Honda Insight Forum banner

21 - 34 of 34 Posts

·
Registered
Joined
·
6,506 Posts
^There's not much I could do to test - besides trying it out, which I definitely would be curious to do. I'll PM you my address, you PM where to send the shipping payment. Sound good?
 

·
Registered
Joined
·
6,506 Posts
I finally got Jc's BCM the other day, for testing. I'll be trying to find time to install it in my car and see what happens... I did open it up, didn't see any obvious problems - was mainly looking for those scorched resistors that happen sometimes with HV shorts I think it is...
 

·
Registered
Joined
·
6,506 Posts
Installed JC's BCM (A03). Drove about 30 miles, 1 hour, didn't see any abnormal behavior - some things are slightly different than the two BCMs I've tried, but there was no 'drain' and background charging kicked-in like it always does... I.e. there was nothing to suggest that this BCM would drain the pack, fail to charge the pack, over some modestly short interval.

I started off by letting the BCM charge the pack to what it considered full (from a level that was already almost full). Nothing too out of the ordinary with that - it did start by charging at idle, at about 6A, and after charging about 10% it stopped, which stopping I haven't seen before. So I rev-charged it until pos recal, which was at about 33% nominal and what looked like pretty typical voltages, around 172V at about 7A, 84F degrees...

It pos recal-ed to 81%, bled charge off (i.e. regen disabled) until about 75.6%, which is normal. Using assist I took it down a bit and when it hit 72% nominal it jumped down to 65% nominal, also 'normal' for certain model year BCMs. My BCMs pos recal to 75% and allow another 6% charge, and they don't have that jump from 72% to 65%.

When it hit 65% it immediately began to background charge. I had AC ON. I took it up to 70.2% and background charge stopped - normal. I took it down again and it started background charging again around 65%. Later, after key-OFF and back ON, with AC and CC off, I had nominal state of charge down around 55%, no background charge. I turned AC ON and lights ON and it started background charging - normal...

So, all-in-all, I don't see anything wrong with this BCM. I'll leave it in the car for some more drives and keep an eye on it.
 

·
Registered
Joined
·
125 Posts
Discussion Starter #25
Have you attempted to recreate the fail? That is driving on flat ground at constant highway speed for some distance without IMA assist. As stated earlier, the battery indication level would fall until I think one bar. I saw this on two consecutive trips. Where I live in TN, roads are level for only very short distances. On I-26 heading East from Columbia, the roads are almost perfectly flat until Charleston, about 110 miles. After changing the MCM and BCM, I took another trip and did not experience the discharge. Maybe I should have sent the MCM as well. After replacing the just the BCM, the battery was definable being managed differently. I changed the MCM before the trip as a cautionary measure.
 

·
Registered
Joined
·
6,506 Posts
Have you attempted to recreate the fail? That is driving on flat ground at constant highway speed for some distance without IMA assist...
Not exactly, but I don't see how that would change anything I've seen thus far. I've seen no charge drain on the OBDIIC&C, for instance, no 'negative current' in the Bam parameter... How many miles do you think I have to drive until I definitely should have seen what you think you saw?
 

·
Registered
Joined
·
6,506 Posts
OK. I took it on a highway drive, about 50 miles, and very soon on the drive I observed something that suggests a possible answer to your problem. It likely has to start with a somewhat dysfunctional/errant pack, but what this BCM does could definitely lead to a gradually discharged pack... There's some weird stuff going on in terms of the 'top-end algorithm' - what determines when the pack is 'full'. I don't think it's a bona fide error, but rather, it's probably some wonky, ill conceived programming additions made by Honda for later model years...

In a nut shell, this mostly hinges on two things: 1) a low pos recal threshold, and 2) pos recal-ing directly to 81% with an automatic 5% drain (regen disabled, 12V load powered by DC-DC and IMA pack).

Shortly after I got on the highway, starting the trip at a nominal 54% SoC, background charge kicked-in, which is par for the course. But then, the pos recal happened quite early, much earlier than it should have, and much earlier than it would have happened were I using my other BCMs. I didn't catch the exact voltage and nominal charge state, but it was around only 60%; voltage was probably around 168V at maybe 10 amps... My pack is in good form, I checked tap voltages before the trip - all even (spread of only 0.02V), etc. - so I know with reasonable certainty where the pack's at and what it should be doing.

So, nominal state of charge jumps to 81% and then the car automatically disables regen and runs the 12V load off the IMA pack, until a nominal 76% charge state is reached, when regen is re-enabled. Thus, 5% of the pack's usable charge gets drained.

IF you have a somewhat dysfunctional pack - such as imbalanced, higher than normal IR, etc. - your tap voltages are going to be high during charge and/or at least one tap voltage will be higher than the others. So, presuming that this algorithm relies on tap voltages, you're going to see pos recals to 81% and automatic drains of 5% more often than you should, so much so that eventually it's possible the pack gets completely drained... Some of the details here start to get a little tricky, but this is the general 'sweep' of the problem.

I saw a few other weird things that might factor-in to this. For example, after a pos recal and some drain, down to about 76.8%, I stopped at a store. When I came back out, nominal SoC dropped to 70% on key-ON. Thus, the 'top-end determination algorithm' was already back into play. In general, this BCM has a way more active top-end algorithm or what have you - plus a low pos recal threshold. Put the two together and your pack's state of charge is going to be pegged to around 164V (more like 167V though; 164V is on the OBDIIC&C, which is about 3V lower than actual). The problem with this is that ~167V - 16.7V at taps - I don't think that's high enough to distinguish a near full pack, especially one that's degraded. On a good pack you can see 16.7V at low actual charge states... I mean, looking at my spreadsheet, my pack should be around 50% SoC; the A03 BCM had the nominal at 75%.

This BCM will end up cycling your pack within a very narrow range, never getting very high, and I think the normal day-to-day degradation that happens in these cells will gradually if not sooner pull the true charge state downward - you'll end up using a lower and lower true charge state, a smaller and smaller capacity window, until eventually a single automatic 5% drain is enough to exhaust all the usable capacity...

I'm not sure how we get directly from a steady-state highway cruise, with no use of assist or regen, to an empty or 1 bar BAT gauge. Things have to be put into play somehow. In general, I'm thinking the background charge has to be going - so at some point charge state has to be around 65% nominal, background charge kicks-in, you get a pos recal to 81% and then a 5% drain. Then, either that amount of drain is enough to empty the pack so you get a neg recal and the BAT gauge drops to 1 bar, or your BCM and MCM are already in some low capacity mode, where the BAT gauge bars are scaled to a small capacity window - draining only 5% or so amounts to the entire capacity in terms of 'bars'... Maybe it has to be more than just this 5% - as you need a minimum of 10% to avoid a P1449-78. But, the main point is simply the overall downward pressure, the down-scaling, of usable capacity that will happen with this BCM, especially with a sketchy pack...

That's the idea at this point...
 

·
Registered
Joined
·
125 Posts
Discussion Starter #28 (Edited)
That’s quite a bit to digest. It sounds like the late model BCM doesn’t like packs that are unbalanced more than earlier BCM units.

It may be helpful if I give a little more history. The battery pack in question was purchased new from Bubble Bee I guess 3 years ago and installed in my 02my CVT Insight. I probably ran it in that car a year until I wrecked the car, scrapped it, but kept the pack. The CVT IMA system was working perfectly before the wreck. No damage was done to the rear of the vehicle and the car didn’t experience high G forces so I don’t think the battery was compromised. When I purchased the 06my MT, it was obvious the battery pack was being managed much differently than With the CVT. The MT used the IMA assist a lot more. I installed the CVT's pack in the MT using the 06my IMA ECUs. Over the next few months, the pack gradually deteriorated and was frequently re-calibrating. A few weeks ago I cycle charged/discharged the original 06my pack and reinstalled it using a 03my MCM and BCM pair I had salvaged from another vehicle. So far the system’s. performance is much improved and holding steady.

Still the car is not quite right. Is it possible an under-performing ICE could effect how the IMA manages the pack and could be why it wants to uses assist so much more? I have never driven another MT so I have no reference as to what is normal.


Sent from my iPhone using Tapatalk
 

·
Premium Member
Joined
·
10,349 Posts
IMA use is determined by the "Load" on the engine. The CVT mostly keeps the rpm of the engine at the min. load requirements. What rpm is your CVT mostly running at? Compare that to the rpm that a lot of the hypermilers are running that have the 5 speeds. I chose 2500 to 3200 rpm as my favorite range.
(MY opinion)
 

·
Registered
Joined
·
125 Posts
Discussion Starter #30
Willy, I can't recall what rpm the CVT ran at. I always left the transmission in normal mode. I can tell you my CVT performed better than the MT although I could never get it to idle properly. The MT idles so smooth you're not sure its even running.
 

·
Registered
Joined
·
6,506 Posts
That’s quite a bit to digest.
The most important things to focus on are a lower pos recal threshold and the automatic 5% discharge.

A lower pos recal threshold means there's much less certainty about just how high the pack is getting charged. It has to do with the way NiMH voltage works: voltages don't reveal higher charge states until they're pretty high, like probably over 17.2V at the tap level, definitely over 16.8V. In general the pack won't be getting charged as high. Plus there'd probably be a downward tendency, meaning every pos recal would tend to charge to a lower and lower absolute level -- the first time you might be at say 4,875 mAh out of a nominal 6500, the 7th, 8th, 9th, and so on times you might be down to 2275mAh, that is, originally "81%" meant 4875 mAh of charge, now it means only 2275 mAh...

And then, the automatic 5% drain kicks-in after every pos recal. That may not seem like a big deal, but if the BCM is pos recal-ing often - AND the charge level to which it's pos recal-ing is low, getting lower, and/or is uncertain - then 5% discharge every time can add up. 5% can become a huge absolute value: 5% means 5 X 65mAh for a normal, good pack, but 5% on a sub-par pack can represent something like 15% of your total usable capacity...

It sounds like the late model BCM doesn’t like packs that are unbalanced more than earlier BCM units.
From what I've seen thus far, I think that could be said. An unbalanced pack means some cells, at least one tap, is more charged than the others. The BCM will stop charging when the most-charged tap reaches the pos recal threshold (call it 16.7V here), leaving the other taps at a lower charge level. So, we start with a lowish, uncertain pos recal threshold, then we throw in imbalance that leaves most of the taps at an even lower level... I think this BCM would be more sensitive to imbalance and would probably have less ability to correct it.

On the flip side, though, there could be good things about this BCM's tactics. To me it looks like Honda tried to adjust the programming to deal with something that I've thought is a problem - too high, too frequent, high charge state usage. Charging lower is probably a good thing overall, it appears that this BCM charges to a lower absolute level and also concentrates usage at lower charge states.

The battery pack in question was purchased new from Bubble Bee I guess 3 years ago and installed in my 02my CVT Insight. I probably ran it in that car a year until I wrecked the car....
Aftermarket cells are a little different than OEM. I don't feel nearly as certain about their behavior as I do about the OEM cells...

Still the car is not quite right. Is it possible an under-performing ICE could effect how the IMA manages the pack and could be why it wants to uses assist so much more?
I've heard that before. It would definitely make sense. No experience with it though.
 

·
Registered
Joined
·
6,506 Posts
JC PM'd me about the A03 BCM, I'm responding here, with excerpts:

...I've read some of your posting concerning this A03 BCM. It seems you've come to the conclusion the NiMH cells originally installed in my 2006 are different than earlier year vehicle cells and the same as Civic. It never occurred to me cell chemistry changed from earlier vehicles and could be the root cause of the failure.
Sources say later Insights used Civic cells, and Civic cells are different from 'Insight cells'...

Currently I'm using the OEM 2006 battery pack with an earlier version of the BCM and all appears to be okay. I never ran the A03 with the 2006 battery pack. When I received the vehicle I immediately installed a Bumble Bee battery since it was less than a year old. Evidently it was a mistake not to change the BCM at the same time.
I know very little about Bumblebee packs or other aftermarket packs. In theory, since the A03 is meant for 2005/06 Insight sticks or Civic sticks, if the A03 didn't work right with your BB pack it's probably a decent guess that the BB pack is more like older Insight packs... Or maybe your BB pack had issues...

I may want to re-install the A03 into my vehicle to see how it works with the 2006 battery. I think when you posted that the A03 maintains the battery at a lower charge level, you're referring to the absolute level as provided by a scan tool versus the relative charge level provided from the vehicle's dash meter display.
Yes, absolute level, real charge state, or at minimum the charge state shown on an OBDIIC&C, not 'the bars'...

I think you determined the purpose of this lowered charge level improves efficiency and perhaps battery life as well. But only for the later battery....
Not really. I only used it with my 2002 Insight pack. I thought the way the A03 BCM charged that pack, with a low pos recal threshold and generally forcing usage at the bottom-end of the charge state range, might have been done on purpose. That usage pattern fit with my pet theories about low charge state usage, which I think can be good, so it made sense to me. BUT, it turned out that it was all about the different cells, older Insight vs. later. The A03 BCM will under-charge older Insight packs; for example, when I used it I never saw a charge state above about 50%...

1) Were you able to run the A03 using the latest G1 Insight battery and were there any issues?
2) Did you have issues similar to what I experienced when running the A03 with an earlier battery?
I never tried it with anything other than the 2002 Insight pack. I didn't experience any of the issues you had, probably because my pack was in good condition or because you were using an aftermarket pack. At some point people complained of weird charging issues/state of charge issues with aftermarket packs. I don't think there's ever been much resolution around that. In general it's common IC knowledge that 'the aftermarket packs' have slightly different voltage characteristics, so... I don't know, there's been changes with aftermarket packs, I think newer cells are different. I don't like to think about aftermarket packs because it's all shrouded in mystery...

Basically, the questions you're asking really point to some complicated stuff. Running your 2006 pack with an older BCM should be charging it to a higher level. Eli at Bumblebee would like that. He thinks packs fail because they're not charged high enough often enough. Me, on the other hand, I think packs fail because they're charged too high too often, or, the flip side, that they're never discharged low enough (I've been leaning toward the latter)... So, I don't know. I write like volumes per day about this stuff, trying to get to the bottom of it...
 

·
Registered
Joined
·
125 Posts
Discussion Starter #33
eq1
If I were to re-install the original BCM/MCM pair back into my 2006 with its original battery, is there anything you'd be interested in having me look at for the sake of a better understanding? I don't have one of Peter's OBDII C&C tools and can't justify the cost to acquire one so I am limited of what I can provide.
 

·
Registered
Joined
·
6,506 Posts
^ No, it's all pretty clear to me now. Let me know if you want me to send it back... If you don't care all that much, though, I'd prefer to hold on to it and would be willing to buy it off of you. The lower voltage full charge threshold ('pos recal treshold') would be good for a lithium build I might do at some point... Send me a PM...
 
21 - 34 of 34 Posts
Top