Honda Insight Forum banner

81 - 100 of 113 Posts

·
Registered
Joined
·
96 Posts
what is your objective? Is it just to be able to drive without the fear of being stranded? if so, I would just disconnect the IMA battery and drive around that way. you could do this or even less than this and leave the IMA in place but just bypass it: Hybrid Automotive - Insight Battery Bypass Board Installation
Thank you very much sir! This looks very do-able. I checked out a fix from tryingbe and really appreciate it but it required soldering on a little circuit board and looked daunting. I think I can do this and I have an extra battery that I can steal parts from to make the bypass. I will give you an update when I finish. Again, I really appreciate your offering.
 

·
Linsight Designer
Joined
·
1,899 Posts
Discussion Starter #83
Josh, mod works great. Yes, you can instead cut the WHT/GRN wire at the ECM (pin A9)... same result.
 

·
Registered
Joined
·
24 Posts
Adding second battery in IMA bay.

Mudder, Thank you for helping us out. In this thread, you mentioned adding a second battery in the IMA bay. To me with the problems I am having this seems to be my best solution. I have an oversized battery I wish to place in the IMA bay. I was going to remove my existing 12-volt battery and add this oversized battery and run 0 gauge wire to the bay to the cables in the front that the OEM battery connects to.
I ran a test with just the oversized battery hooked up to the OEM cables and it ran for a very long time, at idle, without getting charged with everything on.
In theory:
Can this battery be connected just to the DC-DC + and - from the DC-DC converter?
Could I keep the existing 12 volt upfront hooked up as well?
Will these different sized batteries charge without overcharging one of them?
 

·
Registered
Joined
·
824 Posts
Yesterday evening I picked up a free voltmeter from Harbor Freight and replaced the probes with a cigarette lighter plug, for giggles. Coincidentally my weak battery later would not start the car (ECUs not getting enough current), so after a charge of the 12V it sprung to life.

This morning I was surprised to see the voltage at 11-some volts instead of 13+. With the voltmeter mounted in my phone holder I saw that the battery was not being charged based on its actual voltage but on whether or not I was accelerating.

Having messed with RV battery setups the rule of thumb IIRC is that 12.8 V is charged, 12.4 about 50% capacity remaining, and best to not let it go below 12V etc. And car batteries have thin plates so they make poor RV batteries because the erode and break off. Stopping the discharge at a specific point before it gets too low increases the battery lifetime, based on charts I've seen on the net.

So for a good battery, this cycling does not seem to be an issue, because the recharge event may occur frequently enough. But people have talked here about going with smaller 12V batteries which makes this more of an issue, as would extended engine-off events not factored into Honda's design. And most of our batteries are not new anyhow.

I am wondering if a compromise between wire cut vs not cut is to monitor the battery voltage and to enable the DC-DC once it gets below a certain voltage (say, 12.4 volts?) and let it stay on until it reaches 12.8V (full charge). Obviously you can't tell if it is at 12.8 resting volts while charging so you would need to monitor it with an Arduino and periodically close the wire and observe the eventual discharge event to determine if the battery has been fully recharged. Loads on the system will impact this but that could be measured with a hall effect current clamp in a battery cable and factored in. Or in fact have we learned over the years [email protected] posted this, that in practice leaving the DC-DC converter on full time has negligible MPG impact?
 

·
Premium Member
Joined
·
3,415 Posts
Guidelines for resting voltages for 12V batteries apply after sitting 24 hours from last use.

12.6V is more typical for a fully charged wet cell auto battery. AGM tend to sit a little higher as do deep cycle batteries.

12.4V UNDER LOAD may be fully charged. Heck, 11.8V under a heavy load may be fully charged. Normal running load is probably around 20A, and that can pull a fully charged battery down a bit.

You're basically using a moving target with two variables to establish rules, and unless you install the proposed hall effect sensor to monitor current, it's pointless.

You also don't have a grasp of 12V charging. 12V batteries must be charged at substantially higher than 12.8V. In fact, most need to be pushed to 14.4-14.7V to fully charge them (I have some solar-type deep cycle batteries that require 15.0V). They must be held there for about 3-5 hours and then "floated' at about 13.2-13.5V to maintain full charge. If you cut the current and let the voltage drop, you are now losing capacity to self-discharge.

The bottom line is that even with the cut mod, you'll likely NEVER fully charge your battery. When you hit 14.4V at about 4-6A, the battery is about 80% charged. that last 20% happens over 3-5 hours of the "absorption" phase of 12V charging at 14.4V.

Your 12.4-12.8 scenario would guarantee that the battery is kept at an unhealthy low state of charge and accelerate failure.

Whatever solution you come up with, you need to replace your 12V battery ASAP.

Yes, it's generally agreed that leaving it on full time has negligible mpg impact. Note my 4-6A/14.4V statement... that's 72W of power - less than 0.1 hp. That's less than 1% of the power required to push the car down the road at 65mph.
 

·
Registered
Joined
·
824 Posts
I decided to do the wire cut on my MT without getting fancy and inventing a voltage monitor. I'm very happy with not having to worry about losing my battery to being repeatedly undercharged.

I picked up a CVT recently and got the service history which indicated that the 12V battery was replaced more frequently than one would have liked. I installed a brand new battery, and a quick check showed that it's also not getting charged when under assist, as expected. So, I'm about to do the mod to this car as well.

Question. Has anyone allowed the car to stay in autostop so long that the IMA battery drains completely with this mod in place? What happens? At what time does the DC-DC cut out, or does it run the IMA into the ground? Does the car attempt to start? Or?
 

·
Registered
Joined
·
6,843 Posts
Question. Has anyone allowed the car to stay in autostop so long that the IMA battery drains completely with this mod in place? What happens?
Not with this mod in place, but I don't think anything should be different, as it's the other DCDC wire that comes into play, the one that disables it entirely...

At what time does the DC-DC cut out, or does it run the IMA into the ground? Does the car attempt to start? Or?
I think the DCDC will be totally disabled once the IMA pack reaches 'empty'. Not positive which metric is in play - nominal state of charge, neg recal, pack voltage, tap voltage, amp-hour count, or ?. I've seen neg recals without the DCDC being disabled, so it's not the neg recal itself or the same criterion that causes a neg recal (which is kind of weird now that I think about it)... It won't go lower than 17.X percent nominal state of charge. Maybe it's total pack voltage...

The car does not attempt to start.
 

·
Linsight Designer
Joined
·
1,899 Posts
Discussion Starter #89 (Edited)
Lots of questions asked since I last read this thread. I'll start from the newest and work my way up.

With the WHT/GRN wire cut, if you leave the car in autostop long enough to fully empty the IMA battery, then the MCM will still correctly disable the DCDC (via the GRN/BLK wire). The WHT/GRN wire has nothing to do with the MCM being able to turn the DCDC off when things are too funky in the IMA bay.

...

In a car with a working IMA - whether or not WHT/GRN is cut - the DCDC cuts off as soon as SOC falls below 25%. If you watch the video I posted in #1 all the way through, I show exactly this condition.

...

What S Keith says about lead acid charging is 100% correct; thanks for saving me the time in writing up my own explanation. To reiterate his statements, lead acid battery voltage is not a good indicator of SoC unless the battery has sat unloaded for 24 hours. You can find similar SoC graphs for when constant loads are placed on the battery, but the key word here is 'constant,' and that certainly isn't the case in a vehicle setting.

...

In regards to "compromises" on whether or not to cut the WHT/GRN wire... I propose there's no reason to compromise, as there's nothing to "compromise" in the first place. The more I've measured the actual data with actual calibrated test instruments, the more convinced I've become that cutting the WHT/GRN wire is essentially a no-brainer. To be clear: I am 100% convinced that cutting the WHT/GRN wire has zero negative consequences. As such, I don't believe adding a "compromise" to re-enable the OEM behavior is useful in any way, unless you just like to fiddle with knobs. However, if you really want a knob to fiddle, just put an SPST switch inline with the cut WHT/GRN wire and presto, now you have a knob to fiddle with when you want your 12 volt battery to be severely undercharged.

Anyone coming to insightfest will get to see a pretty nifty technical demo I hacked together last night: It allows us to fiddle any knob we want on the OEM DCDC converter, while simultaneously measuring all results... it won't win any beauty awards, but it certainly has all the technical questions accounted for. We can even use this equipment in-situ, but for empirical data that is unnecessary. Still, if someone wants to put the equipment in their car, I'm totally cool with that.

I'm excited to present my findings on this subject at ICF2019... so that we can finally convince everyone that Honda 100% made an error-in-design by adding the WHT/GRN mode select to the DCDC. I'm all ears if someone can propose one valid reason for this besides "I think it hurts my mileage" (which isn't true, btw, but I'll explain why people THINK this is happening during the presentation).

Here's a couple quick pictures I captured showing my test configuration... what hacked proto isn't complete without an Arduino Uno kludged in somewhere:
82323


82324


FYI: I showed this demo to my wife this afternoon... and she was bored out of her mind.

...

In regards to installing another battery (presumably 12 volts), there's really no reason to do that unless your car sits for months on end without being driven... a better solution in that case would be to install a battery disconnect switch. If you've installed a massive audio system, then you're much better off adding a 1 Farad capacitor, which will take care of the high frequency current surges, while the DCDC handles the base load (including recharging the cap).

As long as you drive the car enough to have it realiably turn on when you put the key in, there's absolutely zero reason you need to install any additional battery capacity to the 12 volt system... in fact, I propose going the other way: I removed my OEM battery and replaced it with a 15 ounce (i.e. less than 1 pound) sealed lead acid battery (I believe it's 4 Ah). With the WHT/GRN wire cut - and a semi-healthy IMA battery - the DCDC is all you need*. I've had this configuration in one G1 for nearly as long as this thread has been around and it hasn't caused any problems; another G1 has an OEM battery with WHT/GRN cut. If you don't want to get stranded when the IMA dies (i.e. you need your backup starter), then get a lithium jump starter battery backup... our 1 liter engines pull well under 200 A when using the backup starter.

*The only legal caveat I'll add here is that the anti-lock braking system - when activated - pulls a TON of current, with very high frequency pulsing on the 12V system. Based on my testing, the 4 Ah battery + DCDC converter provides sufficient current for the ABS system to work properly. However, in an emergency braking event, it is possible that a severely-poorly-balanced IMA battery could cause the DCDC to brown out (or be turned off altogether by the MCM). If this condition were to occur with a smaller 12 volt battery installed, it is possible - albeit unlikely - that the ABS system could pull more current than the 12V bus could deliver, which could result in injury or death. To be clear, any and all statements I make on this website are without warranty and are to be used at your own risk.

^^^The above paragraph is certain to make someone on my legal team happy ;).

...

In regards to installing additional cabling from the IMA bay to the engine bay, there's really no reason to do that, either... the OEM WHT cable runs all the way up to the engine fuse box and can handle way more current than the entire car can consume (80 A fused), EXCEPT for the backup starter... if you're planning to move the OEM 12 volt battery into the IMA bay - and you want to retain backup starter capabilities - then again, I recommend just getting a lithium backup jump starter... you can even remove the OEM battery tray and cables if you want... and just connect the lithium starter directly to the starter if you ever need it. On my cars, I've left the tiny 12 volt battery in the engine bay, but you could just as easily move it into the IMA bay, too (without beefing up the OEM WHT wire).

...

Well, I think I got all the questions answered. If not, post again and I'll keep going.
 

·
Registered
Joined
·
24 Posts
Lots of questions asked since I last read this thread. I'll start from the newest and work my way up.

With the WHT/GRN wire cut, if you leave the car in autostop long enough to fully empty the IMA battery, then the MCM will still correctly disable the DCDC (via the GRN/BLK wire). The WHT/GRN wire has nothing to do with the MCM being able to turn the DCDC off when things are too funky in the IMA bay.

...

In a car with a working IMA - whether or not WHT/GRN is cut - the DCDC cuts off as soon as SOC falls below 25%. If you watch the video I posted in #1 all the way through, I show exactly this condition.

...

What S Keith says about lead acid charging is 100% correct; thanks for saving me the time in writing up my own explanation. To reiterate his statements, lead acid battery voltage is not a good indicator of SoC unless the battery has sat unloaded for 24 hours. You can find similar SoC graphs for when constant loads are placed on the battery, but the key word here is 'constant,' and that certainly isn't the case in a vehicle setting.

...

In regards to "compromises" on whether or not to cut the WHT/GRN wire... I propose there's no reason to compromise, as there's nothing to "compromise" in the first place. The more I've measured the actual data with actual calibrated test instruments, the more convinced I've become that cutting the WHT/GRN wire is essentially a no-brainer. To be clear: I am 100% convinced that cutting the WHT/GRN wire has zero negative consequences. As such, I don't believe adding a "compromise" to re-enable the OEM behavior is useful in any way, unless you just like to fiddle with knobs. However, if you really want a knob to fiddle, just put an SPST switch inline with the cut WHT/GRN wire and presto, now you have a knob to fiddle with when you want your 12 volt battery to be severely undercharged.

Anyone coming to insightfest will get to see a pretty nifty technical demo I hacked together last night: It allows us to fiddle any knob we want on the OEM DCDC converter, while simultaneously measuring all results... it won't win any beauty awards, but it certainly has all the technical questions accounted for. We can even use this equipment in-situ, but for empirical data that is unnecessary. Still, if someone wants to put the equipment in their car, I'm totally cool with that.

I'm excited to present my findings on this subject at ICF2019... so that we can finally convince everyone that Honda 100% made an error-in-design by adding the WHT/GRN mode select to the DCDC. I'm all ears if someone can propose one valid reason for this besides "I think it hurts my mileage" (which isn't true, btw, but I'll explain why people THINK this is happening during the presentation).

Here's a couple quick pictures I captured showing my test configuration... what hacked proto isn't complete without an Arduino Uno kludged in somewhere:
View attachment 82323

View attachment 82324

FYI: I showed this demo to my wife this afternoon... and she was bored out of her mind.

...

In regards to installing another battery (presumably 12 volts), there's really no reason to do that unless your car sits for months on end without being driven... a better solution in that case would be to install a battery disconnect switch. If you've installed a massive audio system, then you're much better off adding a 1 Farad capacitor, which will take care of the high frequency current surges, while the DCDC handles the base load (including recharging the cap).

As long as you drive the car enough to have it realiably turn on when you put the key in, there's absolutely zero reason you need to install any additional battery capacity to the 12 volt system... in fact, I propose going the other way: I removed my OEM battery and replaced it with a 15 ounce (i.e. less than 1 pound) sealed lead acid battery (I believe it's 4 Ah). With the WHT/GRN wire cut - and a semi-healthy IMA battery - the DCDC is all you need*. I've had this configuration in one G1 for nearly as long as this thread has been around and it hasn't caused any problems; another G1 has an OEM battery with WHT/GRN cut. If you don't want to get stranded when the IMA dies (i.e. you need your backup starter), then get a lithium jump starter battery backup... our 1 liter engines pull well under 200 A when using the backup starter.

*The only legal caveat I'll add here is that the anti-lock braking system - when activated - pulls a TON of current, with very high frequency pulsing on the 12V system. Based on my testing, the 4 Ah battery + DCDC converter provides sufficient current for the ABS system to work properly. However, in an emergency braking event, it is possible that a severely-poorly-balanced IMA battery could cause the DCDC to brown out (or be turned off altogether by the MCM). If this condition were to occur with a smaller 12 volt battery installed, it is possible - albeit unlikely - that the ABS system could pull more current than the 12V bus could deliver, which could result in injury or death. To be clear, any and all statements I make on this website are without warranty and are to be used at your own risk.

^^^The above paragraph is certain to make someone on my legal team happy ;).

...

In regards to installing additional cabling from the IMA bay to the engine bay, there's really no reason to do that, either... the OEM WHT cable runs all the way up to the engine fuse box and can handle way more current than the entire car can consume (80 A fused), EXCEPT for the backup starter... if you're planning to move the OEM 12 volt battery into the IMA bay - and you want to retain backup starter capabilities - then again, I recommend just getting a lithium backup jump starter... you can even remove the OEM battery tray and cables if you want... and just connect the lithium starter directly to the starter if you ever need it. On my cars, I've left the tiny 12 volt battery in the engine bay, but you could just as easily move it into the IMA bay, too (without beefing up the OEM WHT wire).

...

Well, I think I got all the questions answered. If not, post again and I'll keep going.
Many Thank You's Mudder. I will continue to look for what I am missing. Good to know that I will not need to run heavy cables and install a heavy 12-volt battery to keep a charge.
 

·
Linsight Designer
Joined
·
1,899 Posts
Discussion Starter #92
Nifty demo screen thingy! How did you do that? Software?
Yeah, the software I'm using is LabVIEW... it's a jumbo jet when all I needed was a car, but I'm super familiar with it, having worked at National Instruments (creator of LabVIEW). I spent about 5 hours total on the demo start to finish, so it's not a polished project, but hopefully the visuals are good enough to get the point across.
 

·
Registered
Joined
·
6,843 Posts
Mudder, would you please twiddle the "Engine Coolant Temp (PWM duty cycle)" slider and report the DCDC output voltage range? Or maybe you know that off the top of your head?

....Anyone coming to insightfest will get to see a pretty nifty technical demo I hacked together last night: It allows us to fiddle any knob we want on the OEM DCDC converter, while simultaneously measuring all results... Here's a couple quick pictures....
 

·
Registered
Joined
·
6,843 Posts
^fyi, I think I figured out the max output voltage: ~15.1V. This is the DCDC output I got when I shorted DVCT to ground. Based on what I think I'm understanding about the DVCT signal, it appears that it should have a lot more upside than downside, in terms of DCDC output potential. The lowest output voltage I've seen while measuring the PWM signal as a voltage, was 13.8V with engine coolant fully warmed up, and the voltage at DVCT was about 4.4V. I measured 90% duty cycle, but I don't know if my device is measuring correctly, as it seems to have values even when I'm not touching the probes to anything. But 0.9 X 5V =4.5V, so I think that might be in the ballpark. But also, what it seems to imply is that there's only about +0.6V of 'headroom' until the minimum output voltage is reached (i.e. at 5V max signal level, the DCDC output voltage will be the lowest it can go). Dropping the signal from about 4.4V to zero (when I shorted DVCT to GND) increased output voltage from about 13.8V to 15.1V, about a 3.4:1 ratio, signal to output. So raising the signal V from ~4.4V to 5V I'm guessing won't drop the output voltage much farther below 13.8V. If it's linear and this 3.4:1 ratio holds, then I'd be able to drop it by another 0.6V/3.4=0.2V or so...

It'd still welcome guidance - if anyone has some...
 

·
Registered
Joined
·
6,843 Posts
^ Overall I'm trying to figure out how to adjust DCDC output voltage to make it easier to use different types of 12V battery. Here, in this thread, I was asking if mudder could use his LabView program to tell me what the potential DCDC ouput voltage range is, more specifically using the 'DVCT' pin. In my post above, I think I've basically figured out the range - probably about 13.6V to almost certainly 15.1V - but it's based on only my half-assed understanding of how DVCT works... My plan is to basically install a voltage converter with an output to DVCT so that I can adjust the voltage at the pin between 0 and 5V, I think that's all I need to do to be able to set the DCDC output voltage to between 13.6V (or thereabouts) and 15.1V.
 

·
Linsight Designer
Joined
·
1,899 Posts
Discussion Starter #97
DVCT is a 10 Hz PWM signal... the DCDC expects values between 10% and 90% duty cycle. You are correct that the DCDC will output 15.x volts when DVCT faults outside that range. DVCT is only used by the DCDC when the ECM wants the DCDC to output 14.x volts; in 12.x mode the output voltage is fixed. So the valid ranges are: fixed 12.x value (WHT/GRN held low), then variable 13.x to 14.x (WHT/GRN held high), then fixed 15.x value during fault condition (DVCT out-of-range.

Note that the DCDC responds VERY SLOWLY to rising PWM duty changes, but responds very fast to falling PWM duty changes; this is a bug in TDK Lambda's firmware (they designed the DCDC).
 

·
Registered
Joined
·
6,843 Posts
DVCT is a 10 Hz PWM signal... the DCDC expects values between 10% and 90% duty cycle....
I was thinking that the PWM signal is equivalent to a voltage, in this case between 0 and 5 volts. But that doesn't seem to be true. Is this right?

If I can produce a 10Hz PWM at 10-90% duty and feed it to the DVCT pin (+ output to DVCT, - output to any ground, maybe the BLU/WHT SG10 GND?), will that allow me to adjust DCDC output voltage within whatever range it can do?*

How about with something like this: https://www.amazon.com/Generator-Adjustable-1Hz-150KHz-Envistia-Mall/dp/B07QNYMNGK/ref=sr_1_15?keywords=5V+PWM+signal+generator&qid=1570256510&sr=8-15

Note that the DCDC responds VERY SLOWLY to rising PWM duty changes, but responds very fast to falling PWM duty changes; this is a bug in TDK Lambda's firmware (they designed the DCDC).
I saw very slow changes when I shorted DVCT to ground - it jumped up to 15.1V, but then took at least a minute or so to drop back down to ~13.8V once the short was removed...
 

·
Registered
Joined
·
6,843 Posts
If DVCT at the DCDC side is already 5V when the line is cut (i.e. cut DVCT, measure voltage between DCDC end and ground), doesn't that mean the output voltage is already as low as it can go?
 

·
Registered
Joined
·
96 Posts
This is probably not the place to ask my question but I haven't been on the site since it has been changed and having trouble navigating it. I typed in BCM problems and wound up here.
I bypassed my IMA a year or so ago just using the end plate and the BCM. It keeps the 12v batt. charged but it is time to smog it and I have a check engine light on and the code says my BCM connection is bad. I unplugged all of the plugs and plugged em back in but that didn't fix it. I am thinking about buying one and just trying it but having trouble finding one. Anybody have ideas to help me?
 
81 - 100 of 113 Posts
Top