Honda Insight Forum banner

1 - 20 of 25 Posts

·
Premium Member
Joined
·
4,389 Posts
Discussion Starter #1
NIMH batteries are tough. Some proof
Take a look at the bottom photo on my battery exposed page.
The 2 pack of prius subpacks started charging at ~85% charged.

It saw an additional 6A for 4 to 5 hours.
http://www.99mpg.com/resources/artic...rypacksexpose/

The subpacks still work just fine, as I used them today to heat some 25A Nichrome ribbon to incandesence, and they seem to be of normal capacity.
Hope to get some time to test them more thoroughly soon.

I think as a result of this accidental experiment, is clear that the packs are a lot tougher than intuition would lead us to believe.
100A out of a D cell for 100K partial cycles says it all. They are tough.

Instead of the careful computer controlled charging with the Triton,
the pack could be put into a balancing charge with a setup as simple as a variac and a rectifier, using the PTC thermal strip to signal when the pack reaches some average temp that says ok everyone is up to max 100 %. Fans running of course.
This is in practice similar to what Armin used to grid charge his whole pack at 300 MA.

Then we let the pack sit and cool before putting it back into service.
When I get the right IMA pack in my lab, that will be a great and simple reconditioning process to try.
In theory one could even do it in the car with the pack disconnected and fans running.

Lets see D cell spec is 500 full cycles to drop to 80% capacity, that means that one could rebalance their pack with a balancing charge every 6 months and still last for 250 years?

In theory one could just look at the PTC strip resistance which is in physical contact with each cell, to tell when all the cells had reached a maximum temperature, thats where all the electrical watts are turned into thermal watts, or 100% balanced and charged.
should work?
 

·
Registered
Joined
·
264 Posts
Mike what are you geting at.. are you saying there's a way to restore older battery packs or is there a way to better charge the battery etc...
 

·
Administrator
Joined
·
12,283 Posts
Interesting

Nice Pics Mike!

I think 'Armins' low current (300ma) top up/balancing charge is the way to go to boost a failing pack without much effort. You don't really have to do much with this, and it will eventually bring that weak cell up to it's max capacity, so perhaps one top up charge a week would be enough to prolong pack life. Def something I am looking at for my wife's car which only has a short commute that is hard on the pack. It doesn't have a recal/capacity issue, it would just be nice to top it off from the mains.

Sadly I just got outbid on e-bay on a nice 1A Variac for this project :(

I appreciate accesing the individual subpacks and cycling them with the triton type charge would be better, but that's much more difficult. I'll probably reserve that for when I get the code of death and have to take it out anyway to resurect it :)

Looking at the pack exposed it is not too difficult to connect a simple charger as had been described with a variac type arrangement.

I was looking at disecting one of those 100w UK240v/US120v converters which are common over here. Add some full wave rectification and decent smoothing to the 120v AC output and voltage is getting upto what we need. 120 x 1.4 = 168v

Let's keep us all updated.
 

·
Premium Member
Joined
·
4,389 Posts
Discussion Starter #4
Meow 77,
Maybe.
There is some evidence that many of the recal issues may be due to a gradual unbalancing of the 120 cell series string, an occasional full charge to equalize them, could improve the condition.
This experiment started us down this path.
http://www.99mpg.com/Projectcars/mimapackwhack/
Then IamIan did a subpack by subpack full cycle of a badly unbalanced pack and corrected the condition.
http://www.99mpg.com/workshops/mikessaturdayhybri/
The next test will have to be on someone elses pack, since the 4 we have access to are now working normally.
Peter,
The Insight pack with main switch OFF, is pretty safe to handle, as none of the exposed terminals are hot. With the switch ON, one can easily connect to the + pack output on the pack side of the main disconnect and the - directly on the negative terminal. The relays do not need to be on. Care is the key word here.
The stock pack can drop to 120VDC and get up to 177VDC at full charge current.
 

·
Registered
Joined
·
264 Posts
Hi Mike.. well... I hear you... but how can we apply this in layman's terms.. for practical useage... In other wods how can I do this on my own pack. A. without killing myself B. is there an easy way to do it?

What I've done inadvertently is that I didn't use the car for such a long time that the ima completely drained out. Then I recharged it however with the stock recharging system so that means to only 80% capacity.. how could I get around this to 100% in an easy fashion.
 

·
Registered
Joined
·
1,819 Posts
More experience with the pack rebalancing: I've just finished doing mine, and so far (one trip over the Sierra) it seems to have worked.

For background, I've owned my 2000 MT for a bit over 4 years (115K miles total). For the first 3 years or so, I had about 3/4 use of the battery. That is, in climbing up a mountain road, I could drain the pack down to about 1/4 shown on the gauge, at which point it would go quickly to zero. About a year ago, that shifted fairly quickly to where I could use less than the top 1/4 of the gauge. Further draw, or full assist, would cause a recal and often an IMA code that needed to be reset by pulling the fuse.

After doing the pack balancing per Mike's web site, I seem to be back to original condition: I can draw the pack down to the 1/4 level, where I get a quick drop to 1 bar, and then recharge. I did work it fairly conservatively, using M*MA to limit assist on climbs, so we'll see how it holds up as I drive more.

Caveat for those who want to do it: it's slow. Using a 3A charge & discharge, it took about 4 days to do all the subpacks. I started taking the pack out Monday pm, got it back together & running Saturday pm.

I also ran a temperature pickup (simple indoor/outdoor thermometer) into the pack, and found that it doesn't seem to heat much. (Which was one of my theories for the IMA codes.) Going over Carson Pass I climb from 4800 ft to about 8500, do a couple of ~1000 ft down & up over ridges, then a long downhill to near sea level. This should work the pack pretty hard, but I never saw temps over 80F, with the interior temp about 65.
 

·
Premium Member
Joined
·
1,431 Posts
I also bought one of the Triton chargers and will be doing mine soon. I have been practicing on some 'dead' NiMH laptop batteries. Had one 7.2v string that would only charge to 100 mAh after 1 charge, but after 10 cycles it went up to 600mAh which is still not good, but an improvement.
 

·
Premium Member
Joined
·
4,389 Posts
Discussion Starter #9
As a first test of another balancing technique, Ian and I ran an experiment.
We set up a string of 7 prius subpacks, and charged them in series at Armins 300MA.
The idea was to see if that charge level would heat the cells or develop high internal pressures.
The 300MA did not generate any measurable heat with the wide open cooling that we had, and after 20+ hours of charging, I finally saw some pressure buildup.
The subpacks had a small initial voltage difference that disappeared once we reached eight volts per cell, and they remained matched right to the millivolt up to the final 8.47V when I stopped charging.
After settling overnight, each subpack has exactly 8.27V.
The idea would be to use a constant current charger to gently top off the pack and rebalance whenever necessary. A simple two wire charge connection could do it right in the car?
We will follow this development here:
http://www.99mpg.com/Projectcars/mimapackwhack/
;)
 

·
Registered
Joined
·
1,819 Posts
A little more on the balancing. Discharging, of the 17 subpacks I got accurate numbers from, most had a charge between 3000-3800 mAh. 3 were just under 3000 mAh, and one was only 2000 mAh. (This was with the pack showing near full charge when removed from the car.)

After discharge, all but 3 accepted more than the rated 6500 mAh charge, with about 1/3 accepting 7000 (the Triton's safety cutoff). 2 were just slightly below the 6500 level, while the 2000 mAh one only took about 5700 mAh. (A second cycle took it above 6000, though I don't have the exact figure.) That suggests that the problems were due to the one weak subpack. If the problem comes back, I may hunt for a junked pack and pull out a subpack or two to replace the ones below 6500 mAh.
 

·
Registered
Joined
·
1,819 Posts
Mike Dabrowski 2000 said:
The subpacks had a small initial voltage difference that disappeared once we reached eight volts per cell, and they remained matched right to the millivolt up to the final 8.47V when I stopped charging.
I wonder, though, if equal voltage means equal charge? It doesn't look that way from my readings: the lowest voltage subpack did have the lowest discharge mAh, but other low voltage ones had average mAh, and some lower mAh ones had voltages on the higher side of the range.

If you had a lower-capacity cell charged to the same voltage as a higher-capacity one, wouldn't its voltage drop faster than the rest? And the BCM detect this, and limit the pack by the weakest link?
 

·
Premium Member
Joined
·
4,389 Posts
Discussion Starter #12
There are two aspects of the pack that we are exploring.
An imbalance of SOC, and an imbalance due to memory effect decreasing the useable AH capacity of a subpack.
The procedure that you used will give the subpacks a complete cycle to potentially recover some lost AH capacity. The topping off/equalization I am also suggesting would only get all subpacks at the same 100% SOC.
I will be running several more test including a discharge test, to see how well the subpacks track.
While voltage alone cannot tell us an absolute level of SOC, so far it seems that it will show the relative SOC between subpacks of the same temperature.
More testing is necessary, so keep us posted on how your pack is behaving.
;)
 

·
Registered
Joined
·
1,819 Posts
Update on the pack balancing, after about 1000 miles & 3 trips over the Sierras: It doesn't seem to have restored the pack to new condition. It did at first, but a full-assist passing of an RV on an upgrade seems to have kicked it back to something like its previous behavior, although it does have quite a bit more battery capacity available before hitting the recal (about 1/2 the gauge, vs less than 1/4 before), and it's much less likely to turn on the IMA light when it happens.

Bottom line is that it's worth doing, but not the perfect solution I'd been hoping for. I think the next thing to try will be to look around for an old/salvage battery pack, and replace the several weakest sub-packs. Anyone know of an available pack?
 

·
Administrator
Joined
·
12,283 Posts
My spare pack is on charge

Just bench testing my homemade psu at the moment and it's pushing 150ma into my old spare IMA pack. I am soak testing at the moment and using a simple 0.5a variac and constant current at 175-180v, My pack had been out of the car for over a year and was all over the place, so I'm doing a gentle charge over several days to bring it up to full, then I will cycle a few subpacks to see what happens. Once I am happy with my charger I shall use it on my cars! Both are working fine and have no IMA battery issues so I might not be the best to test the project.

My thoughts on the charger aspect of this project on the other thread are important, any ideas/comments? I have put out the feelers with several psu makers for a quote for 50-100 Universal AC input 175-180v DC output 250ma CC supplies. I'll keep you updated.

If after a pack overnight charging session the IMA light stays off for say 500-1000 miles with adequate IMA performance in a known weak pack then one/two charges per tankful of gas might keep it rolling along for ever 8)
 

·
Registered
Joined
·
1,819 Posts
Yes. Most put out about 3500 mAh on discharge, and accepted over 6000 mAh on charge, which is about what would be expected from the specs. One only put out about 2000 mAh, and accepted about 5700. I'm guessing that's the weakest one, and that when stressed, the IMA logic somehow picks up the low voltage from the weakest link, and behaves accordingly.
 

·
Premium Member
Joined
·
4,389 Posts
Discussion Starter #17
Ian and I have started to work on an isolated battery monitoring data acquisition system that will allow recording of each subpacks voltage as well as the current in and out of the pack while driving.
I had a relay isolated system set up years ago, but it was too noisy, and scanned the whole pack too slow for my taste.
We hope to be able to see if a subpack is of lower capacity, or gather some clues as to what is really happening dynamically. It seems that this may be the best way to get an understanding of all of the variables involved in the degradation process.
;)
 

·
Administrator
Joined
·
12,283 Posts
Mike Dabrowski 2000 said:
Ian and I have started to work on an isolated battery monitoring data acquisition system that will allow recording of each subpacks voltage as well as the current in and out of the pack while driving.
I had a relay isolated system set up years ago, but it was too noisy, and scanned the whole pack too slow for my taste.
;)
Not as easy as it looks, I can vouch from experience here, I had a relay system originally on my Li-Ion EV battery pack. I then looked at some Supertex HV20822FG & HV20220FG IC's which seemed to peform adequate isolated switching of individual cells/subpacks but never got round to building anything with them. I even got them to send me a few samples. Have a look at the chip specs, if you want them I'll pop them in the post free gratis as it's in a good cause ;) I have a couple of each I think. They are expensive to buy!!

http://www.supertex.com/pdf/datasheets/HV20822.pdf
http://www.supertex.com/pdf/datasheets/HV20220.pdf

Peter
 

·
Premium Member
Joined
·
4,389 Posts
Discussion Starter #19
Yep it will not be easy.
Thanks for the offer, but I think we will go with the following approach.
Take a look at the MIMA schematic.
http://www.99mpg.com/Data/resources/dow ... ematic.pdf

The MAP, TPS, and battery amp inputs are isolated with a precision analog opto isolator HCNR201. cost about $3 each. The plan is to use one of these per subpack. The accuracy is very good. The subpack under test will power the hot side.
The 20 isolated outputs will reference to a single ground, and for starters, I will use labview and a multifunction I/O card in my notebook for the data logging.
Should work fine, just a lot of work to build. We are thinking of making a board for the base circuit for a single channel, that way we can use the same isolator board for any number of subpacks.
 

·
Registered
Joined
·
88 Posts
Mike,

Shouldn't the input impedance on the MAP / TPS / other engine signals be high enough to preclude the need for isolation? Or just being extra cautious? 8)

I'd always read that analog opto-isolators are generally poor and non-linear - nice to know there are good ones out there.

It seems to me like you'd be able to bench test the pack once you have the DAQ setup working - probably safer also? Maybe start with the Prius subpacks and work your way up.

Any thought on what you expect to see?
 
1 - 20 of 25 Posts
Top