Honda Insight Forum banner

181 - 200 of 603 Posts

·
Registered
Joined
·
6,732 Posts
Discussion Starter #181
I started this thread in part based on some stuff I read in a serious book about batteries, with a chapter and sections on NiMH. In a nut shell, it argued that the memory effect was caused by over charging and evolution of an "amorphous" nickel-hydrogen-oxygen state that has a lower voltage potential than the normal stuff. Capacity basically gets locked up because of this, and deep discharges get rid of the amorphous state and bring capacity back...

One question that has puzzled me is, if this were true, and many of our packs suffer because of this, why are our packs getting over charged? I have one idea:

Basically, since only pairs of sticks are monitored, the car only sees the aggregate of the cells' behaviors, not the behavior of each cell. I think it might be possible, or maybe even likely, that some cells get over charged while others are getting weak; you could have cells with low voltages and cells with high voltages, yet the sub-pack pair as a whole has the 'normal' voltage. The car continues to charge normally based on the sub-pack pair voltage - yet some cells are already charged full, while others are not... The fully charged cells get over charged, that amorphous stuff evolves, capacity gets locked up...

I get the impression that something like this might be happening, watching IMA parameters on my OBDIIC&C during charge, assist, etc. Very subtle changes take place in the way, for instance, voltage behaves over time (weeks of driving) after an initial grid charge. Basically, these 'subtle changes' make me think that individual cells are slumping, while the rest of the cells are being made to make up for the weak ones... After a low current, slow grid charge, when presumably all the cells are as charged as they'll ever be and performing as good as they can, the pack is basically 'crisper', 'tighter'. After some weeks, I may hit the same overall pack voltage targets or what not, but it takes more effort - more charge/regen, more time; voltage rebound slumps; pack voltage drops and stays down a little more, a little longer, etc...

I guess the basic point is, I don't see how the overall pack voltages could be 'hitting the targets' etc. when some cells are slumping - unless some cells are being over driven, over charged... Meanwhile, this circumstance would seem to fit the Huggins' memory effect argument...
 

·
Registered
Joined
·
946 Posts
Last time I grid charge my battery, I drain it before. To drain it, I was driving in fifth gear at very low rpm, and I was stopping on neutral I did this until I see only 1 bar. It make's a big difference. After 2 weeks of use, the SOC it always at the top. The lower it goes is around only 3 bars missing.
 

·
Premium Member
Joined
·
1,069 Posts
I started this thread in part based on some stuff I read in a serious book about batteries, with a chapter and sections on NiMH. In a nut shell, it argued that the memory effect was caused by over charging and evolution of an "amorphous" nickel-hydrogen-oxygen state that has a lower voltage potential than the normal stuff. Capacity basically gets locked up because of this, and deep discharges get rid of the amorphous state and bring capacity back...

One question that has puzzled me is, if this were true, and many of our packs suffer because of this, why are our packs getting over charged? I have one idea:

Basically, since only pairs of sticks are monitored, the car only sees the aggregate of the cells' behaviors, not the behavior of each cell. I think it might be possible, or maybe even likely, that some cells get over charged while others are getting weak; you could have cells with low voltages and cells with high voltages, yet the sub-pack pair as a whole has the 'normal' voltage. The car continues to charge normally based on the sub-pack pair voltage - yet some cells are already charged full, while others are not... The fully charged cells get over charged, that amorphous stuff evolves, capacity gets locked up...

I get the impression that something like this might be happening, watching IMA parameters on my OBDIIC&C during charge, assist, etc. Very subtle changes take place in the way, for instance, voltage behaves over time (weeks of driving) after an initial grid charge. Basically, these 'subtle changes' make me think that individual cells are slumping, while the rest of the cells are being made to make up for the weak ones... After a low current, slow grid charge, when presumably all the cells are as charged as they'll ever be and performing as good as they can, the pack is basically 'crisper', 'tighter'. After some weeks, I may hit the same overall pack voltage targets or what not, but it takes more effort - more charge/regen, more time; voltage rebound slumps; pack voltage drops and stays down a little more, a little longer, etc...

I guess the basic point is, I don't see how the overall pack voltages could be 'hitting the targets' etc. when some cells are slumping - unless some cells are being over driven, over charged... Meanwhile, this circumstance would seem to fit the Huggins' memory effect argument...
eq1,

I'll have to say that your discussion above had me looking through your text several times to capture what you are saying.

I would think the basic cell-by-cell imbalances you are talking about makes perfect sense. And that some cells could possibly really degrade in performance, and this makes the rest of the pack seem very weak as well.

It surely begs for a BCM with 120 wire taps running into it!

Jim.
 

·
Registered
Joined
·
6,732 Posts
Discussion Starter #185
My battery manual said this about memory effect, which is quite different, but this short cycling effect is something we do to our packs all the time....
Yeah, I've seen this before. The author of the chapter I posted somewhat argues against the short cycling idea, saying that such arguments have been incomplete, inconclusive, etc. Instead, he tries to go to the heart of the matter by explaining the electro-chemistry... I don't know enough to say what's right. Yet 'my author' seems to produce a more convincing, deeper cause-and-effect type argument; he gives reasons why over charging will reduce capacity as well as reasons why a deep discharge brings it back...
 

·
Registered
Joined
·
6,732 Posts
Discussion Starter #186
eq1,
I'll have to say that your discussion above had me looking through your text several times to capture what you are saying....
Are you saying that what I'm writing isn't very clear? If so, my apologies. It's hard to summarize a bunch of nascent ideas, especially knowing that forum users usually hate long posts... Hard to capture the main ideas without getting bogged down in all the details...
 

·
Premium Member
Joined
·
1,069 Posts
Are you saying that what I'm writing isn't very clear? If so, my apologies. It's hard to summarize a bunch of nascent ideas, especially knowing that forum users usually hate long posts... Hard to capture the main ideas without getting bogged down in all the details...
Not at all.

I was trying to say (and not very clearly) that what you had said, had a lot of content, and I was rereading it in an attempt to take it all in !

It would be awesome to have individual taps for each cell, and be able to see real-time what is happening to them over time.

Thanks for the input !

Jim.
 

·
Registered
Joined
·
6,732 Posts
Discussion Starter #188
Ah, OK then...

While I'm here I might as well inject a few things. It's really hard to capture the obsessive details of my HV battery venture. I keep track of all the OBDIIC&C IMA data, recording after each trip. Overall I've been trying to get my pack to operate more or less normally at lower states of charge; the 'deeper-than-normal' discharge and grid charge are strategies I've explored (I started with stick cycling over a year ago, too). Right now I'm mainly 'exploring' the autostop discharge I did a few weeks ago. There I was able to discharge more than ever (~3000mAh), and get the pack voltage down lower than I had before (137V @0.3 amp vs. maybe 142V @10 amps). And since then my pack has been performing noticeably better.

At the moment I've dropped the state of charge down to 44% and it's still functioning normally; pack voltage (static) is something like 147-152V. When you get down there voltage-wise, voltage behaves a little differently than when it's higher, like above 158V or so; it tends to hover around a particular voltage over a much broader state of charge range, like it's been around 147-152V between something like 44% to maybe 65% SOC. When voltage is higher it tends to drop and stay down. I imagine different car's have a different 'central tendency' voltage, depending on pack condition, what's wrong with it, if anything, etc...

But anyway, I can use the pack much more liberally since I did the autostop discharge; before that I really got white knuckles when the SOC would drop below 60%. Then again, I usually drive at night, with a lot of accessories on, and background charge kicks in almost always at 65% under those circumstances - so I don't get much below 60% very often, due to driving conditions. In the daytime without any accessories, background charge isn't kicking in yet, even at 44% - though it did when I turned on the climate control...

Even if my pack punks-out on my next trip, dropping below 44%, the performance has been much better up to this point... Though I've done a lot of things to try to get the pack up to speed, I'm pretty certain the autostop discharge followed by grid charge has done the most. I kind of wish I had another ailing pack that's never had anything done to it, cuz then I could do just the autostop discharge and grid charge and see if it helps a great deal...
 

·
Registered
Joined
·
2,015 Posts
It would be awesome to have individual taps for each cell, and be able to see real-time what is happening to them over time.
While taps on each cell might be interesting, I don't think it is necessary.

Taps on each stick, or ready from the 10 two-stick factory taps would be much more useful.

Right now people don't fix things beyond the stick level.

So if I could see the stick voltage and have it logged on a chart/graph/csv dump while I was doing a pack discharge, or a pack charge, I would be in heaven.

Until we get to the point that we are running out of sticks, I do not see any reason to go beyond that level of monitoring.
 

·
Registered
Joined
·
958 Posts
Paradigm shift

Last time I grid charge my battery, I drain it before.......
The meticulous observations by Eq1, talonmike and others should compel us to rethink the notion of "Charge/Discharge". Probably "Discharge/Charge is the more reasonable sequence and for the occasional grid charge we should program the charger for a discharge cycle to PRECEDE the grid charge cycle. The Discharge/Charge program makes perfect sense considering that in most other human activity we empty before we refill.

The levels of safe discharge and soak charge appear to be undetermined at present. Also undetermined is how often or when should this be done?
 

·
Premium Member
Joined
·
1,069 Posts
Ah, OK then...

While I'm here I might as well inject a few things. It's really hard to capture the obsessive details of my HV battery venture. I keep track of all the OBDIIC&C IMA data, recording after each trip. Overall I've been trying to get my pack to operate more or less normally at lower states of charge; the 'deeper-than-normal' discharge and grid charge are strategies I've explored (I started with stick cycling over a year ago, too). Right now I'm mainly 'exploring' the autostop discharge I did a few weeks ago. There I was able to discharge more than ever (~3000mAh), and get the pack voltage down lower than I had before (137V @0.3 amp vs. maybe 142V @10 amps). And since then my pack has been performing noticeably better.....
I've been doing exactly the same thing !! But for different reasons. Still interesting non-the-less.

I have a stick with an unusually high self discharge, and normally the BCM stops all MIMA activity as high as 155 volts, sometimes higher.

So over the last week or so, I do the following:

1) Park the car at night with this stick discharged as low as the BCM will allow
2) Drive the car normally the next morning drive and regen at most stops.
2a) at work the voltage is typically close to 160 or so.
3) Continue regenning on about half the way home, then start using assist for the remaining half.
4) About two miles from home, I start running out of assist, continuing as allowed until the BCM cuts things off.
5) Let the car rest with the stick depleted again over night.

Doing the above has slowly lowered the cutoff voltage from over 155 volts to close to 150 during this time.

I will continue the above and see just how low the voltage will go.

Jim.
 

·
Registered
Joined
·
6,732 Posts
Discussion Starter #192
The meticulous observations by Eq1, talonmike and others should compel us to rethink the notion of "Charge/Discharge". Probably "Discharge/Charge is the more reasonable sequence and for the occasional grid charge we should program the charger for a discharge cycle to PRECEDE the grid charge cycle. The Discharge/Charge program makes perfect sense considering that in most other human activity we empty before we refill.

The levels of safe discharge and soak charge appear to be undetermined at present. Also undetermined is how often or when should this be done?
I think Mike D.'s doing the most "meticulous" work at this time trying to determine the impact of deep discharges and how low to go. From my reading he's found that going quite low is beneficial and so far doesn't pose too much risk for someone with a pack that's already seemingly on the way out.

As I see it, there's not really a 'paradigm shift' here, but rather, a little re-thinking, perhaps. The main thing I believe that's been missing from the pack rejuvenation/cycling talk is the significance of a deeper-than-normal discharge, preferably a very deep discharge. You usually read that going below, say, 0.9-1.0 volt per cell on average is very harmful (sometimes that even 0.9V per cell is too low). I don't think that's true, at least not with an ailing pack.

Also, I'm leaning toward believing that the discharge side of cycling is more important than the charge side, though I believe a deep discharge followed by a low current soak charge is the best for rejuvenation. Basically, if I had to draw a conclusion, I'd say that the deeper discharges free-up active material in the battery that has heretofore been locked up, and then the low current soak charge gets that active material charged up most completely, and the balance you get from the top-up is most complete...
 

·
Premium Member
Joined
·
1,631 Posts
As I see it, there's not really a 'paradigm shift' here, but rather, a little re-thinking, perhaps. The main thing I believe that's been missing from the pack rejuvenation/cycling talk is the significance of a deeper-than-normal discharge, preferably a very deep discharge. You usually read that going below, say, 0.9-1.0 volt per cell on average is very harmful (sometimes that even 0.9V per cell is too low). I don't think that's true, at least not with an ailing pack.
I fully agree. My two packs were coming up with IMA codes like 47 / 49 constantly after IMA resets last November and I deep discharged both packs below 100v. After posting, some responses were that I may have damaged my pack with cell reversal but you got to think "What do I have to lose?" I can't run the pack with constant codes because it will not pass inspection in my state (New Hampshire) and bypassing the IMA, it has the same problem with state inspection so those two were not solutions. If I damage the pack, heck, I would have to buy a new one anyway. I had nothing to lose.

Since doing those two packs in mid December, I have not had any IMA codes at all. Heck, I have not even had a single negative recal still to this day and it appears to be strong. One day in the future I may get an IMA code but I am all set to discharge again. I has been a success for me and it appears others as well

JoeCVT
 

·
Moderator
Joined
·
7,181 Posts
It's really interesting because I stumbled upon the deep discharge thing about 9mo ago, just as I was gearing up to start Bumblebee.. but I was so busy that I didn't really do anything but pin a(rather large) sticky note in my mind.

But that's the great thing about data. It's data. I know I have it around here somewhere.. LOL. Probably on my other machine. The two graphs were pretty profound though.

I recently tried to replicate the experiment, but wasn't able to. I think it was because of the particular stick I tried it again on - it was just an old worn out stick and had problems unrelated to the locking up of electrochemical material. So the extent to which this works will likely depend on individual packs, naturally...

Will continue poking and prodding at the cells in my spare time(lol).
 

·
Premium Member
Joined
·
4,389 Posts
I continue to try and determine where the limits are.
My current belief is that the longer you reverse a cell the more damage one can do, which my cell level testing had somewhat confirmed. I am charging to soak to get all the cells full, discharging until drop out, and then continuing to discharge until the second discharge plateau, then recharging, to avoid reversing those first cells any longeg than necessary, on the second discharge the drop outs from deep cycle 1 tend (so far on every pack)to hang on and all drop out together, so I can discharge deeper on the second cycle before they drop out, and therefore reach a deep discharge on the next to drop out, and in that fashion work my way down with no cell being reversed for very long.

I had run my old red Insight pack through a deep discharge 3 or more weeks ago. and then it just sat for much of that time. I had the guy from NC insight pack out, so I used the empty battery place to put my red pack in and run a few Illegal runs around the block (borrowed the plate off my red Insight) After the positive recal, I ran hard up and down the hill I live on, and the pack which had originally pretty much given up before the procedure , always with an immediate1449 IMA code and no assist or regen. held up very nicely with only a couple of bars drop in 5 full assist runs up the hill.
The pack I am conditioning right now (from the NC pack) had a whole group of sticks drop out starting at 148V, and then leveled off at ~ 125V, I discharged to 120, and it recharged last nite, so this morning I go deeper. This pack had progressed to the no 12V charging and both 1449 and 1447 codes with no IMA action at all, so it is a good candidate for this deep cycle testing.
The results will be interesting for sure.
 

·
Registered
Joined
·
128 Posts
Now the weather has finally improved and enables autostop discharging, I decided to try it out on my 2 2000 silvers. Both have IMA problems and both are equipped wit a 10mm spanner for resets.
#2 is the younger one with 200.000 km. It occasionally produces a 12V battery warning (which can be reset in motion by turning the ignition off and back on again). It produces (almost) no IMA and engine lights.
It lasted for 10 minutes on autostop with headlights and heater fan on max. Then the 12V battery light came on. The IMA battery showed no discernible improvement when driving afterward.
#1 has 300.000 km and regularly throws IMA and engine lights. I don't use the car much for long journeys because its a pain in the neck to stop every 50 km for a reset.
It lasted for 15 mins under the same conditions, the the 12V battery light came on. I restarted and then went for a drive. The IMA and engine lights came on immediately. Stop, reset, but no noticeable capacity improvement.
So the autostop disharge produced basically the same symptoms as seen while driving. But..... why the difference? Does anyone have any ideas to explain the 2 different behaviour patterns?
 

·
Registered
Joined
·
128 Posts
Hi Mike, thanks for the suggestion.
Where is the best place to connect the artificial load and voltmeter without toasting myself?
I don't (yet) have a grid charger, thought it was already too late for much improvement. I have almost zero assist. How do you see the chances of regaining some capacity?
 

·
Read Only
Joined
·
2,009 Posts
I don't (yet) have a grid charger, thought it was already too late for much improvement.
I would build or buy and use a grid charger before deep discharging. Deep discharging seems to help really bad battery packs that likely already have some damaged cells, but deep discharging may damage some good cells in the process. I would resort to deep discharging only after balance charging with a grid charger failed to help. I think it's likely after so many years that the charge levels of the cells in your battery packs are very unbalanced.
 

·
Registered
Joined
·
128 Posts
Hi Art, I feel that you must be right with the grid charger. I hadn't looked into the theme too much because it would mean putting harnesses into 2 cars, investing in or building the grid chargers, and to be honest, I thought until now that the improvement would not be significant. But now with warning lights coming on at every inconvenient moment I am beginning to rethink.
So I think it would be a good idea to exactly find out what state my battery (ies) are actually in..
So that brings me back to my question originally directed at Mike:
Can someone tell me where/how to safely connect a voltmeter and how can I deduce the state of the battery from the reading?
 
181 - 200 of 603 Posts
Top