Honda Insight Forum banner

1 - 20 of 40 Posts

·
Registered
Joined
·
6,153 Posts
Discussion Starter #1
Someone PM'd me about grid charging his pack. I responded with kind of a summary of what I think about that stuff these days, including a couple links to some underlying nut n' bolts. I thought it might be useful as a new thread, despite the constant drip of that kind of thing around here already. Mainly, there's a few ways I look at it that seem important to me but aren't usually represented in what people advise to others, such as the limits of grid charging-only, how a healthy pack should actually perform, and the function of deep discharging... Here's most of that response:

In general, I'm not nearly the fan of grid charging that most other IC'ers are. I don't think it does most of what people think it does. Here's a link to a thread where I've posted some graphic representations of the process, grid charging and grid charging + deep discharging. https://www.insightcentral.net/forums/problems-troubleshooting/65322-graphic-illustrations-grid-charge-deep-discharge-process.html#post724378

I feel pretty confident about the general sweep of what's claimed there, though a lot of the specifics are out-dated... The graphics should help you wrap your brain around the bare 'mechanics' of charging and discharging a pack of cells in series... The gist of it is, simply, that if you need to revive or recondition your pack, a grid charge alone isn't going to do much...

As far as what your voltage readings say about the health of your pack? They don't say much of anything. The Insight NiMH cells will settle at a voltage around 1.318V at a wide variety of charge states and conditions. So, your pack voltage suggests just about that -- 159-160V / 120 cells=1.329V per cell. It just doesn't tell us much. We might be able to say that it's unlikely you have a totally failed cell...

You can't tell pack health very well unless you read and/or log pack voltages under load under various circumstances. For example, if you can bring state of charge down below 50% nominal, hit the pack with say 20 amps of assist, with the pack around 60 degrees F, and voltage stays above about 141V, your pack is probably pretty good... If you can bring charge state below 50% and still get about 4 seconds of about 80 amps of assist, at full throttle, your pack is probably pretty healthy... Anything less than these levels suggests some degree of... funk.

This is what's problematic with stuff you see/read at IC: no one looks closely enough at their pack performance to really know how they're actually performing. Most people are like, 'it doesn't neg recal' or 'there's no IMA light' - 'my pack is working great!' No, a great working pack has to hit certain metrics, certain benchmarks, in order to qualify as working great...

This is also what's insidious about the stock management and feedback about IMA functionality in the car - it's next to non-existent, you just don't know how well or bad the IMA is performing unless you dig in and take a closer look. The way the car manages the pack masks both poor management practices and poorly performing packs.

I also think the stock management causes at least some of the degradation/poor performance that creeps in seemingly fairly quickly... It charges too much, too often, too high. From my experimentation and observations over the... years, I'm coming to the conclusion that you almost can't force the pack to operate at too low a state of charge; i.e. the more you do/you can force the car to use lower charge states the better. The problem is, though, I think most packs are probably already too 'crudded-up' to get charge state very low and to have useful power/energy down there, so some initial reconditioning is needed...

Personally, I'm a proponent of 'ultra deep discharge' - very low current, prolonged, as disaggregated as possible (i.e. less than whole pack) discharging, followed by a grid charge. With this strategy I don't think it matters whether you grid charge first or not, as the load used is so small and you're preferably only dealing with at most 12 cells in series rather than 120, that cell reversal isn't an issue... Here's a link to some theory underlying this 'ultra-deep discharging' process: https://www.insightcentral.net/forums/honda-insight-forum-1st-gen-discussion/81778-tinkering-non-working-ima-battery-8.html#post1206666


I am reading grid charging threads. I am wading thru the technological boilerplate of them.... I “might” be starting to get a small amount of it.... My charger is in, and on as of 9 this evening. The starting voltage was 159-160 at 299 ma. What does this say about the health of my battery which just sat idle for 4 1/2 weeks? The temperature here is about 45f.
 

·
Registered
Joined
·
27 Posts
I agree with you about deep discharge. I bring the pack down to about 5 or 10v, and then a full grid charge. I only need to do this about once every 2 years to keep the pack running normally. Don't have any detailed metrics as you mentioned.
 

·
Registered
Joined
·
630 Posts
This is what's problematic with stuff you see/read at IC: no one looks closely enough at their pack performance to really know how they're actually performing. Most people are like, 'it doesn't neg recal' or 'there's no IMA light' - 'my pack is working great!' No, a great working pack has to hit certain metrics, certain benchmarks, in order to qualify as working great...
IMO the data is not readily accessible. I've tried to get more info.. My ECU scanner can access this information but at only a few messages per second, it's not really possible to get great data. Does one of the member diagnostic tools log enough data to do data logging of enough metrics (current demanded, current supplied, etc etc) to create a bunch of data that can be mined for trends? Or are we looking at reading data lines to get commands sent from the ECU to BCM, and reading voltages and currents to see the result of those commands?

Also, has anyone with any significant mileage been able to keep their original battery alive with grid charging and/or discharging, or are most folks on replacement batteries?
 

·
Registered
Joined
·
6,153 Posts
Discussion Starter #4 (Edited)
I agree with you about deep discharge. I bring the pack down to about 5 or 10v, and then a full grid charge. I only need to do this about once every 2 years to keep the pack running normally. Don't have any detailed metrics as you mentioned.
What discharge rate/load do you use?

The reason I ask is because it took me a long time of experimentation to see, to somewhat understand, that there's worthwhile gains doing very low current, very prolonged, very deep - "ultra"-deep - discharge. For example, looking at some graphs yesterday, I saw that my old, merely 'super deep' methodology considered something like 200mA down to 0.2V cell-level as sufficient. I was doing cells with a hobby-type discharger that would start discharging at say 1A down to 0.2V, but would then reduce current, holding 0.2V constant, until current reached 200mA. But later I shifted to 'ultra deep' - basically something like a 33Ω resistor at cell-level for as long as it'd take to not have voltage rebounding above about 0.9V max over a 12 hour period after the resistor was removed. 19.5 amp discharge graphs after each of these types of treatments show that the latter treatment produced much better results...

Which brings me to my main point. The main type of "gain" from this ultra deep discharge is obvious (in graphs) only if you're looking at a high rate discharge and the voltage. Capacity alone tells you little to nothing. In other words, looking at, for example, the graphs I mention above, the cells/sticks put out the same capacity both times, after each treatment. BUT, the curve of the second graph, after the ultra deep, is much loftier over the second half of the discharge [edit: actually it's loftier over the whole curve, but especially the second half. Compare first inflection point, after discharge begins, we see about 2.4V for the 'super deep' treatment vs. about 2.45V for the 'ultra-deep']. The cells also track better.

In a nut shell, the ultra-deep discharge can improve the power output of the cells (in addition to capacity - if they're capacity-stunted as well)... I think this is important when it comes to how the pack performs in the car, but it's often/always overlooked: If your cells can't maintain voltage under around the 50% charge state, your pack is quickly going to trigger recals. Etc., etc...

I guess I should post those graphs; this is just one example of what I've seen in other cases... These are 19.5 amp discharges on a stick, with voltage monitored on pairs of cells. Note how the roughly mid-point voltages for the lower pane are about 2.40V to 2.42V, whereas they're only about 2.33V to 2.36V for the top pane...

 

·
Premium Member
Joined
·
3,421 Posts
Personally, I'm a proponent of 'ultra deep discharge' - very low current, prolonged, as disaggregated as possible (i.e. less than whole pack) discharging, followed by a grid charge. With this strategy I don't think it matters whether you grid charge first or not, as the load used is so small and you're preferably only dealing with at most 12 cells in series rather than 120, that cell reversal isn't an issue... Here's a link to some theory underlying this 'ultra-deep discharging' process: https://www.insightcentral.net/forums/honda-insight-forum-1st-gen-discussion/81778-tinkering-non-working-ima-battery-8.html#post1206666
I agree with you about deep discharge. I bring the pack down to about 5 or 10v, and then a full grid charge. I only need to do this about once every 2 years to keep the pack running normally. Don't have any detailed metrics as you mentioned.
You agree with him, but you do it anyway? Likely that a single low current discharge to 96V (.8V/cell) will get you just as much good with far less risk.

IMO the data is not readily accessible. I've tried to get more info.. My ECU scanner can access this information but at only a few messages per second, it's not really possible to get great data. Does one of the member diagnostic tools log enough data to do data logging of enough metrics (current demanded, current supplied, etc etc) to create a bunch of data that can be mined for trends? Or are we looking at reading data lines to get commands sent from the ECU to BCM, and reading voltages and currents to see the result of those commands?

Also, has anyone with any significant mileage been able to keep their original battery alive with grid charging and/or discharging, or are most folks on replacement batteries?
Peter's OBDIIC&C permits logging to excel. SoC, pack voltage, current in and out, net Ah, etc.
 

·
Registered
Joined
·
6,153 Posts
Discussion Starter #6
Does one of the member diagnostic tools log enough data to do data logging of enough metrics (current demanded, current supplied, etc etc) to create a bunch of data that can be mined for trends? Or are we looking at reading data lines to get commands sent from the ECU to BCM, and reading voltages and currents to see the result of those commands?
Not sure I understand the second part of your response. But yeah, the OBDII&C is the tool to use to get this kind of info/data. You can log 8 parameters of your choice, such as amps, volts, state of charge, pack temp, whatever, yet I don't think you really need to log to read the health of your pack. You just need to take some quick peeks at the digital readout of amps and volts when you're subjecting the pack to, say, 20 amps of assist... Hit a slight incline, have it in 4th gear, press the throttle moderately, enough to invoke assist, and hold it there - it usually pans-out to about 20 amp assist that's easy to hold. What does the voltage readout say? You'd also be reading pack temp - cooler temps you'll get lower voltages, so you factor that into the interpretation, for instance. You'd look at state of charge (Soc) - as I mentioned, you really want to be able to be under 50% yet still get the full assist values - the more assist, the lower the SoC the better. Etc etc...

But, I don't think the OBDIIC&C is absolutely necessary to get a much better read on pack health. I think just looking at total pack voltage with a multimeter while you're driving would work. Maybe bring the BAT (charge state) bars down, well, at least off the top few bars, hit a slight incline in 4th, invoke assist at modest throttle, and it's almost always around 20 amps or so - 18 amps, 27 amps - whatever, somewhere around there. Look at your voltmeter readout and it should be holding fairly steady above about 140V. Ideally it'd be higher, but 140V is probably a good ballpark, cutoff-type value. Even just noting the change is good - voltage shouldn't drop precipitously; it should be holding fairly steady with only a little bit of gradual decline, as you hold assist as constant as you can...

To get even less technical but not so less useful, you can simply watch voltage a bit when you're holding whatever assist level as steady as you can: voltage should hold fairly steady. If voltage drops like 1 volt, 1 volt, 2 volts and lower still, that's a sure sign your pack needs attention. Good packs hold voltage fairly steady, just a little drop, gradually, and even down to quite low charge state levels... Bad packs - voltage drops obviously, never seems to stabilize. You'll hit say 147V at first few seconds of modest assist, but shortly thereafter it's just drop, drop, drop...

And the thing is, this can all happen - "this" being the drop, drop, drop, etc. - and there really isn't anything the car, the BCM, the gauges, the trouble lights - will do to tell you your pack is in bad shape. That is, until it's really bad...

Also, has anyone with any significant mileage been able to keep their original battery alive with grid charging and/or discharging, or are most folks on replacement batteries?
I don't know. I have an original battery that 'threw codes' at one point, that's when I picked it up from someone. I've done my things to it and it has, for the most part, been kind of blowing my mind. I've been at this for quite a long time now, and until this particular pack and 'these treatments' I've simply been unaware of the high performance capability of used Insight packs. I mean, I was kind of just throwing this pack together - I had been using a couple of the sticks as a 12V battery for months, for instance, and was thinking I'd be using some sticks from another pack. But then I just decided to keep it all original, despite having what seemed to be a few weak sticks plus the dubious reuse of those 12V sticks along with the others that had no such different usage. Etc. Well, after some treatment, monitoring, etc., this original pack is doing full-blown full assist at like 40% state of charge, for instance, at not very high temps (you can only get true full assist above something like 77F, so I've seen say 90 amps at 77F+, and some lower amount at lower temps, but still high values)... So, I don't know, considering that the pack was failing when i got it, just like so many other people's packs have failed and are failing, I don't see why other people's failing packs can't be reconditioned to perform like mine is... I guess that's partly why I'm bothering to write anything here at all, it makes me kind of sad to see so many people complain about their packs when I'm like going full throttle with a 60 degree F pack at 40% SoC and the pack's not batting an eye...
 

·
Registered
Joined
·
226 Posts
The moron at the back of the class

I have waded through all the threads I have found so far. It’s a bit like reading a foreign language I barely speak. But I’m learning a lot, about large and small Chrystal’s and their relative surface areas available for the electro chemical reactions that make electrons flow and their relationship to battery function, memory, flaking off at the positive electrode etc. etc. etc. you guys are doing the
Deep digging to figure out finally how to effectively keep these batteries alive. And you have been doing it for years the earliest threads start, what five, four years back. You use diagnostic tools I would need training to learn how to use.
I don’t think I am unusual in that I will not go to that extent to maintain my battery or modify my car. But I can read directions and plans. Your threads tell me you right now have it mostly figured out and if I may:
It’s this: the way the car manages the battery doesn’t exercise the battery through its full range of levels of charge from high to low. This results in the electrolytes chrystalysing in ways that limit the batterie’s range of effective operation to a narrow range of charge, say just for example from 50% to 80%
Resulting in only 30-40% of its potential power out put. But the car continues to operate like it expects 70-80% output. This stresses what’s left of the battery and in the end wears it out prematurely. Again I’m a moron when it comes to this. But I don’t think I’m alone. The take away tho, is that at this point you have come to the conclusion that a slow very deep discharge and then a complete charge afterward, done from time to time erases these chemical “memory” effects” to a great degree. Allowing the battery to last a much longer time and perform much better. What I think I see though I don’t begin to see it
In the detail some of you do. Is that you recently brought it into sharper focus than at any time before. What I think, is it might be time to creat a new up dated recipe for battery maintainance for the rest of us. An idiots guide as it were. Something simple, maybe not perfect but attainable and adequate.
 

·
Registered
Joined
·
6,153 Posts
Discussion Starter #8
a-hem, I think we're a long ways off from ever figuring out much. I don't think any of 'us' do much deep digging any longer, I know I don't - plus, there's only so deep you can go until you bump into having to know a whole field of study, chemistry, and within that some nuances, weird stuff, that no one has ever spent much time working to figure out... And then this doesn't even include all the stuff you need to know, all the stuff you'd need to do, to figure out what the car electronics and computers are really doing. Etc etc... But, having said that, I do think the ideas at the thread I linked to earlier go the farthest in explaining most of the deterioration phenomena our cells/packs see - and these explanations are based on what I gather are fairly straight forward, common concepts among chemistry people.

The point here is that you probably shouldn't take whatever I write too seriously...

Your threads tell me you right now have it mostly figured out and if I may, it’s this: the way the car manages the battery doesn’t exercise the battery through its full range of levels of charge from high to low. This results in the electrolytes chrystalysing in ways that limit the batterie’s range of effective operation to a narrow range of charge, say just for example from 50% to 80%, resulting in only 30-40% of its potential power output....
I don't really know what's going on, but I have some semi-educated hunches. The car doesn't use the full range, but I don't think that's the problem, exactly. If anything, it's that the car uses a very narrow range - nothing even close to full - and that range of use is concentrated sort of toward the top, call it 65% to 80%. It might be higher, like 75% to 90% actual, but nominally it's around 65 to 75 or 80. I don't know exactly what the link is between this kind of usage and the deterioration we experience, but it's probably more or less due to what I quote below, regarding 'large crystals'.

In general though, it just seems to me, based on the tidbits of stuff about NiMH that I've read and what I've seen experimenting with charging and discharging sticks, packs and cells, that this chemistry screams for more balanced usage. To my eye, the cells need to be used in the middle of the charge state range most of the time, and occasionally stretched out to the ends - to top and bottom. It does seem to me that, once you can start using a lower and lower charge state range, the more you use it the better. But that could simply be an artifact, a repercussion, of having the cells used in the upper charge state range their whole life. It could be that, if the cells had been used in the middle from the get-go, they wouldn't benefit from concentrated use at the bottom...

Anyway, it's not the electrolyte that crystalizes. sser2 in that linked thread mentions that, among the three or so causes of deterioration he suggests, larger crystal formation is one of them. Here's exactly what he writes about it:

"1. Large crystals of Ni (hydr)oxides with Ni oxidation states above 2. These large crystals form because the battery is never fully discharged in the car. Due to their low surface-to-volume ratio, they cannot support higher discharge currents, but can still support low discharge currents. At low discharge currents, these large crystals will be eventually converted into Ni(OH)2, and this way the larger crystals will be broken up into smaller entities, which, during the subsequent cycles, will be able to support higher currents. Another name for this phenomenon is memory effect."

So, the car concentrates usage in the upper charge state range, whatever parts of the cells correspond to that range get used more, while other parts get used less. Larger and larger crystals form in these unused portions, they support lower and lower currents, which means power output decreases in those lower charge states. This also means that capacity (amp-hour capacity) effectively shrinks - because the car's lower cutoff is in part based on voltage and voltage is now sagging more and more, earlier and earlier, in that lower range. So the car starts charging the pack earlier, more often, and usage gets even more concentrated in the upper charge state range. And the cycle keeps repeating...

I'm not sure this is the big kahuna of failure modes, but it's a big one. Somewhere in the process I think uneven cell usage has to come into play, where some cells perhaps get hotter and wear differently - because imbalance seems to be the more common problem... I think I'd say this is the most glaring problem with the car's management and probably spawns all the other problems that rear their heads along the way to neg recals and IMA lights...

....it might be time to create a new updated recipe for battery maintainance for the rest of us. An idiots guide as it were. Something simple, maybe not perfect but attainable and adequate.
I might add some material that explains my 'tap-UDD' process, ultra-deep discharge at the voltage tap level. I've thrown some of it out here and there already. I think it's probably the easiest, safest (from a cell-damage perspective), most effective way to recondition. Not much labor input, simple to do, should/seems to achieve what needs to be achieved. It does take a relatively long time though, like a couple weeks of one driving in IMA bypass mode...
 

·
Registered
Joined
·
27 Posts
What discharge rate/load do you use?

The reason I ask is because it took me a long time of experimentation to see, to somewhat understand, that there's worthwhile gains doing very low current, very prolonged, very deep - "ultra"-deep - discharge. For example, looking at some graphs yesterday, I saw that my old, merely 'super deep' methodology considered something like 200mA down to 0.2V cell-level as sufficient. I was doing cells with a hobby-type discharger that would start discharging at say 1A down to 0.2V, but would then reduce current, holding 0.2V constant, until current reached 200mA. But later I shifted to 'ultra deep' - basically something like a 33Ω resistor at cell-level for as long as it'd take to not have voltage rebounding above about 0.9V max over a 12 hour period after the resistor was removed. 19.5 amp discharge graphs after each of these types of treatments show that the latter treatment produced much better results...

Which brings me to my main point. The main type of "gain" from this ultra deep discharge is obvious (in graphs) only if you're looking at a high rate discharge and the voltage. Capacity alone tells you little to nothing. In other words, looking at, for example, the graphs I mention above, the cells/sticks put out the same capacity both times, after each treatment. BUT, the curve of the second graph, after the ultra deep, is much loftier over the second half of the discharge. The cells also track better.

In a nut shell, the ultra-deep discharge can improve the power output of the cells (in addition to capacity - if they're capacity-stunted as well)... I think this is important when it comes to how the pack performs in the car, but it's often/always overlooked: If your cells can't maintain voltage under around the 50% charge state, your pack is quickly going to trigger recals. Etc., etc...

I guess I should post those graphs; this is just one example of what I've seen in other cases... These are 19.5 amp discharges on a stick, with voltage monitored on pairs of cells. Note how the roughly mid-point voltages for the lower pane are about 2.40V to 2.42V, whereas they're only about 2.33V to 2.36V for the top pane...

This is the original battery on 2006 with ~138K miles.
I use a pair of 470 ohm, 50W resistors (mounted on a large heatsink). I run them in series at first, then parallel when V< ~70v, so always < 300mA. (Of course, the pack voltage quickly recovers after the load is removed to ~70-100v). I usually grid charge first, discharge, then grid charge as suggested. I, too, found that grid charging alone doesn't do much, but discharging below 20v does appear to help (for me).
 

·
Registered
Joined
·
6,153 Posts
Discussion Starter #10
I use a pair of 470 ohm, 50W resistors (mounted on a large heatsink). I run them in series at first, then parallel when V< ~70v, so always < 300mA. (Of course, the pack voltage quickly recovers after the load is removed to ~70-100v). I usually grid charge first, discharge, then grid charge as suggested. I, too, found that grid charging alone doesn't do much, but discharging below 20v does appear to help (for me).
This sounds about the same as the typical deep discharging that people do, though most people I think have shifted to a 3 cycle approach, with progressively lower cutoff voltages but never very low... There are potential gains with your way and the 3 cycle approach, but I don't think either of them do what an ultra deep discharge at a fairly disaggregated level can do. That's partially what the graphs above illustrate: The 'super deep' discharge is more or less analogous to the typical deep discharges people are doing, though even better. Yet still, it doesn't live up to the ultra deep discharge. The kind of difference illustrated in the graphs I think makes a sizeable, worthwhile difference in how the pack performs in the car.

I think part of what I'm saying is sort of like this: 'There's a process that can fix the cells. This process has a natural start and end point. Most people's deep discharges are part of this process and it can help, often a lot. Same with your process. The ultra deep discharge goes as close to the end of this process as possible, and there's worthwhile gains from going there.'
 

·
Registered
Joined
·
140 Posts
Before Christmas I was having IMA light problems, I did a grid charge over Christmas [IMA gauge only on one bar!] and then last week did another grid charge in a warm garage. The results has been the gauge is staying much higher, no recals, and the battery is defo performing better.
Now I have plans to do a deep discharge, Peter says take it down to 100 volts then grid charge up, and then discharge it to 75 volts and grid charge it up again. So in the near future I will be doing just that, however for the moment the battery pack is doing just fine...............fingers crossed;)
 

·
Registered
Joined
·
6,153 Posts
Discussion Starter #12
Well Mike invested considerable effort in trying to understand NiMH batteries.... You have said that you don't always understand what is happening. Most folks who have tried to understand the "art," if it can be called such, have walked off shaking their heads - plus, OPINIONS have changed over time. No need to throw away everything that has been written.
Yeah, I don't think I'd argue with what you say here. I guess I'm trying to some degree to focus on the few things I mention in the first post rather than rehashing all the trivia, all the circumstances, etc., loosely related to 'Insight packs' and 'NiMH'... Let me see if I can summarize - really pull-out the crux of the matter...

The three things I mention in the first post:

1) limits of grid charging-only,
2) how a healthy pack should actually perform, and
3) the function of deep discharging.


In my view these things tend to get lost (or have never actually gelled) in the advice usually provided and the discussion that may follow when people come to IC asking 'What should I do with my IMA?'.

I've been at this for 6 years now, early-on quite concerted efforts to figure out what's going on, and then later just kind of hodge-podge following through as best I can with whatever aspect regarding reconditioning pops up in my life. Based on the last couple years or so of experimentation with ultra deep discharge and the results, I'm becoming more confident in the idea that ultra-deep discharge can make a significant - and perhaps crucial - difference in the outcome of reconditioning efforts. I want to say that the extent to which grid charge-discharge types of reconditioning fail or don't last probably has at least a lot to do with the extent to which the grid charge-discharge process has been carried out to its fullest. Pack management and pack performance in the car hinge seemingly delicately on the balance of cells - they need to perform as closely as possible. To me it looks like one of the only ways to achieve maximum balance is to do ultra-deep discharges. Deep, super deep, sort of deep - they inevitably end up treating cells differently, cells still end up being too different, and the pack will fail sooner rather than later...

A lot of my early efforts with whole pack deep discharge didn't quite meet the mark, and then, when more people started to try whole pack deep discharge, we saw that some people's packs ended up failing (likely too-deep reversals at too high a current for too long). Then the kind of standard approach became suggestions like 'just grid charge, that should be enough,' or the 3 cycle grid charge-discharge approach with not-so-deep end voltages... I don't think I'd argue that these don't help or can't help or aren't the right thing to do in some cases. I'm questioning whether these approaches are really just a part of what seems like the real-deal process and perhaps whether it'd be better to do something else most of the time, like a tap-level ultra deep discharge...

Part of the problem is that the bar is set too low in what constitutes a reconditioned pack, a healthy pack, a pack that now works yet didn't before. Take what Peterboat writes above as a case in point:

"Before Christmas I was having IMA light problems, I did a grid charge and then last week did another. The results has been the gauge is staying much higher, no recals, and the battery is defo performing better. I have plans to do a deep discharge, Peter [Perkins] says take it down to 100 volts then grid charge up, and then discharge it to 75 volts and grid charge it up again. So in the near future I will be doing just that, however for the moment the battery pack is doing just fine."

Here the metrics used to determine pack health are the gauge staying much higher, no recals, a butt-dyno sense that the battery is performing better. Peterboat thinks his pack is "doing just fine." But is it really? One thing I'm saying is that we should use better metrics to measure pack health - and that it's not that hard to do. I mentioned some criteria earlier.

I think this is really important, maybe even the most important. Because a pack can seem to work just fine very easily yet still be seriously under-performing. It took me a long time to see that, to figure it out, and that's me being into it and really looking for it... I don't think I'm concerned so much about performance for performance's sake, but rather, I'm concerned most about the effectiveness, the longevity, of the reconditioning process.

I kind of think that, if one's reconditioning process can't get the pack to put out about 80 amps at 40% charge state, then the reconditioning process won't last long and/or simply wasn't good enough. My hunch is that few packs reconditioned according to the 'standard' approach can probably put out 80 amps at 40% charge state. My hunch is that few packs reconditioned according to the standard approach could even get state of charge down to 40% without having a neg recal... And it's not that this performance is so crucial or wonderful - it's that it means the pack will be performing better, longer in the normal day-to-day type operational circumstances. It will stay balanced better and longer. It will have less self-discharge. It will run cooler. Etc etc...
 

·
Registered
Joined
·
630 Posts
...when more people started to try whole pack deep discharge, we saw that some people's packs ended up failing (likely too-deep reversals at too high a current for too long).

...

Part of the problem is that the bar is set too low in what constitutes a reconditioned pack, a healthy pack, a pack that now works yet didn't before. Take what Peterboat writes above as a case in point:
...

Peter [Perkins] says take it down to 100 volts then grid charge up, and then discharge it to 75 volts and grid charge it up again. So in the near future I will be doing just that, however for the moment the battery pack is doing just fine."[/i]

...

One thing I'm saying is that we should use better metrics to measure pack health - and that it's not that hard to do. I mentioned some criteria earlier.
Sounds like part of the problem is that if only a final voltage is specified, and not the rate needed to get there, and monitoring for cell reversals doesn't happen, yeah, you are going to damage stuff? Perhaps if we get can get more precise with the charge/discharge procedure, and the rest will follow?

Says the guy who still hasn't finished building his grid charger. However, I do have connectors to let me tap the lines to the BCM, waiting for me to build a circuit which can measure each stick's voltage during charge and discharge, and referring to graphs plotted by others, automatically detect the onset of cell reversal. (And then what? Cease charging?)

I'm also considering conditioning my worst stick first, alone, and seeing if it subsequently its performance improves relative to the rest of the pack. Would it make more sense to attempt to bring the worst performing stick more in balance with others in the pack, before performing a full pack reconditioning?
 

·
Registered
Joined
·
6,153 Posts
Discussion Starter #14
Sounds like part of the problem is that, if only a final voltage is specified yet not the rate needed to get there, and monitoring for cell reversals doesn't happen, yeah, you are going to damage stuff?
Well, not really... I think the bar is set too low, but if people do something and the pack goes from non-working to working, it's hard to argue that that's a problem. There's all sorts of problems... People only have so much time and experience, their packs are at different conditions, we often resort to suggesting a one-size-fits-all strategy when a bit of finesse is likely called for, etc. etc... Really, I'm probably the only one who thinks there's any problem. I think there's probably a better way, with better results, and, like I said earlier, it 'makes me kind of sad' to hear how people suffer with bad packs when I'm tearing about with an 18 year-old one...

So, anyway, more to your points... Specifying a final discharge voltage like people do is fine, and at the same time, people are saying what rate needs to be used more or less, such as what wattage light bulb. The relatively higher voltage suggestions, such as what Peter suggested above - 100V, 75V - and the multiple cycle approach, is a conservative, most widely used strategy that probably picks the low hanging reconditioning fruit fairly well yet poses little risk of damage. The stepped cycles and high voltage thresholds limit exposure to cell reversals at relatively high rates, for instance... You don't need to monitor for reversals; only really severe reversal is noticeably bad...

But, from my perspective, from my experience, this kind of program just doesn't recondition as well as it should. To me, this program is full of... slop. Why are we wasting time 'cycling' 3 times at grid charge and light bulb discharge rates? Just to avoid cell reversals? Because we think the act of cycling is part of the reconditioning?

I think most of it is because of the cell reversal thing, though I think some people think the cycling in and of itself helps. I don't think it does, but I don't know for sure. I'd rather short the voltage taps for several days, 5 at a time, and then grid charge to full. Done... Maybe do it again some time later. The minuscule load and only 12 cell strings make reversal a non-issue. And at the same time, you're actually taking all the cells down to a level and at a rate that I see as essential for thorough reconditioning. When you do full pack discharging AND you're using say a 25 watt light bulb - it's just too much of a load, and you're not going low enough anyway...

I'm being a bit hyperbolic here in the interest of trying to make some points stick out a bit better. In reality I'm not so antagonistic toward the 'existing programs'. I can see why they are what they are...

Perhaps if we can get more precise with the charge/discharge procedure the rest will follow?
May be.

Says the guy who still hasn't finished building his grid charger. However, I do have connectors to let me tap the lines to the BCM, waiting for me to build a circuit which can measure each stick's voltage during charge and discharge, and referring to graphs plotted by others, automatically detect the onset of cell reversal. (And then what? Cease charging?)
Um, monitoring and stuff is cool, but I don't think it's necessary. Just use a really low discharge load, as low as you can stand to wait for. Even on a full pack, I think a low load very deep would be better than the typical loads not-so-deep. The real risk exists for people who are first time reconditioners, with packs at God only knows what level of imbalance. Typically you can have cells that are almost empty and others that are half full. The full pack deep discharge on such a pack, starting with say a 100 watt bulb, can ruin the pack if you're not paying attention. I just wouldn't bother - too many details: 'Use this bulb to this level, swap this bulb for another, go down to this voltage, grid charge, swap back the other bulb, go down to this voltage,' etc etc...

I'm also considering conditioning my worst stick first, alone, and seeing if subsequently its performance improves relative to the rest of the pack. Would it make more sense to attempt to bring the worst performing stick more in balance with others in the pack, before performing a full pack reconditioning?
If you want to dig into stick level work because it seems like it will be interesting, there's no harm with that. But I don't think it's necessary. There's a bunch of us who have graduated from the School of Insight Sticks, only to find that it was probably a waste of time... OK, a waste of time in terms of becoming a practical exercise for the masses... I learned a lot, a lot of it was interesting. I mean, really, you have to look at what's going on at the cell-level to understand what needs to be prescribed at levels above. My whole shtick here is about achieving cell-level results yet not having to go to that extreme... But, to answer your question, I don't see how you can do what you suggest without a lot of work... How have you identified your worst performing stick? Does that mean you know how the rest are performing? (These are rhetorical questions).

In general, it makes sense to have balanced sticks before you do any full pack work. But then, it makes no sense to go about balancing your sticks, digging into the pack... That's why there's the conservative 3 cycle, stepped, not-so-low voltage grid charge/discharge regime. And why I'm suggesting a tap-level ultra-deep discharge one...
 

·
Administrator
Joined
·
10,771 Posts
Not much to add.

But I have just left a pack deep discharging on my bench while i'm away for ten days..

Basically I cycled it as per my original thoughts, and now when it is discharged to around 144v nominal I have attached a fixed 3.2k 10w resistor too it to continue until empty at <50ma rate and decreasing quickly as the voltage dwindles.

The low current and long period i hope will minimise any damaging cell reversal.
It's almost like an exacerbated/faster self discharge..

I'll report on the on/off load pack voltage when I get back.
 

·
Registered
Joined
·
6,153 Posts
Discussion Starter #16
That sounds about right... When I was working with cells, I found that about a 20mA rate at about 0.7V was enough to 'cut through the fat' in a reasonable time, but not too much to blow right through it. At 3.2k, you'd be looking at about a 26mA rate at 84V... As far as I can tell, once voltage gets really low (discharging a string), voltage on 'stubborn' cells will rise, such as out of slight reversal, and those cells 'burn' again, until they all even out. Cells sort of oscillate up and down in voltage until they peter-out in the low double-digit mV level...
 

·
Registered
Joined
·
140 Posts
In my case I am lucky to have Peter Perkins on the doorstep to help, also my pack has only done 51k [december 2005 car imported from Japan] so it seems to have responded to grid charging.
I will be following Peters instructions using a low wattage bulb to see if that helps, however over the last couple of days I have done 700 miles and the IMA light has remained off.
I will plug my diagnostics into it today and interrogate it to see if the IMA fault is still present.
I have to say that this forum is the reason I bought this car as it provides the help to run what is a very rare car with specific needs that move it out of mainstream motoring.
On another point we are thinking of buying a CZR however I am thinking of either an 2010 one or going later for a 2013 model year with the lithium battery pack, what do the experts think?
 

·
Registered
Joined
·
6,153 Posts
Discussion Starter #18
In my case I am lucky to have Peter Perkins on the doorstep to help, also my pack has only done 51k so it seems to have responded to grid charging....
You have an original pack with only 51k? If that's the case I wouldn't think it'd need much to be reconditioned into being a top performer... Do you have an OBDIIC&C? If so, you should do some testing - pull the state of charge down (preferably below 50%), hit it with full assist: What's the voltage? Is it fairly steady or does it drop rather fast?
 

·
Registered
Joined
·
140 Posts
You have an original pack with only 51k? If that's the case I wouldn't think it'd need much to be reconditioned into being a top performer... Do you have an OBDIIC&C? If so, you should do some testing - pull the state of charge down (preferably below 50%), hit it with full assist: What's the voltage? Is it fairly steady or does it drop rather fast?
I will have a check at it today if I get time, I am busy with my boat fitting an electric drive so that is the most important thing at the moment.
 

·
Registered
Joined
·
6,153 Posts
Discussion Starter #20
Thinking I should post sser2's chemistry explanations of why ultra-deep discharge works directly in this thread. This was from the end of 2015, originally posted here: https://www.insightcentral.net/forums/honda-insight-forum-1st-gen-discussion/81778-tinkering-non-working-ima-battery-6.html#post912394

Lightly edited, highlighted:

"Although I didn't want to theorize, I cannot help doing it because the chemistry behind the deep discharge seems pretty straightforward. To my excuse, I was a chemistry geek at high school, and took extensive courses of inorganic and physical chemistry as part of my undergrad studies. I am a PhD in biochemistry, and chemistry is a significant part of my professional activities.

So, in lay terms:

The negative (MH) electrode is of little concern from the point of view of cell's deteriorating capacity: it is purposely designed to have larger capacity than the positive (nickel) electrode. However, upon reversal down to -1.25 V, atomic oxygen will corrode it, causing reduction of its capacity. Atomic oxygen is a very potent oxidizer. Nothing can stand to it but noble metals like platinum.

Since the positive electrode is limiting to cell's capacity, its degradation is the key to capacity loss.

There are three major modes of degradation:

1. Large crystals of Ni (hydr)oxides with Ni oxidation states above 2.
These large crystals form because the battery is never fully discharged in the car. Due to their low surface-to-volume ratio, they cannot support higher discharge currents, but can still support low discharge currents. At low discharge currents, these large crystals will be eventually converted into Ni(OH)2, and this way the larger crystals will be broken up into smaller entities, which, during subsequent cycles, will be able to support higher currents. Another name for this phenomenon is memory effect.

2. Formation of γ-NiOOH. The normal isoform of the NiOOH in the cell is β, which is readily converted into the Ni(OH)2 during discharge. γ-NiOOH, which slowly accumulates as battery ages, is converted to Ni(OH)2 with more difficulty, and only after all the β isoform is gone. γ-NiOOH accumulates because the battery is never fully discharged in the car.

γ-NiOOH is bad in two more ways. It crystalizes with considerable amount of water, which sequesters water from the electrolyte, making the electrolyte less conductive. The consequence is increased internal resistance. When γ-NiOOH is eliminated following deep discharge, this sequestered water returns back to the electrolyte, and internal resistance decreases.

Sequestration of water in γ-NiOOH dramatically increases its volume. Whereas the β-NiOOH<=>Ni(OH)2 transition is associated with only 1-2% volumetric change, the γ-NiOOH<=>Ni(OH)2 transition alters the volume by a whopping 40%. It is therefore much more destructive to the active layer of the electrode.

3. The active Ni (hydr)oxide layer of the positive electrode may lose contact with Ni foil support, or flake off, which takes it out of commission. The charge-discharge reactions on the positive electrode involve small, yet consistent volumetric expansion and contraction of the (hydr)oxide layer. This eventually causes material cracking and flaking, which in turn reduces the capacity of the electrode. γ-NiOOH exacerbates flaking of the active layer.

Ultra deep discharge remedies conditions 1 and 2 by eliminating the large crystals and γ-NiOOH. Interestingly, ultra deep discharge may also remedy condition 3. As the cell gets below 0.25V, the nickel metal support of the positive electrode forms a galvanic pair with the negative electrode, and conversion of nickel into Ni(OH)2 begins:

Ni - 2e- = Ni2+
Ni2+ + 2OH- = Ni(OH)2

As Ni2+ ions are formed, they immediately react with OH- ions (from KOH) to form the insoluble Ni(OH)2, which is deposited at the electrode. Thus, the lost active layer of the positive electrode is replenished."

* * *

All of these explanations are pretty consistent with what I've read in other places. For example, Huggins discusses γ-NiOOH formation. Mike D. used to come across dried-out, patchy separators when he opened-up cells, which led him to believe that electrolyte loss was one of the main failure modes. That's consistent with what sser2 says about γ-NiOOH, electrolyte sequestration, major increases in volumetric expansion and contraction, and cracking and flaking. I've seen the 'cracking and flaking' idea suggested as one among few major failure modes in NiMH posed in research papers... etc etc...

Of course, at the 'street-level', a lot of the interest in 'deep discharge' stems from what stick putzers have seen after dealing with deeply self-discharged sticks: charge them up and they're usually miraculously healed. In the old days, I used to cycle sticks 3-5 times at about a 1C rate - that was the program (everybody was doin' it). The cycles didn't go lower than about 5.4V - the 0.9V per cell cutoff. You can do that all day yet never see the gains you get simply from charging a deeply self-discharged stick. Again, it's like there's a full process, from start to finish, of which these types of cycles are merely a part...

At some point maybe a year ago or so, I experimented with ultra-deep discharge on already deeply self-discharged cells and sticks. I found that even ultra-deep discharging cells that were already deeply self-discharged would produce worthwhile gains... Typically, self-discharged cells end up at a voltage above about 0.5V, I have usually seen right around 0.68V. Perhaps any gains going even lower have to do with what sser2 suggests in point 3 above, where going below 0.25V can end up converting Ni to Ni(OH)2...
 
1 - 20 of 40 Posts
Top