall, but at the end of it the IMA light and check engine light came back on. This mixed in with the "Testimonials" of failures one the insight by people driving once a week (which i did to recieve the codes initially) brings me to the theorey that to show a failure your insight has to be put into a stress zone, Either Full battery discharge, or minimal battery charge. It would be interesting if the people who have high millage insights, who perseave battery problems to put their millage aside for one day and try and reproduce what I did and confirm my theory. If we can come up with a standard for how stress test your pack yourself, it may be extremly benificial to other owners

What you saw was that assist cut off because the 12v battery needed charging. I belive that you can't charge your 12V and have assist at the same time. At least thats what I see on my 03 civic.

Maybe the DC/DC converter need replacing. Ask them nicely to check it out. I believe there are test proceedures for it.

My hydroponics experiment
My carputer

You ever notice how hard it is to lip read cartoon characters?

No it couldnt be because For the 3 days leading up to this not once did such a thing happen where the voltaged dropped. This was only after an extended period of full assist (like 3-4 times the normal amount of time it takes to reach these speeds)

Also, If what you were saying was true, the voltage of the 12V system wouldnt Drop in one instance from 13.9v to 12.1v, but whould steadily drop from 13.9 to 13.8 to 13.7 etc etc. Also, the gain back up from 12.1v to 13.9 would also steadily increase and not do a sudden jump.

If the DC-DC Converter went down, it would be IMA Failure related. If the IMA malfunctions the MCM may take out the DC momentrarily

Edit:

What i am ultimatley getting at is there may be a great number of people with faulting packs, they just havent put their insight into that "Failure zone" to exploit such problems

So doing this procedure made the IMA light come on, eh?
Interesting stess test idea. But's hard to consider trying it since I've got a good tank going.

Eric Powers
www.EVPowers.com
EV Powers - A Green Vehicle workshop in Madison, WI
608-729-4082

I've driven my IMA down to one bar several times (long, steep hill at 65 mph). But no "Check IMA" light ever came on. But at least now I know what to look for. My HCH now has 96k miles and has had only one IMA-related failure.

I have some speculation why the IMA light comes on under those circumstances.

Some cells hold more charge than others, and when you run the pack down, you run those cells dry, and in fact are forcing current through them in a discharged state, which is known to cause damage to NiMH cells.

The computer detects this and disables the IMA system, which then disables the charge circuitry to the 12v battery, thus your precipitous voltage drop.
Once the recal gets going, the charging circuitry is re-enabled, even though the IMA is "offline" for the duration, hence the sudden voltage return.

The computer expects a pack with limited variation in cell charge capacity, which is why they are cap matched at the factory. If they age uniformly then the recal will work many times, but if the recal detects too much variation for it to handle, on go the IMA and check engine lights.

So it sounds like your pack has a weak cell or two, but they aren't entirely dead.
Therefore your test seems to be a legitimate method for provoking the problem.

And you get to drop the car off under warranty with the IMA light on, which is pretty much the only way to convince a dealer anything is wrong.

The HCH-II's recently got a software update to increase the reserve under zero SoC on the meter, and chances are that allows a recal to deal with a wider cell variation and nurse more life out of an aging pack. Too bad the Insights didn't get such an update, as it would be helpful here. You could possibly simulate it by avoiding draining the pack too much, not that hard in an MT if you are willing to downshift a bit, but since it is under warranty, better to just get a new pack.

Once the IMA light goes on however, I thought that 12v charging was disabled, so you might want to keep an eye on that voltage meter. Maybe there is a state where the pack isn't "dead" enough for the computer to disable charging, but it will still enable the IMA light.

Hey Hondasucks,

Why don't you use spell check? Your message becomes completely unbelievable when there are so many misspellings.

That was uncalled for.

Stop now. Both of you.

My hydroponics experiment
My carputer

You ever notice how hard it is to lip read cartoon characters?

kmh3,

My 2000 5-speed Insight got a new hybrid battery pack (and MCM and BCM units). This is my plan to give it a long life:

• No long assists
• Keep the 12-volt battery so the hybrid battery pack does not go dead trying to save the 12-volt battery

In other words, avoid geting the charge low enough to do a recal. If you have other ideas, I'm listening.