[*sean*]

To answer your third question, it's just the 3 orange wires to the 3 leads on the 3-phase rectifier in any order, then two leads from the two on the rectifier to the input on the meanwell, and then connect 12V output from your car to the output leads of the meanwell. There's a little adjustable screw on the back of the meanwell to adjust output voltage.

This is the rectifier I bought: https://www.amazon.com/Baomain-Heatsink-Shape-Bridge-Rectifier/dp/B01JKRIPUK/

It has a heat sink built into it and stays cooler than the meanwell itself.

There is also a tiny parasitic drain (100mA IIRC?) from the meanwell when the car is off. To stop this some people use a relay; I'm using a manual kill switch. If it's just temporary you should be fine if you drive the car at least once every few days.

Thanks a ton for this info, mpg numbers dude! It will make things a lot easier.

 

[mpg_numbers_guy]

Deleted by poster.

 

[*sean*]

Charging question

My recently acquired CVT hasn't been driven in two years. I likely have a misbehaving cell in a stick. After multiple grid charges, and a selective discharge of the sticks holding the "bad" cell down to 7 volts (using @eq1 's method through the BCM connector; this is about 7/12 volt per cell - further than I wanted to go) I have one cell that's still about a volt less than the rest after a day or two. The IMA not happy, giving me a code about a detiorating stick.

I am wondering if it just needs some healthy doses of current.

Before I tear apart the pack...

I'm pretty sure I can fool my spare BCM into thinking everything is hunky-dory, and get assist and regen working. If I monitor the voltage of that stick pair and try not to over-discharge it - maybe even try to detect cell reversal - does exercising the pack with healthy doses of current sound like a good idea?

(I would then follow it with several pack discharge and charge cycles to condition the pack).

 

[*sean*]

Charging question

My recently acquired CVT hasn't been driven in two years. I likely have a misbehaving cell in a stick. After multiple grid charges, and a selective discharge of the sticks holding the "bad" cell down to 7 volts (using @eq1 's method through the BCM connector; this is about 7/12 volt per cell - further than I wanted to go) I have one cell that's still about a volt less than the rest after a day or two. The IMA not happy, giving me a code about a detiorating stick.

I am wondering if it just needs some healthy doses of current.

Before I tear apart the pack...

I'm pretty sure I can fool my spare BCM into thinking everything is hunky-dory, and get assist and regen working. If I monitor the voltage of that stick pair and try not to over-discharge it - maybe even try to detect cell reversal - does exercising the pack with healthy doses of current sound like a good idea?

(I would then follow it with several pack discharge and charge cycles to condition the pack).

Decided this is a bad idea, after reading battery charging posts where high current discharge into reversed cells led to worse, not better performance.

 

[eq1]

Decided this is a bad idea, after reading battery charging posts where high current discharge into reversed cells led to worse, not better performance.

What are you trying to do, exactly? If you have the pack apart, and access to cells, such that you can actually tell that you do indeed have one "bad" cell, why are you discharging through the voltage taps, or why did you? I'm trying to read between the lines here, it sounds like you have access to the sticks and cells, but maybe you don't...

If you do, you should just short the bad cell with a single resistor, at most about a 47 ohm one, let it go for as long as you can stand to wait, but at least long enough so that voltage doesn't rebound above 0.9V within a 24 hour period after you remove the short. The shorted cell should be below 0.1V after taking a while to get there... You might have to remove the short, see that it rebounds fast and/or high, and then put it back on again... Someone with chemistry knowledge explained before at IC that once voltage gets below 0.25V it can restore nickel back to the plate. Having worked with completely self-discharged cells, it looks like something palliative does indeed happen at very low voltages, as further discharge of already-self-discharged cells results in better performance, such as less voltage sag under discharge load. etc etc...

Oh, and you definitely shouldn't be subjecting a whole stick with a bad cell to high discharge current, reversing a cell at high current.

 

[*sean*]

What are you trying to do, exactly? If you have the pack apart, and access to cells, such that you can actually tell that you do indeed have one "bad" cell, why are you discharging through the voltage taps, or why did you? I'm trying to read between the lines here, it sounds like you have access to the sticks and cells, but maybe you don't...

If you do, you should just short the bad cell with a single resistor, at most about a 47 ohm one, let it go for as long as you can stand to wait, but at least long enough so that voltage doesn't rebound above 0.9V within a 24 hour period after you remove the short. The shorted cell should be below 0.1V after taking a while to get there... You might have to remove the short, see that it rebounds fast and/or high, and then put it back on again... Someone with chemistry knowledge explained before at IC that once voltage gets below 0.25V it can restore nickel back to the plate. Having worked with completely self-discharged cells, it looks like something palliative does indeed happen at very low voltages, as further discharge of already-self-discharged cells results in better performance, such as less voltage sag under discharge load. etc etc...

Oh, and you definitely shouldn't be subjecting a whole stick with a bad cell to high discharge current, reversing a cell at high current.

I am trying to recover a pack that is setting the IMA light and throwing an IMA pack imbalance code and causing the IMA system to be disabled. The pack is still in the car.

I have one stick pair that is consistently 1.25 +/- 0.5V below the others, while charging the pack from ~157V - 173V.

At some point I will take the pack out of the car, and find and replace the stick with the bad cell, then attempt to recover the cell. (Perhaps I should experiment on some NiMH AA cells I have around the house to refine a technique!)

I would prefer, however, to first take two bad Honda packs, and build one good pack, to have ready to go when I pull this one.

 

[eq1]

^ OK, gotcha... Truth is, if that cell in the bad stick can be fixed it should be with just the tap discharge and following charge. How long/deep did you go with the tap discharge, and what did you do after?

 

[Teletech]

Hey there, new to the forum and hybrid-novice. I do have a background in tinkering with cars and electronics though.
Bought a gen-1 with IMA issues that is running sucessfully in bypass mode (thanks Honda), my first car made in the OBDII era. I've ordered a scanner but for the moment it's stone knives and bearskins, or rather variacs, voltmeters, and a paperclip. Pulled the battery and grid-charged it to a very mild 165V, added taps so I can do more in future, and set it back in the car. I got assist for most of a 10 mile drive, but no charge until I'd driven it for a bit. Once the battery got below about 75% full the system threw a code and then no assist was forthcoming. Blink codes for IMA 78 (battery deterioration) and 46 HV bypass contactor. I'll grid charge to a higher voltage and check the strings, then let it sit and re-check, but not sure what to do about the HV contactor. Also an ABS code (blink code 61) and engine (blink 69), I didn't find breakouts for the blink codes apart from the IMA. So, help with decoding the ABS or engine codes would be welcome and of course suggestions about the HV contactor.

 

[*sean*]

Welcome! First OBD2 car? Awesome! With a car repair and electronics background you will have lots of fun.

I will leave discussion of the IMA to others, though personally I am learning about mine, and am taking a conservative approach. I would prefer to recover sticks at the pack (and maybe cell) level and replace bad sticks than buy a new pack (which would cost multiple times what I paid for my cars).

Curious to learn what pre-OBD2 cars you have and (optionally) what side of the country you are on since folks tend to meet in person from time to time.

 

[Teletech]

I'm working on getting a good grid charge so I can hunt for bad sticks but it looks like a thing that will take a week or so of charge and discharge here and there before I really know the extent of what's bad. I'm planning on just doing the gross stuff first, grid charge, deep discharge, grid charge, then let it sit and check banks for self-discharge and replace those. I'm hoping to avoid characterizing individual sticks since that would take many hours of time over many days/weeks and I'm too deep into too many projects just now. If I can get a pack that works well enough to not throw codes more than monthly I'll be satisfied for now.
I've got about a dozen other cars so a list might be long, four Saabs from '78 to '91 and a few Datsuns (210x2 and 1200) are likely the ones more interesting to this crowd. The Datsun 1200 was touted as the most fuel-efficent car sold in the USA in 1972 at 28.5MPG, it weighed 1608 lbs and the engine produced 67HP, rendering it pretty similar to the Insight in several respects. In fact, I can't help but notice the engine bay seems like it would accommodate an Insight powerplant.

I'm more often in the Monterey-bay area of California. Though I have a couple other diverse locations in CA and AZ, hence the Insight. The Saabs tend to top out just under 30MPG and while the 900 turbo is happy to cruise at any speed I'm willing to ask of it they get pretty loud by 70MPH. Datsun 210s do better on fuel consumption (the 5-speed has often gotten me 39MPG on road trips) they are even louder and slower with 70MPH being about as fast as the suspension is really suited to and loud at anything over about 60.

 

[Mark W]

The P0171 code eventually returned, as did the loss of assist, but with no IMA codes.

So later I reset the code and loosened the gas cap, and no P0171 code ("fuel metering too lean") but of course I get P1456 from having the fuel cap off. However, I'm calling this inconclusive.

A friend suggested that fixing the leak in the flexible joint between the two cats may have changed the amount of exhaust flowing across the sensor, and oxygen coming in, etc etc, and that may be affecting readings, and I should force the ECU to relearn the 02 sensor values.

After more searching, I learned more about how to read the short and long term fuel trim info. This might crack the puzzle.

This offered a list of things to examine for a P0171 issue, but doesn't talk about how to use ST/LT trim to identify which might be the culprit:


This article provides reasoning for resetting the ECU:


This video talks about how to read ST & LT fuel trim, what they mean and the kind of problem they may point to, with a great demonstration:

This is an even more in-depth discussion, giving even more info on fuel trim, also with a vacuum leak example, and how you can use short and long term trim to figure it out:

part 2 talks about a weak fuel pump, and dirty MAF sensor:

so now I have some ideas on where to look for causes of a lean condition:

• vacuum leaks: large LT trim near idle
• weak fuel pump and dirty MAF sensor: large LT trim at higher RPM

s

Sean,

I have a 2005 CVT with 140K miles on it. Last July I replaced the IMA battery with a greentec battery after it had sat for roughly two years. Initially, I was getting 50 mpg but it was back firing (forget the code). It eventually dropped to 33mpg and the IMA assist only worked the first mile or so. I changed the plugs, oil, transmission fluid, fuel filter and coil packs. I also ran a can of Sea Foam and several cans of Heet through it.

I was a little frustrated; so I took it to the dealership. They cleared a P0171 code (lean condition) and suggested they could not do anything with the IMA battery without replacing it with a new Honda battery. I told them that in my opinion there was nothing wrong with the battery. After some give and take, they suggested the first step was a 250 dollar resetting of the valves. I drove her the hundred miles home and searched the forum for a similar case. Your post above was the closest one to my case. Long story short, any suggestions on next steps?
Thanks,
Mark

 

[*sean*]

I have a 2005 CVT with 140K miles on it.

I've only been doing this for a year on my MT but I've noticed or read that parts like the EGR valve and the LAF sensor can contribute pretty significantly to poor mileage - and the life of both of these is limited. You could have a lot of other things going on. This can include vacuum leaks, grounding problems, bad sensors, deeper engine problems... that should be checked out and eliminated as causes - a real diagnosis. If you are paying a mechanic to sort them out, you should ask the forum to find one who is near you who knows Insights (@KLR3CYL on the west coast).

And you can do the valves in an hour, including learning how to do it. Adjusting the valves may make a BIG difference. This added 5 MPG and smoothed out the idle on mine. The cost is that of a feeler gauge and 10 mm wrenches. You may find they are way out of whack. I found that on both my CVT and MT.

I ended up replacing (fixing, actually) the EGR on mine, after building a tester to confirm it was bad. After testing a few other valves @jime and I concluded that these don't last very long. There are a lot of posts about EGR valves as a result, and a lot of cheap EGR valves sold on Ebay and Amazon. Some of these don't work. I know. I think @Natalya can point you in the right direction. It might not do anything, but at least it will eliminate a variable or a failure that will eventually happen.

I also replaced the LAF sensor with a Denso, which is the OEM version for cars starting in late 2001. These you pretty much can't test (well, I just thought of a way, but it would take a LOT of work to build the tester) and if you have another problem, like a leaky head gasket or a ring problem, contamination may trash your new one. I spent a lot of time eliminating those as possible problems. I did eventually replace mine and the mileage increased quite a bit.

I got my CVT not long ago and I'm still learning to drive it. I can squeak out 60+ MPG if I baby the heck out of it but 43-49 MPG seems to be where it wants to live. I have not touched the EGR or LAF on these but I think the LAF sensor may be new. I need to run tests to see if my other sensors are performing correctly. I am also not running the super high mileage RE92s tires either.

Also, I'm still looking into this, but consider running 87 gas from a Top Tier station (google it) and maybe Techron (black bottle). If you are buying the cheapest gas you can find, well, it's a variable that you can control. My borescope and piston-top-cleaning experiments suggested that the best high-additive high-octane gas might not be the best for the car. But this is probably a minor factor at best.

 

[10-A-SEE ME]

One of my CVT's has an issue where the dash light gear indicator
is not in synch with the shifter. Example " Indicator says car is in
Neutral" when the shifter is actually in "Drive" or indicator says
Reverse when it is in Neutral. Can this be adjusted somehow to
your knowledge ? thanks

 

[Rainsux]

I think this switch is under the console next to the shifter.

Sam

 

[jime]

Hmmmm. I don't think so Sam. The Electrical Troubleshooting Manual has a picture of the "Transmission Range Switch" which shows it on the top of the transmission. Picture attached.

Can it be adjusted? I think you will have to tinker a bit to see. I don't have a CVT here - sold it.

 

[Willie Williford]

Cable adjustment?

 

[jime]

Yeah, seems like a good bet.

 

[10-A-SEE ME]

OK will check cable

 

[HLK]

Hi guys,

I took my insight to get a battery replaced (local battery replacement place), took them a while but today they put it in and gave me a call saying "nothing happened".

He then asked if there was a reset procedure to do after dropping in the battery, me being less knowledgeable than they were, said sorry I don't know what to do, I've only turned my IMA on and off via the switch.

I don't have the car with me yet, I have to pick it up later tonight, but can someone run me through what i could do to try getting the IMA working? I did find a post saying the IMA and check engine light could be reset by pulling a fuse.

thanks!

 

[Willie Williford]

Pull fuse number 16 under the bonnet for app. 15 sec. re-insert the fuse. Start the car and let it idle till the SOC is full.