Honda Insight Forum banner

1 - 20 of 150 Posts

·
Registered
Joined
·
6,747 Posts
Discussion Starter #1 (Edited)
I've been meaning to try this for a long time but never got around to it, until recently...

OVERVIEW
Rainsux was asking about an aftermarket ~$100 lithium 12V battery a few weeks ago, which got me thinking about it again. http://www.insightcentral.net/forums/modifications-technical-issues/84969-lithium-jump-starters.html

That battery was only 6 amp-hours (compared to something like 30Ah for the typical lead-acid that goes into our Insights). I wasn't sure how that size would fare in our cars, regardless of battery chemistry. So I figured an easy way to test was to forget the expensive lithium, try a 2-stick-based 12V battery (actually, 2 sticks but only 10 of the cells), and start observing/logging the demands and stuff...

In a nut shell, if you're interested in shedding some unnecessary weight - about 22.5 lbs - it definitely works. In some sense I think it's actually a much more appropriately-sized 12V solution for the hybrid Insight -- which uses the IMA pack and motor for starting under normal conditions and a DC-DC converter running off the IMA motor and pack for 12V accessory loads most of the time when the car's running...

On the other hand, I think it will have trouble powering the auxiliary/backup starter motor, might depend in part on how good your sticks are. I haven't tried it and I don't intend to.

The back-up, 'traditional' starter motor is supposed to be rated at 1000W. Julian Edgar of Autospeed, in an IC thread, wrote that he measured the starter motor load as being 132 amps, rather than the approximately 1000W/12V=83 amps the 1000W rating sort of suggests. Both of these would be quite high currents for the NiMH cells; 83 amps would be doable - about the same load the cells in the IMA pack see at full assist, for instance. Obviously good packs can do that.

But 132 amps? - that's quite a bit more. My guess is good sticks, warm, could probably start the car off the backup starter motor without much trouble or harm. But then, the only time we'd need it to do that is when it's not warm, but rather very cold. Plus, we're likely to be using not-so-great sticks for this mod...

The only time I've ever heard the 12V starter is when I turned the pack off. But I also live in a temperate climate (pacific northwest)...

Anyway, if your driving environment and circumstances don't rely on the 12V starter motor, a 2-stick 12V NiMH replacement seems like a very good alternative to me... I can actually feel the difference, too, from shedding that 22.5 lbs - the steering feels just a tad lighter (my lead-acid battery is 27 lbs, NiMH replacement is about 4.5 lbs, weight savings is the difference)...

METHODS IN BRIEF
The mod itself is quite easy - I'd say it'd be perfect for the only slightly adventurous, DIY type, with pretty minimal skills. It's totally reversible - you remove the lead-acid battery, its box and hardware, reuse half the hardware, and can put it all back if you want...

-You use the stock positive cable and buy a 'starter switch to starter' cable (about 19 inches, 12 inches might work too, at local big box or auto parts store, ~$5) to replace the shorter OEM negative cable and clamp...

-I scrounged up everything I used, stuff I had laying around. For instance, I had some 1/8 inch PVC sheeting left over from my under-engine aero panel mod - and used that for the black mounting clips.

-I had some leftover something like 1/2 inch round aluminum rod (from my rear shocks crossbar pin mod) that I made into a small boss/spacer that fits into the stock positive cable terminal clamp; the 'boss' is like the lead-acid positive terminal post and fits into the clamp. You drill a hole through that boss, so then you can just stick a bolt through it to attach to the positive stick terminal...

-I used 1/2 inch water supply copper tubing as the busbar connecting the two sticks in series, just squash the ends and drill a couple holes...

-I used 1/2 inch copper tubing as well for the negative clamp that fits around one of the cells. I couldn't find any decent copper flat bar - copper flashing would have worked, but I couldn't find it. I just split the tube in half lengthwise, flattened it out, cut it in half lengthwise again, and had two fairly thick copper strips. I bent one of them into a round clip that fits around cell 3, the negative terminal...

Etc etc...

'PERFORMANCE' AND STUFF
Your new 12V NiMH battery should stay charged to about 85% state of charge. So far, from my minimal testing, you'll have about 5.5Ah of usable capacity. Sitting in one place with the ignition on uses about 5 amps; so you'd be able to do that for only about an hour, for example.

S Keith wrote in another thread that the 12V drain with key OFF is 40mA. I read elsewhere that it's supposed to be less than 50mA. I tested voltage drop when hooking-up the battery, and saw none, so the drain has to be pretty small. At 40mA 'parasitic' drain, you'd be able to leave the car parked for about 5 to 6 days and still be able to come back and start the car. I think I'd use 5 days at this point as a ballpark figure, as the cells will also have self discharge... Frankly, in a pinch you could move the negative terminal clip to the 1st cell instead of the 3rd for extra energy/power... [edit: Mudder, 'Linsight' guru, just posted in the thread where I asked what the key-off drain is - and he says it's only 15mA. At that rate you could let the car sit for about 15 days and still be able to start it. http://www.insightcentral.net/forums/problems-troubleshooting/85466-12v-electric-drain-key-off.html#post947745 ]

A couple things, caveats:
-The way I have it setup in the images, the left cells get hotter than the right (by maybe about 15 degrees F). That's not optimal, don't know how much trouble it might cause in the long run... I don't have a catalytic converter heat shield installed, that's probably part of the issue. But the left cells in general are closer to the engine, too. At minimum, I will need a heat shield under the left cells come summer - something, like shiny aluminum or galvanized flashing, to deflect radiant heat from the engine and cat area away from the cells... It's possible that a more extensive under tray might be needed, to deflect radiant heat but also to help keep all the cells at about the same temperature.

-I've driven in sloshy, wet, rainy conditions and didn't see any spray onto the battery area... But, of course, the cells need protection, heat shrink, to be covered, etc., not only to protect from the elements, but to protect from shorts... I'm using the stock heat shrink, slit down the middle neatly, in the picts - because I've needed to monitor/check cell voltages to see how things are doing. But ultimately I'll use new, good heat shrink (if I don't enclose them totally). 54-55mm flat width heatshrink is the size; don't buy 64mm, it's too big and wrinkles a lot... It works, but looks fuggly...

-Make sure your windshield washer fluid nozzle and tubes are in good condition and don't leak. The driver-side washer nozzle is right over the battery area. The way I have it setup, though, that nozzle and tube are over the space between the two sticks. Regardless, you don't want washer fluid dripping into the area...

 

·
Registered
Joined
·
56 Posts
Very cool. Could you do all of us IMA-less cars a favor and try driving with your IMA switch off to see if it would be worth or efforts to try.
Thank you.
 

·
Premium Member
Joined
·
3,415 Posts
Very nice. I've always dismissed it because of the 12 cells.

How did you cut down to 10 cells? I swear I see 12 there.

I suspect the starter current disparity is due to the typically large voltage drop.
 

·
Registered
Joined
·
6,747 Posts
Discussion Starter #4
Very cool. Could you do all of us IMA-less cars a favor and try driving with your IMA switch off to see if it would be worth or efforts to try.
I don't think I'm quite ready to try that - because I'm almost 100% sure these two sticks I'm using won't like a 132 amp discharge (they may pull it off, but they won't like it). I couldn't really imagine using this setup for a car with a bypassed IMA pack. It's not only the starter load that might be an issue; it's also the 12V accessory loads. Without the IMA pack in place as a buffer, the DC-DC shuts down more often and thus the 12V load runs directly off the 12V battery more often... It's not that the sticks can't take those accessory loads (max is about 40 amps; typical is about 20-30 amps, at night); it's that you've only got about 5.5 amp-hours to 'play' with. For perspective, with climate control on, headlights on, and whatever the normal background 12V load, you're looking at 11 minutes of 12V battery run time (~30 amps); auto-stop for 11 minutes and your 12V battery will be dead... I've never tried a bypassed IMA - so I don't know how much, how often, the 12V battery is getting taxed directly in that situation... Seems like it could be a little tight...

I'll have to think about it. I'll consider trying the 12V starter -- with warmed-up sticks, at least...
 

·
Registered
Joined
·
6,747 Posts
Discussion Starter #5 (Edited)
Very nice. I've always dismissed it because of the 12 cells. How did you cut down to 10 cells? I swear I see 12 there. I suspect the starter current disparity is due to the typically large voltage drop.
Duh, Steve, How could you ask me that? I just connected to the 3rd cell with a ring clip around the cell body - which becomes the negative 'terminal'. The 12 cells are still there; just that, the two on the end aren't used (I'm explaining that for others who might not have caught on - because I'm sure you get it)...

Here's a pict of an earlier version, which shows the two unused cells covered in black and the connector at the '3rd' cell (didn't like the galvanized ferrous metal nor the cable connection on that connector)...
 

·
Premium Member
Joined
·
3,415 Posts
Ah... the cable-y bits at the end in the original picture looked like you were connected at the end. I see clearly now.

Very nice.
 

·
Registered
Joined
·
6,747 Posts
Discussion Starter #7 (Edited)
You should try it. It's a lot of fun particularly for someone who's worked with NiMH sticks. If you have an OBDIIC&C you can monitor 12V stuff, like DC-DC shut-off ("Dcf," not "Dci"), 12V current ("Eld"), and voltage. If anything, it's been a good excuse to monitor these things and get to know what's going on with that system... It's actually a bit weird.
 

·
Premium Member
2001 5S "Turbo"
Joined
·
10,579 Posts
Neat set-up, job well done. Be careful of sitting at a red light with your foot on the brake. the battery will probably drain quickly.
From past experience I think the DC inverter shuts down around 9 volts.

I can "trigger" a P-1443 code, if my LM battery gets down to around 10 volts and I keep my foot on the brake for an extended time. With the brake on at idle, the battery voltage starts
dropping real quick and around 8.5V I get the dreaded brake/battery lights. (And NO performance)
Going to town today to get a new one, lawnmower type. Mine was mfg in 4-12, so it is on the way out.

HTH
Willie
 

·
Registered
Joined
·
4,008 Posts
If you're worried about the 12v starter, you could just use 4 sticks in 2p config. Still a third the weight of the lead lump, and probably something like 8ah capacity from well-used sticks (assuming 4ah remaining per stick).

Sam
 

·
Premium Member
Southern California
Joined
·
791 Posts
It is okay to charge NiMH with constant voltage? Grid chargers use a current-limited source, but I don't think the line going to the 12V battery is current-limited. It just connects straight to the DC-DC's output, right?
 

·
Premium Member
Joined
·
2,885 Posts
It is okay to charge NiMH with constant voltage? Grid chargers use a current-limited source, but I don't think the line going to the 12V battery is current-limited. It just connects straight to the DC-DC's output, right?
As long as the CV is bellow the fully charged battery terminal voltage .. yes.

At the high end of the DC-DC output voltage you might eventually over charge the 10 cells a little bit .. if concerned you cold move the clamp to 11 cells .. and than even the highest DC-DC voltage won't be high enough to over charge them.
 

·
Linsight Designer
Joined
·
1,877 Posts
It is okay to charge NiMH with constant voltage? Grid chargers use a current-limited source, but I don't think the line going to the 12V battery is current-limited. It just connects straight to the DC-DC's output, right?
The line to the battery connects straight to the DCDC converter. As eq1 mentioned, as long as the DCDC's maximum output voltage is below the NiMH stack's 90% SoC (10% safety margin), the NiMH cells will naturally stop sinking current. FYI: Linsight will have complete control of the voltage on the 12V rail. Cool mod.

eq1: A note on the starter power mismatch you mentioned above... the 12V battery voltage drops substantially while starting - especially at the motor input leads - so it's possible the motor was pulling 1000 W. For example, the 132A figure you quoted makes sense if the starter motor input voltage is 7.6 V and the motor is pulling 1000 W.
 

·
Registered
Joined
·
6,747 Posts
Discussion Starter #13
....Be careful of sitting at a red light with your foot on the brake. The battery will probably drain quickly....
A quick note on this idea. I can't really tell what determines when the DC-DC is enabled or disabled. The few times I've driven in the day with most accessories off, watching this stuff, the DC-DC usually?, sometimes?, turns off during assist and during auto-stop. In these cases the 12V load is running off the 12V battery only. At night with lights on, maybe climate control too, the DC-DC never seems to turn off...

In general it seems like if the 12V accessory load is 'pretty high' than the DC-DC stays on. If you've got all 12V accessories off, except for climate control (set in AC Off Econ mode), the DC-DC turns off during assist and during auto-stop, and 12V accessories are being power by the 12V battery only...

So, sitting at a light with the brake on shouldn't pose much if any issue; the load will likely be relatively small if you're running off the 12V battery alone in the first place. So far, based on what I've observed and what I know about the 12V NiMH cells and setup, I don't think there should be any capacity problems - besides extraordinary circumstances, like using the radio for an extended period with the car not running, bypassing the IMA pack, and using the backup starter to start the car...
 

·
Registered
Joined
·
6,747 Posts
Discussion Starter #14
At the high end of the DC-DC output voltage you might eventually over charge the 10 cells a little bit .. if concerned you cold move the clamp to 11 cells .. and than even the highest DC-DC voltage won't be high enough to over charge them.
I thought about that... Until a few days ago I was under the impression that the DC-DC voltage almost always ranged between 14.1V and 14.3V; at a peak of 14.3V your 10 cells are getting charged to about 1.43V/cell; 11 cells would only be getting charged to 1.3V/cell. 1.3V is probably too low...

However, a couple days ago I saw the DC-DC output voltage range shift to 14.3V to 14.5V. It was a pretty solid shift, consistent, rather than just a minor fluctuation. There might be a couple different 12V charging algorithms depending on circumstances, like perhaps if it's really cold and/or you're running a lot of accessories, the DC-DC outputs a higher voltage?. It was cold and I was running a lot of 12V accessories when I saw the shift... If 14.5V were the peak, then 11 cells might be better...

I might shift to 11 in the summer, when the cells will be hotter - keep the state of charge lower and in the cooler, safer voltage range...
 

·
Registered
Joined
·
1,787 Posts
As a crazy test a few of us once wired up a 10 cell AA size NiCad battery pack and started a Honda 500 cc motorcycle. Nothing blew up or got hot in starting it a few times even with the headlight ON.

I tried one of those garden tractor lead acid batteries on my Lotus replica a few years ago. The engine is a GSXR-1000 super bike motor/transmission etc. If it isn't started every few days the fuel injection system doesn't inject enough fuel to act like a choke and the engine is difficult to start.

Over a period of two days using the car the battery shorted most of the cells and was shot. Unfortunately it had taken me more days of part time work make the battery box then warranty was for. :(

So I have my doubts how well the tractor batteries are for starting an Insight if the IMA system is disabled.

Otherwise good work eq1. I'd give the car a test to make sure it will start with your Ni-MH pack just in case you need to do it later. It's better to find out at home rather than somewhere else. I have an idea it will work fine.

Does anyone know what kind of cells are used in those emergency battery starter packs?
 

·
Registered
Joined
·
6,747 Posts
Discussion Starter #16
I'd give the car a test to make sure it will start with your Ni-MH pack just in case you need to do it later. It's better to find out at home rather than somewhere else. I have an idea it will work fine.

Does anyone know what kind of cells are used in those emergency battery starter packs?
They used to be lead-acid, but it seems like there's a shift toward lithium for a lot of things these days...

On starting with the NiMH 12V setup - it strikes me that there's almost never going to be a time where I'd absolutely have to start the car with that battery. As long as the IMA system is functioning, I need very little functionality from that battery in order to get the car started (and if the IMA system isn't functioning, I could probably fix it, or know what's wrong).

Also, it occurred to me that even when close to dead, NiMH hold a high voltage - like 1.2V per cell, so 12V for this battery. That would be enough to get the car's electronics fired up, and then I'd have a few seconds to start the car off the IMA pack... On top of that, I have a manual transmission - so unless there's something odd about this car that I'm unaware of, I can always push start it... Plus, I could always clip into all 12 cells instead of just 10 and get some extra juice... Seems to me the risk is very very low - that not being able to start off of this setup is almost superfluous...
 

·
Premium Member
Joined
·
2,885 Posts
On starting with the NiMH 12V setup - it strikes me that there's almost never going to be a time where I'd absolutely have to start the car with that battery. As long as the IMA system is functioning, I need very little functionality from that battery in order to get the car started (and if the IMA system isn't functioning, I could probably fix it, or know what's wrong).
I actually at one point had a 12V battery that was soo weak it wouldn't even light up the dash cluster lights anymore .. I plugged another 12V battery into 12V outlet .. which gave it enough 12V to boot up the car 12V electronics .. than I used MIMA to force start the ICE via the IMA .. Once ICE on DC-DC kicked in .. and car functioned fine.
 

·
Premium Member
Joined
·
678 Posts
... And how do you keep that engine compartment so dog gone clean?
 

·
Linsight Designer
Joined
·
1,877 Posts
There might be a couple different 12V charging algorithms depending on circumstances, like perhaps if it's really cold and/or you're running a lot of accessories, the DC-DC outputs a higher voltage?
Honda's wiring harness routes the engine coolant temperature to the DCDC converter (RED/YEL, pin [email protected] connector). This is a 0:5 volt PWM @ 10 Hz.

The service manual also describes a PWM signal that controls DCDC output voltage (WHT/GRN, pinA9). However, in my testing, that signal is always either 0 volts or 5 volts; I've never seen PWM, only:
0V=no charge
5V= [email protected] volts*

Given that I've certainly seen higher (and lower) voltages, either I haven't captured PWM data yet, or the DCDC converter itself modifies output voltage based on the engine coolant PWM signal.
 

·
Linsight Designer
Joined
·
1,877 Posts
This seems like the post to mention a (somewhat crazy) derivative idea for the really weight-conscious, who know they'll never need to use the backup starter:

Background:
1a-Linsight has complete control of the DCDC converter.
1b-Linsight can connect the IMA battery to the DCDC converter whenever it wants (as long as there's at least 7.5V on the 12V rail.

2a-The DCDC converter's load regulation is very responsive: in my bench testing, the DCDC responded to a 0 to 30 A load step in 2ms.
2b-Thus, the 12 volt rail only needs to supply current for 2 ms before the DCDC's analog circuitry responds to a 30 A load step.
2c-Thus, the 12 volt battery only needs to source 840 mJ while the DCDC converter catches up. Worst case (0-75 A load step) is still less than 2 joules.

3a-A 0.5 farad capacitor stores 49 Joules at 14 volts.
3b-A 0.5 farad capacitor stores 30 Joules at 11 volts.
3c-Thus, a 0.5 farad capacitor can source 19 Joules as it drops from 14 to 11 volts (i.e. to handle a load step while the DCDC analog circuitry responds).

Hypothesis:
A properly programmed Linsight would allow the user to replace the 12 volt battery with a 0.5 farad capacitor (or less). This is because '3c' is much greater than '2c', and thus the 12 volt rail's actual voltage wouldn't substantially fluctuate each time the current changed suddenly (i.e. high di/dt), as long as the DCDC converter was always (and I do mean ALWAYS) enabled.

With the key 'OFF' (i.e. only 15 mA sourced from 12 volt rail), if the DCDC converter were turned off for more than 3 minutes*, the 12 volt rail would drop to below 7.5 volts, and thus Linsight would turn off (and not be able to turn back on).

As an emergency backup, you could keep a standard 9 volt battery in your glove box to pull the 12 volt rail back up above 7.5 volts. FYI: a 9 volt battery can source about 10 kJ (@800 mA) before dropping below 7.5 volts... that's more than enough to charge all the various capacitors throughout the insight, including the 0.5 farad one you'd add. As soon as Linsight came online (~7.5 volts), it would connect the IMA battery to the DCDC converter, and then enable the converter, as required to keep the 12 volt rail operational.

However, since the 12 volt rail should have remained operational (i.e. not turned off in the first place), the theoretical 9V battery solution wouldn't actually be an 'emergency backup', but rather an "I want to grid charge the (dead) lithium batteries and thus need to restart Linsight." For example, if your insight wouldn't start in a parking lot, the 9V battery wouldn't save the day, because Linsight turned off for a reason in the first place (i.e. the lithium batteries were dead).

*Approximate math (because I'm too lazy to integrate):
35 Joules = 15 mC/s * 12 J/c * ~190 seconds
Note: 'mC/s' is an SI equivalent for 'mA'.
The proper way to do this would be to integrate the voltage to determine the energy under the curve as the voltage slowly dropped. I fixed the voltage to 12 J/c (i.e. 12 volts), which is close enough.

In order to keep the DCDC converter enabled, several relays must remain energized. These relays pull a combined 4.38 W when powered. To stay on the safe side, I'll round this up to 10 W. Given that the 'recommended' 36S2P Leaf pack is about 8 kWh, you could conservatively leave the car 'OFF' - and the DCDC converter enabled - for 33 days before Linsight finally turned the DCDC converter off (to prevent IMA discharge).

Theory:
You can replace the 12 volt battery with a 0.5 farad capacitor if you reprogram Linsight to always keep the DCDC enabled. In this configuration, you would need to recharge the lithium batteries at least once a month (via grid charging or driving the car with regen enabled).
 
1 - 20 of 150 Posts
Top