Honda Insight Forum banner

201 - 220 of 271 Posts

·
Registered
Joined
·
7,309 Posts
Discussion Starter #201
...Given voltages mentioned in this patent, it makes me think it's possible, or likely, that when a cell dips below say 1.1V, it might degrade the 'cobalt conductive matrix'. And since, in the car, there's never anything approaching a re-formation type of charge, that cobalt is lost. Self discharge and the cell's ability to cope with future incursions below 1.1V will erode the cobalt matrix even more. Eventually you end up with a cell or more with super fast self discharge and a failed pack... At least, that's my vague theory...
This statement makes it sound too much like I'm concerned about minor 'blips' below 1.1V. That's not the case. There's really two main things I'd be worried about:

-in connection with what I've learned from 'rock-bottom' charge state usage and taking tap measurements (such as the charts strewn along in above posts), I'm concerned that 'the car' doesn't actually catch cells all the time when they drop well below 1.1V.

-uneven self discharge: it's what makes risk to single cells all the more likely.

My concern is that it may be all-too easy for a single cell to be regularly, perpetually driven low. There's two 'bad' things that I'm thinking result from this (beyond the already-too-fast self discharge):

-the cells that remain high get voltage depressed; they need to be discharged down to at least 1.1V. The one low cell disables the discharge before these other cells get even close to 1.1V.

-the one low cell gets driven below 1V on a regular basis; the new idea is that this might damage the 'cobalt matrix', and this damage would result in smaller capacity, lower efficiency (i.e. worse performance), but most importantly, even faster self discharge and even less ability to cope with this kind of abuse.
 

·
Administrator
Joined
·
12,520 Posts
-the one low cell gets driven below 1V on a regular basis; the new idea is that this might damage the 'cobalt matrix', and this damage would result in smaller capacity, lower efficiency (i.e. worse performance), but most importantly, even faster self discharge and even less ability to cope with this kind of abuse.
We knew this general principle (maybe not the names or exact technicalities) from years of owners anecdotal evidence. Thanks for fleshing it out.

Once a pack throws codes if you do nothing other than reset it (punishing the weak cells) then it's inevitable demise usually follows.
 

·
Registered
Joined
·
1,711 Posts
One other thing I should probably add here. One of the long-lived factoids at IC says the BCM won't let you discharge the pack all the way. The common thing to say is that 'the BCM allows pack usage between about 20% and 80%.' I won't dig into the weeds with this (I could), but rather, let me merely point out that, with an OBDIIC&C, using the SoC reset function, you can drain the pack to pretty much truly empty. In all these years I never quite realized that was possible, until now. The '144V rebound thing' seems to be a tell-tale sign.

Even without the OBDIIC&C/resetting SoC, you can drain the pack completely, just that it takes a lot more work because you have to fight the nominal-SoC-triggered throttling behavior and either use the Calpod IMA disable switch repeatedly, to disable regen and background charge, or do some fancy shifting into neutral. Also, I should point out that it appears some BCMs do allow full discharge as a matter of course, such as the A03 I might've mentioned some posts up.
I have a soc reset box and try tricking the ima to full ,it does work but when I disable the ima with the calpod switch I still see the 4 bars of regen,so it’s hard to get it to drain down.
 

·
Registered
Joined
·
7,309 Posts
Discussion Starter #205
Eq1is there a Udd thread post you started about discharging with the paperclips,,I sent you a conversation,,,thank you,
Here's a post, in this thread, that briefly documents the process: The quintessential Insight NiMH voltage thread

Here's a link to a thread where I reexplained the process and helped someone work through it. I think that starts some time after the first few volleys of exchanges, somewhere down the thread:

My methods are always slightly evolving, but I think the gist of it remains the same as what's explained in these locations.

I have a soc reset box and try tricking the ima to full ,it does work but when I disable the ima with the calpod switch I still see the 4 bars of regen,so it’s hard to get it to drain down.
I think there might be two major 'kinds' of background charge/ICE charging, one that the Calpod switch can typically disable, the other it can't, or at least has a harder time doing it. The latter is the same or akin to the 'forced charge' or '4 bar mandatory charge' that happens after a bonafide neg recal. I think once the pack is truly near empty (or if a cell is near empty), the BCM detects that and thereafter the background charge becomes much harder to disable with the Calpod, even if you have an OBDIIC&C or 'reset device' and reset state of charge high.

Since I now routinely use my pack at low charge state, and I've been experimenting with draining it almost completely everyday, I'm often struggling to control the IMA with the Calpod - it just doesn't work as reliably 'down there'. I often have to hit it twice or more, or at just the right time to get it to work. It's weird. It's like the clutch switch check in the 'algorithm' happens at a certain time in a sequence of checks, and once the pack is truly low it just doesn't check as often - like it forgets that it's supposed to check for the clutch switch - but, once in a while, it still does...
 

·
Registered
Joined
·
1,711 Posts
After reading both the 1449 post and this one, you seem not as sure about the Udd method now as you was at the start, you highlight the need to really monitor the last 3 to 5 hours of the charge ,would you say this method is for the more advanced than the average method of charge and discharge we see on ic here. Also why would this method not work on non factory packs.
 

·
Registered
Joined
·
7,309 Posts
Discussion Starter #207
After reading both the 1449 post and this one, you seem not as sure about the Udd method now as you was at the start...
That's probably true in the sense that originally I thought it was highly unlikely-to-impossible to damage cells with this method, and now I'm not totally sure. And originally I thought it was possible that taking a cell below 0.25V was actually a part of the reconditioning, that a certain reaction probably happens and helps, but now I... kind of doubt it.

But in general it's still a lot better than working with a full pack.

you highlight the need to really monitor the last 3 to 5 hours of the charge
Generally anybody grid charging should do that. If you want to be careful you do it, otherwise you do whatever you have time for, whatever you care to do... There might be some subtle differences grid charging after a thorough ultra-deep discharge vs. any other time that would make it prudent to monitor the end of charge more closely; after an ultra-deep discharge, the end of charge voltages will be higher, there appears to be reconditioning that happens that results in higher voltages than usual. So, if you want to better gauge if what you've done is likely to have been effective, then you'd probably watch the end of charge...

would you say this method is for the more advanced than the average method of charge and discharge we see on ic here.
More advanced? In the sense that sticking paper clips in terminals and causing shorts seems like a radical thing, seems weird compared to 'discharging with a light bulb' and 'doing cycles', yeah, it might seem like an 'advanced' thing to do for a lay person, I guess. But it's really just a natural extension of full pack methods, but a better, more thorough, safer, etc. approach. Full pack methods and cycles are a very blunt and risky approach. Tap-level is way less risky and in some sense easier, as you don't have to fuss with changing light bulbs, worry about multiple voltage levels, etc.

The main thing is that full pack risks reversing individual cells for long periods and at relatively high discharge rates; tap-level reduces that risk by a couple orders of magnitude, in a couple different ways. You reduce the series from 120 cells to 12, and you have a load that's a tenth the size.

Also why would this method not work on non factory packs.
My understanding is that aftermarket cells have a slightly different formulation and do not respond favorably to deep discharges. Or the flip side: OEM cells seem to have a special formulation that allows them to be deep discharged not only without damage, but with potential benefits.
 

·
Registered
Joined
·
1,711 Posts
Well first attempt went not so well ,got the pins in right place around lunch time,20 mins later went to start car,and the 12 volt battery on and almost drain the battery that fast,what am I missing,I have the calpod switch in disable and ima pack on or off didnt make a change,,glad I wasnt far away anywhere,at work so it started of starter which I'd expect with bcm unhooked. Fill me in,do I need to just park the car for 2 weeks and unhook 12 volt battery..that would suck.
 

·
Registered
Joined
·
7,309 Posts
Discussion Starter #209 (Edited)
I've copied the above post over to that P1449 thread I linked to earlier. Makes more sense there:
 

·
Registered
Joined
·
1,711 Posts
First 5 taps finished at 4 days now moving to 2nd set of taps. No problom driving the car today to work after sitting all weekend. Bcm unhooked on both sides,starts off starter and no 12 volt drain on battery. So far so good. Am following 4 straight days of tap drain this first time around..
20200518_172110.jpg
20200518_172118.jpg
 

·
Registered
Joined
·
7,309 Posts
Discussion Starter #213
Here's a link to some contemporary 'voltage' analysis of someone's pack (cells), some new graphs, data, etc. Includes cell-level self discharge data and cell voltages after a couple grid charges...
 

·
Registered
Joined
·
7,309 Posts
Discussion Starter #216 (Edited)
Some questions about 'imbalance' have cropped-up over the last few days so I've been thinking about that a bit. I realized I now have quite a stretch of experience measuring tap voltages, keeping track, etc., and that I have something more definitive to say about it.

Looking back over my charts of tap voltages, I see that I've driven with some pretty extreme tap voltage imbalances. I can now conclude that tap voltage imbalance alone is not an issue for the BCM. You can drive with major tap voltage differences and the BCM will not bat an eye. OK, maybe it bats an eye, but it doesn't trigger any codes or the like.

Of course, imbalance will become an issue at the bottom and top of the charge state range, simply having one tap charge higher/more or discharge lower/to empty before the others. Also, imbalance can increase as you cycle up and down if you have major imbalances or mismatches. But these are different issues. Very small tap voltage differences, like as little as +-100mV, can indicate a real, underlying problem. Just that, the BCM doesn't care (or doesn't tell you it does)...

Here's one chart showing probably one of the more extreme tap voltage imbalances I've driven on. The details don't matter, you can just see that the lowest taps are around 14.2V at beginning of trip (dark grey bars) and the highest are around 15.4V, a difference of around 1.2V. The outlined bars are tap measurements after the trip, such as in auto-stop with a -1A load. Never got any codes, never have, etc...

87919


I should probably add that, yes, more extreme voltage differences can matter. For example, we know that one of the trouble codes will trigger if a tap's voltage deviates from the others by more than 1.2V for something like 24 seconds or longer within a narrow usage current range (like between -10 amps and +20 amps). But that would signal a major major cell failure. It looks like anything less than that, though, doesn't really matter, the BCM won't make a fuss...

There's also another really extreme voltage tap deviation that will trigger a code, a P1568-66. I forgot that I probably got that once due to an extreme voltage deviation. The DTC sheet says a difference of 4V for 2 seconds or more. One time, I started a drive with a couple taps that were deeply discharged. I had grid charged for about 5% worth of charge, but I think that was probably not enough to bring the cells up to the normal operational 'voltage plateau'. Upon assist, it's possible most taps were around say 15V while these deeply discharged ones momentarily fell to about 11V, or just below 1V per cell, which can be typical under the right circumstances. So, the difference in tap voltage would have been more than 4V.
 

·
Registered
Joined
·
351 Posts
Questions about your paper clip method for deep discharging:

Do you recommend grid charge first, or do you feel it's not necessary?

Is there an advantage of discharging 5 taps at a time versus discharging the whole pack with a 380 ohm power resistor?

Thanks!!!
 

·
Registered
Joined
·
906 Posts
Last night I drove for a little more than an hour on a route that gave me about a 650' elevation gain, 450' elevation loss, 150' gain, and 350' loss, with a bunch of little hills and valleys in between. I completely filled and completely exhausted the pack at least once.

The voltage measurements were not made on the pack itself but on each stick pair. This gave me 10 voltage samples per measurement period. I took about 17 measurements of all ten stick pairs every second. All in all, there were 750,000 voltage measurements taken in an hour or so of driving. Measurement started just before engine start and ended just after engine shutdown. During the last third of a drive was a two minute stop in autostop and small amount of surface street travel. The rest was at highway speeds.

I captured this (and have made several other captures) to understand the range of voltages that the IMA will cause a battery to see. I am particularly interested in this because of a potential future migration to LTO. The second plot is the same as the first, but the X axis is scaled to show what each cell in a 72 volt pack would encounter.

Some considerations:
  • This pack is not bad but not great either. The lowest and highest voltages were likely made while the pack was under heavy load (assist or regen) and include the data from the weakest stick pair. If all cells were well balanced, it's not known if the high and low voltages would reach as far as they do.
  • Similarly, one must be careful and not automatically assume that the lower chat is what a 72 cell LTO pack would see if it replaced the NiMH pack. One consideration is that the lower internal resistance of a well balanced LTO pack may reduce the voltage rise and drop during heavy assist and heavy regen. The other is that the periods of low and high voltage must not be dismissed however infrequent - they do appear and an unbalanced cell need reach the damage voltage only once to permanently degrade the whole pack.
  • Which brings to the peak measured voltages. While infrequent, I have measured 2.7 volts to be above the knee in my LTO cell's charge curve (at room temperature) where a small additional charge greatly increases the cell voltage. A cell sufficiently imbalanced on the high side could rise well above this and possibly reach a damaging voltage level. Similarly, the low voltage near 1.6 volts is on the far end of the curve on the low side, and imbalance here could lead to a cell reversing.
  • In my mind, this raises some questions about scaling the LTO pack voltage that some accomplish with the VPIN resistor and the series resistor with the BCM fooler. It would raise the actual pack voltages and may put the top pack voltages uncomfortably close to (and maybe above) the knee of the LTO charge curve.
  • This was done using a BCM ending in A01. There are a wide range of BCMs and different behavior of different BCMs, although anecdotal, have been noted in InsightCentral. So this may not reflect the voltages that might be seen in a system with a different BCM version.
To dos:
  • This is only one drive. At first glance, it seems to supports a potential match between a NiMH pack with 200 cells and an LTO pack with 72 cells. However, the match may be problematic at either end of the voltage curve. The impact of the "tails" (below 1.3 volts and above 1.45 volts) may actually be greater than shown because the steep hills that produced most of the high current assist and regen (which these voltages reflect) were only a small part of the total drive. It's possible than when one considers operation with high assist for a nearly empty pack or high regen for a nearly full pack, the breadth of voltages may be uncomfortably wide or too wide for LTO.
  • Similarly, plotting the voltages only when the assist or regeneration current is sufficiently large will likely emphasize the tails. My logger is entirely in the BCM and does not monitor current (yet).
  • Those building LTO systems may want to install a long term data logger to capture the cell voltages reported by the LTO BMS over the long haul. This could be accomplished with a Teensy 3.5 (Arduino compatible with SD card interface) and an MCP2515 module (CAN bus interface).
  • We need measurements from an actual LTO system.
87985


The lower chart is a copy of the upper chart with different scaling of the X axis. It suggests what an LTO pack might see but may be completely different for many reasons.

The upper and lower excursions are more visible when the Y-Axis is scaled to clip off the more frequently occurring voltages. The second image, again, the same NiMH data scaled differently.

87986
 

·
Registered
Joined
·
906 Posts
Here's another view, of each stick pair voltage, measured about 14 hours after the car was last driven. This pack has lit the IMA light after a long period of sitting idle, but has not done so for probably over a year. It will perform a recalibration if it has been sitting long enough.

This is essentially the exact same measurement as measuring the voltage taps on the BCM connector.

Anyone wish to guess which are the strong and weak stick pairs in this pack - the ones that are the most imbalanced under load - based on this measurement?

87987
87988

The voltages bounce around slightly due to noise in the measurement technique so I've included two screenshots 17 seconds apart.
 

·
Registered
Joined
·
7,309 Posts
Discussion Starter #220
Questions about your paper clip method for deep discharging:
Do you recommend grid charge first, or do you feel it's not necessary?
In general it depends. First, lately I don't think I recommend doing super- or ultra-deep tap discharges at the get go. I define super-deep as all cells down to about 0.5V, and ultra-deep is basically down to near zero, all cells. Merely "deep" I think of as about 1V all cells... Rather, in general, most people should probably start by measuring tap voltages, seeing if there's major low voltage outliers, and then discharging the higher voltage taps 'some', incrementally, maybe for a day or two, if there are those outliers. After that, drive some, see how much or IF anything has improved. Come home, measure again, decide if you need to do more...

I'd call this 'tap-balancing discharges'.

I think typically most pack problems probably at least in part stem from pre-mature empty and full caused by imbalanced taps, so an attempt to balance taps seems like a logical first step. And I guess I generally don't like the idea of blindly grid charging all cells to try to achieve that balance - especially if one is like getting a P1449-78 code, where some cells are near empty and others are near full. Grid charging under those conditions means overcharging a lot of cells, simply to get probably only a handful of cells up to speed...

If this light, incremental tap-balancing approach doesn't help, then I might recommend something more drastic - maybe a modest but full grid charge, followed by some driving, draining to near empty (at least neg recal), and then perhaps a deep to super-deep tap-level discharge. At this point I don't think I recommend 'ultra' deep any longer, at least, not until other lighter stages have been tried and found to be deficient.

Is there an advantage of discharging 5 taps at a time versus discharging the whole pack with a 380 ohm power resistor?
There should be. When you discharge a whole pack, you have 120 cells in series, whereas when you discharge a tap you have only 12 cells in series - and those 12 cells have been managed together, as whole. So there's like 10 times+ greater chance of subjecting a single cell to the discharge characteristics of another cell when you do the whole pack rather than a tap. I think chances of intra stick-pair cells being closer together in terms of charge level and general characteristics are higher than inter stick-pair cells.

Basically, I think what we want to avoid is subjecting single cells to extremes - the proverbial 'discharge 119 full cells and 1 empty', where the one empty gets reversed throughout the whole discharge. The 1 empty is likely to end up ... warped, different, than the others, and that difference will continue to perpetuate in subsequent usage. I think the same concept applies to grid charging.
 
201 - 220 of 271 Posts
Top