Honda Insight Forum banner

141 - 154 of 154 Posts

·
Administrator
Joined
·
10,596 Posts

·
Moderator
Joined
·
6,916 Posts
Discussion Starter #142 (Edited)
I'm super behind on posting data with all the OBDII Gauge stuff, but I did do a 2nd grid charge this weekend on the weak pack.

Grid Charge #2 Tap Voltages

The big dip in the middle is an 8 hour rest while I had to sleep. There are many interesting things to note...

Notice how the taps come into fairly good agreement after the rest, all within 0.15V of each other. This seems typical. Many of us have tested our subpacks, saw a mere 0.1V difference, and thought, "Hey, that's not so bad."

But remember that this charge is only ~350mA, and we have a ~0.26V spread. This gives you an idea of what the difference would be under 45A of regen. Not a good thing.

It seems that Tap #4 is the weakest. I wonder if a stick there is holding this pack back?

I was able to keep the pack temperature fairly well regulated this time by putting the pack fan on a potentiometer.

Grid Charge #2 Pack Temperature

I think the full pack voltage is interesting. I think what is happening here is cells reaching dV, but notice how the overall pack voltage continues to climb in our 22nd hour of grid charging. I had to raise the voltage on my grid charger three times now. It was originally only set to 176V. I don't think this is enough. It's probably set to something like 182V now, to keep pumping current into those straggling cells.

Grid Charge #2 Total Pack Voltage

It's also possible that the slight differences are due to the small perturbations in the pack's temperature. I am not sure.
 

·
Premium Member
Joined
·
4,389 Posts
You charger was calibrated today, and will ship tomorrow, with about 15 others.
Sounds like a cell that has a very high internal resistance that gets worse with heat and charging like the yellow cell # 5 here:
http://99mpg.com/blog/whatactuallygoeswr/subpack19/
The rest of the stick is pretty good. I figure that this cell was the one that set the IMA code that shut the pack down. Someday I will cut out that cell and weld in another, and give the stick some more traces.
The grid charger has an adjustable max voltage that you can set. If the charge ends because if that, rather than detecting the topping plateau, a simple voltage comparison test across the sticks with the 350 ma running should show the bad stick. pull the stick ,change the cell, and you may get some more years out of the rest of it with regular exercise.
 

·
Moderator
Joined
·
6,916 Posts
Discussion Starter #144
Thanks Mike! Looking forward to it. :)

Looking at the data more, I'm thinking the slight voltage variations must just be temperature related. It seems that most sticks mimic eachother, which shouldn't be the case if we had cells topping off and reaching dV. Hmm...
 

·
Premium Member
Joined
·
4,389 Posts
I hope to dig out my cell level tester, and spot welder once the chargers are in the mail, and do more analysis.
Without getting to the cell level, you could ruin the good cells in a stick trying to cycle the bad ones. Rip off the heat shrink, and only cycle the bad cell.
The beauty of the Insight cylindrical subpacks is how easy it is to get to the cell, strip off the heat shrink, make a copper tube press in high current clip, with screw on ring terminals, and do some tracing of the sticks that fail the over all stick level test.
Try that with a prius subpack.

A pack may fail for only one bad cell, but expect that any pack that has seen Arizona/Texas summers, may have overall degradation dur to venting cells?

Another advantage of the cell level tracing is that a 10V data acquisition system can read each cell with high resolution and no isolation is required. I have the power supplies to make a high current tester that can charge at 40A and discharge at 100, at the cell level, never had time to assemble it.

On your subpack traces, I expect that the lines staying more or less parallel in slope, but offset in voltage is reflecting the difference in average internal resistance of the sticks, if there was a big difference in AH capacity, I would expect the slopes to drift apart as a low capacity stick would charge faster than a higher capacity one?

Not enough time
 

·
Premium Member
Joined
·
215 Posts
Eli, during all your data- logging, do you have any figures showing how many
Hours it takes to bring the battery from soc 20 up to soc 75 using one of
The 350 Milliamp gridchargers.?
 

·
Registered
Joined
·
1,616 Posts
I think the charge time would greatly depend on the health of the battery. A weak battery will reach peak very fast. The longer it takes, the more capacity you have.

I use a 350 ma current limited charger and it charges to 174 volts as read on Peters battery condition meter in the dash. It takes me from 12 to 18 hours to get to 99% on an 11 year old battery, going on 12 years old.
 

·
Moderator
Joined
·
6,916 Posts
Discussion Starter #148 (Edited)
Really odd behavior with these cold temps.

This has happened twice now. It's like a partial negative recal? The SoC resets to 35%, and about 5 bars on the dash. This happened under only ~25A of assist, with the battery at over 80% SoC initially, and quite charged.. battery temp was 22 degrees. It was quickly followed by a positive recalibration up to 75%.

With a normal negative recal, the SoC resets to 25% and 0 or 1 bars on the dash.

I wonder if this is the bug they fixed with the early BCM revisions that could allow overcharging when very cold? The "Cold temp" background charge predictably stops at 82.0%. But now that the SoC is has reset back to 75%, it will allow it to go up to 82.0% again. The battery is quite topped off and is resting at 174V.

Battery temp was 13F this morning, and the systems would only allow ~11A of regen and 15A of assist.
 

·
Moderator
Joined
·
6,916 Posts
Discussion Starter #149
This strange cold temperature SoC behavior is quite repeatable..

Seems if the battery temp is below somewhere around 25 degrees, commanding assist(either with MIMA or the pedal) makes it fall to 35% SoC within a few seconds. This starts a background charge and the battery positive recals back to 75% in a few minutes.. once this happens, commanding full assist will make it drop back down to 35% again.. I've been able to do this 2-3 times on my short little 2 mile commute.

Really odd, and it's really overcharging my battery. I'm seeing 188V under a mere 12A of regen at this point. It just keeps charging it up over and over without letting any significant amount be taken out. :-?

This has got to be the cold weather bug they fixed. Really interesting to see it in action.. But I don't think the battery is too pleased.
 

·
Administrator
Joined
·
10,596 Posts
Is it a bug? It may just be one of the ways the system tries to warm the batteries up.

The fact they are being over charged doesn't really matter IMHO as long as the battery temp is within limits. In fact you could call it a bit of a balancing charge, as the repeated charging trying to warm them might be good as it would tend to remove any imbalance when they are fully charged.

Which battery is this BB or an old one?
 

·
Moderator
Joined
·
6,916 Posts
Discussion Starter #151 (Edited)
It's with the old weak pack.

Yeah, I'm not really sure what's going on. From what I've read, it was only the very first revision of the BCM that had this cold weather bug. I'll have to check, but I think the BCM I'm using is a couple of revisions in.. so maybe you're right.

Maybe I'll try putting in a late revision BCM to see if it still happens. It's nice being able to see IMA data without having to rely on the BCM Gauge now. :)

Overcharging a NiMH cell at a fraction of C isn't a big deal(like the 350mA grid charge), but high amperage(1C+) overcharging isn't a good thing. It does seem the car is pretty good at preventing any dangerous over or undercharge situation though, so maybe it's OK.
 

·
Premium Member
Joined
·
4,389 Posts
I put a scope on the HV during one of those events, many years ago, and I recall that instead of a steady current, it was a sawtooth, so I am not sure if the OBDIIC&C current is the peak or average when that is happening?
I expect the pack PTC strips and the temperature sensors keep things from getting too hot, and from what I have seen, the temperature issues are the main reason that overcharging at higher currents are a problem.
Most NIMH and NICAD chargers for power tools charge at several amps until the temperature raises to a setpoint, and then just stop.
A look at these Prius subpacks
http://99mpg.com/blog/batterypacksexpose/whathappenstoapriu/
and one would have expected them to be toast, but I have been using that same set of two subpacks as my lawn tractor 12V battery for several years, and they work great.
Sit all winter, one grid charge in the spring, and a summer of use. My lawn is big, so when using the bagger I need to stop to empty the bags and restart the lawn tractor 20-30 times in an hour mowing, and the batteries crank the 13HP engine as well on the last start as on the first.
Tough little batteries.
 

·
Premium Member
Joined
·
1,067 Posts
I put a scope on the HV during one of those events, many years ago, and I recall that instead of a steady current, it was a sawtooth, so I am not sure if the OBDIIC&C current is the peak or average when that is happening? ....
Mike,

I would not be at all surprised that the waveforms between Assist and Regen are different.

This is because I put on a 200mV mini-voltmeter display in the dash to monitor pack voltage, with a custom voltage divider at the pack taps.

Anyway due to electrical noise, the meter only reads a steady voltage when the engine off during a FAS. At all other times, the meter fluctuates all over the place.

However that fluctuation is completely different when Regenning into the pack. The voltmeter refresh rate has got to be over 20Hz and you can see "patterns" in the display fluctuations during regenning that are not present at any other time when the engine is running.

Jim.
 

·
Premium Member
Joined
·
4,389 Posts
Jim,
I have a real time DVM with a rotary switch to monitor all the taps as well as the full pack voltage. I get a nice steady voltage reading even when driving.

I used a 60 mhz storage scope to capture the regen waveforms. It was a sawtooth until the pack got over 40F then it was DC.

Nothing to do with Assist.
 
141 - 154 of 154 Posts
About this Discussion
153 Replies
23 Participants
Mike Dabrowski 2000
Honda Insight Forum
We’re the ultimate Honda Insight forum to talk about Honda’s hybrid car and its fuel economy and specs!
Full Forum Listing
Top