Honda Insight Forum banner

121 - 140 of 154 Posts

·
Moderator
Joined
·
6,916 Posts
Discussion Starter #121
Added Kilowatt Hours Output to the #3 post, forgot to post it initially.

But yes it's very interesting. I'm curious to see where this ends. As soon as the Pack Whacks stop helping, I'll grid charge the pack so we can see how that helps.
 

·
Moderator
Joined
·
6,916 Posts
Discussion Starter #122
The Anatomy of a MIMA Pack Whack #4

Unfortunately, no, it turns out that run #3 wasn't some kind of breakthrough for this pack. In Run #4, we see the pack returns to behavior similar to the first runs, although there is improvement in some areas.

Car #3 Pack Whack Run #1, 2, 3 & 4 - Assist Volts, Temp, State of Charge, Amps Overlay

The improvement continues. Recalibration point is now down to 58.6% from 63.1%, an improvement of 7.6%. Since the beginning of this test, we've seen an improvement of 22% with the recalibration point moving from 71.3% to 58.6%.

Car #3 Pack Whack Run #1, 2, 3 & 4 - Assist Kilowatts Output Overlay

As you can see here, we actually took a small step back in both peak output and overall output, though obviously total kilowatt hours improved. The overall discharge curve of run #4 is more like runs #1 and #2, and run #3 clearly sticks out above the rest. Peak output dropped from 10.193kW to 10.177kW. I'm not sure if we can read into these small variations so much, as the car's systems may just not be consistent from one run to another, even if all else were equal. I'm not sure.

Car #3 Pack Whack Run #1, 2, 3 & 4 - Assist Amp Hours Overlay

Overall amp hours output improved from 1.53Ah to 1.79Ah, or 17%. Again though you can see that Run #4 is more like the first runs, and Run #3 stands out above the rest for whatever reason.

Car #3 Pack Whack Run #1, 2, 3 & 4 - Assist Kilowatt Hours Overlay

Here we begin to perhaps see evidence of the progress leveling off. Notice how in each previous run, 0.04kWh was gained on both portions of each run. However, we only improved 0.03kWh this time. Overall output increased from 0.20kWh to 0.23kWh, or 15%. Once again we can see how Run #3 sticks out above the rest. Perhaps it was just an anomaly.

Car #3 Pack Whack Run #1, 2, 3 & 4 - Assist Kilowatts Closeup

You can see how Run #4 started out OK, but then fell completely on it's face. Fairly interesting. Once again, I don't know if this is just variations in how the car's systems behave, or is actually representative of a pack condition. It could also be showing that we're reaching the ceiling of how much the pack whacks are helping this pack.

Regen:

Car #3 Pack Whack Run #4 - Regen Volts, Temp, State of Charge, Amps

I tried to be a little more aggressive with the charging this time, and as you can see the systems didn't like it. Recal occurred at a paltry 34.5% SoC, only slightly better than Run #2.

Car #3 Pack Whack Run #1, 2, 3 & 4 - Regen Amp Hours Overlay

We can see that the overall input rate was only about half of the first two runs, but also twice as much as the anomalous run #3.

Car #3 Pack Whack Run #1, 2, 3 & 4 -Regen Kilowatt Hours Overlay

Same overall profile as amp hours, with only a marginal improvement all around. The IMA light was tripped again this run, despite inputting 0.93Ah/0.16kWh before the positive recal.

To be continued....
 

·
Administrator
Joined
·
10,596 Posts
Nice data. Very pretty as well. very colourful :)

I'm interested in how the BCM gauge internal ah/wh etc calculations compare with your spreadsheet, do I/we need to make any adjustments, what is the variation like?

If after gridcharging and no further improvement is seen you could isolate/identify the weak stick it would be interesting to see what aggresive cycling say 6.5ah charge and 15ah discharge minimum using a standalone nimh charger like the robotronic via croc clips to the stick end bolts would achieve. Fan running of course.
 

·
Moderator
Joined
·
6,916 Posts
Discussion Starter #124 (Edited)
The variation is pretty much non-existent. When you round to the nearest hundredth, the Ah/Wh in/out is nearly always the same as my calculated value.

I'm graphing Run #5 now, here's how it breaks out:

BCM Gauge:
1.35Ah Out
0.168kWh Out

Spreadsheet:
1.35496Ah Out
0.168595kWh Out

So I guess when I round, that will be 0.169kWh. But really that's inconsequential for our purposes here.
 

·
Moderator
Joined
·
6,916 Posts
Discussion Starter #125 (Edited)
The Anatomy of a MIMA Pack Whack #5 - Assist

This latest run was quite interesting. The pack seems to be getting much stronger, although we're still struggling with capacity issues. Let's take a look.

Car #3 Pack Whack Run #1, 2, 3, 4 & 5 - Assist Volts, Temp, State of Charge, Amps Overlay

We'll start at the top with voltage. Notice how the systems don't back off immediately. The pack stays at 120-122V for the entire run before the negative recal. This is a pretty profound change compared to the previous four runs. To me this is a sign that the pack is definitely becoming healthier, in some ways at least.

We can see that the negative recal happened at only 65.9%, quite a step back from the last run. You can also see that the rate of decline of the SoC is faster than all the previous runs.

The amperage discharge curve is much improved. Gone is the sawtooth pattern that I noted at the beginning of this test, and the curve has taken on a more plateaued look that we've seen from healthier packs.

Car #3 Pack Whack Run #1, 2, 3, 4 & 5 - Assist Kilowatts Overlay

Kilowatt output is also much improved. On this latest run, the car sustained above 9.8kW of output for a full 4 seconds, quite the contrast to the 0.25 seconds of peak power that we saw on the first run. Peak output increased slightly to 10.206kW.

Car #3 Pack Whack Run #1, 2, 3, 4 & 5 - Assist Kilowatts Closeup Overlay

Closeup snapshot of the first few seconds of assist. You can see that the latest run is very close to the profile of the strongest previous runs, but it climbs in power faster and maintains it much longer.

Car #3 Pack Whack Run #1, 2, 3, 4 & 5 - Assist Amp Hours Overlay

Overall amp hours decreased significantly, but as you can see the rate of output increased quite dramatically.

Car #3 Pack Whack Run #1, 2, 3, 4 & 5 - Assist Kilowatt Hours Overlay

Again we can see that the overall rate of output increased significantly, but the total output was barely better than Run #2 in this test.

Regen to come soon...
 

·
Moderator
Joined
·
6,916 Posts
Discussion Starter #126 (Edited)
I do have one more Pack Whack run to post, but this weekend I've started grid charging the battery.

Some interesting notes. The amp sensor on this car seems to be wonky. Is that even possible? For some reason it always reads +0.04 to +0.08A, even when nothing is happening. I'm pretty sure I've seen a consistent 0.00A in my other car when nothing is happening? My 350mA grid charge oscillates between 0.39 and 0.43A because of this anomaly. Cycling the key does nothing. It seems it's self calibration is off? What could cause that?

Because of this the amp hours the BCM Gauge is counting are 22% high for this grid charge. No big really, just thought it was very interesting.

I started at 162V, 72.0% BCM SoC. I'm sitting at 10.5 hours and about 3.6Ah actual input now. Battery at 174V on the BCM Gauge and 55F.

Battery is still VERY out of balance. I measured the ten taps a few minutes ago, this is what I saw:

Tap 1 = 17.43
Tap 2 = 17.43
Tap 3 = 17.44
Tap 4 = 17.39
Tap 5 = 17.45
Tap 6 = 17.58
Tap 7 = 17.57
Tap 8 = 17.61
Tap 9 = 17.68
Tap 10 = 17.64

175.22V actual.

The imbalance is fairly profound. I'm going to let the charge soak longer and see if that ~0.3V gap closes.

I wonder if there's a stick in tap 9, or possibly tap 4 that is holding this pack back the most.
 

·
Administrator
Joined
·
10,596 Posts
I would give it 20hrs minimum.

The battery current sensor never really normally operates at the low level we use when grid charging so i'm not surprised it's off at the low end of the scale. I can't remember but there may be a little more resolution we can squeeze out of the sensor code in the bcm gauge. I'll see how it goes with the OBDIIC&C code as that develops. That will shortly be occupying my time. ;)
 

·
Moderator
Joined
·
6,916 Posts
Discussion Starter #128
More resolution would be excellent. :)

What I was saying though is that the amp sensor oscillates between +0.04A and +0.08A even when the engine isn't running and with no grid charge input, just with the key on. :-?

When I've grid charged in Insight #1, the BCM Gauge would read 0.2 to 0.3A. In this car, it's reading 0.3 to 0.4A, and the logged output is 0.39A to 0.43A. If we minus the error, we get 0.35A

It's showing 4.0Ah inputted at 9.5 hours, or 0.42A average input. But that is impossible since the LPC-20-350 limits current to 350mA.
 

·
Moderator
Joined
·
6,916 Posts
Discussion Starter #129
176V on the BCM Gauge, 57F, 12.5 hours total, ~4.375Ah inputted.

Tap 1 = 17.51
Tap 2 = 17.51
Tap 3 = 17.53
Tap 4 = 17.50
Tap 5 = 17.55
Tap 6 = 17.66
Tap 7 = 17.65
Tap 8 = 17.70
Tap 9 = 17.75
Tap 1 = 17.74

176.1V actual.

Gap closed a little bit, down to 0.25V from 0.29V....
 

·
Premium Member
Joined
·
4,389 Posts
Nice research Eli.
Not sure you saw the blog where I took one of the current sensors apart to see how it worked.
http://99mpg.com/blog/batterypacksexpose/batterycurrentsens/
A Hall effect sensor determines when a null balance between the actual current, and an opposing current with more turns, so it has a scale factor of ~.25V/A.
So the hall sensor is operating in the switching mode, with an analog current null amplifier.
The small current offsets you are seeing near the zero point, are likely falling into the noise zone, and what you see may be the op amps trying to find a null.
Another factor with any current sensor with iron, is a small residual magnetic field, which would cause some hysteresis when small currents are flowing.

I am seeing in our calibration process where we have tested 50 or more of the chargers that the LPC-20-350. and the LPC-35-700 have some deviation from their nominal current.
We have seen the 350ma as low as 335ma, and as high as 358ma, and larger differences in the 700ma. still a CC, but with slightly differing actual output currents.
 

·
Moderator
Joined
·
6,916 Posts
Discussion Starter #132 (Edited)
Interesting Mike. I have seen that blog, I had just never noticed an error like that before. I'm pretty sure it's not present on Insight #1.

Sure will be nice to have your charger. :)

I had to sleep, so the pack just rested 6 hours. I wanted to continue monitoring the tap voltages as I think it's interesting.

168V BCM Gauge, battery at 55ºF, 6hr rest:

Tap 1 = 16.81
Tap 2 = 16.78
Tap 3 = 16.79
Tap 4 = 16.81
Tap 5 = 16.77
Tap 6 = 16.85
Tap 7 = 16.86
Tap 8 = 16.92
Tap 9 = 16.92
Tap 10 = 16.91

168.42V actual voltage. Difference of 0.15V at rest. Resuming charging.
 

·
Moderator
Joined
·
6,916 Posts
Discussion Starter #133 (Edited)
1Hr into resuming charge, 14 hours total, about 4.9Ah inputted. Battery temp at 62F.

Tap 1 = 17.51
Tap 2 = 17.51
Tap 3 = 17.52
Tap 4 = 17.49
Tap 5 = 17.51
Tap 6 = 17.63
Tap 7 = 17.63
Tap 8 = 17.67
Tap 9 = 17.72
Tap 10 = 17.67

Total pack voltage 175.86V. Spread of 0.23V. Closing the gap :)
 

·
Moderator
Joined
·
6,916 Posts
Discussion Starter #134
Since ambient is continuing to rise as the day progresses, so has the pack temperature. This is causing the voltage to be depressed over what we saw with battery temps in the 50s.

1:30PM, 15 hours total, ~5.25Ah inputted, 68 degrees battery temperature:

Tap 1 = 17.51
Tap 2 = 17.52
Tap 3 = 17.52
Tap 4 = 17.48
Tap 5 = 17.53
Tap 6 = 17.67
Tap 7 = 17.62
Tap 8 = 17.67
Tap 9 = 17.73
Tap 10 = 17.70

175.95V pack voltage. Spread increased back to 0.25V.

--------------

2:30PM, ~16 hours, ~5.6Ah inputted, battery temperature up to 71 degrees:

Tap 1 = 17.48
Tap 2 = 17.49
Tap 3 = 17.49
Tap 4 = 17.43
Tap 5 = 17.51
Tap 6 = 17.64
Tap 7 = 17.58
Tap 8 = 17.64
Tap 9 = 17.69
Tap 10 = 17.67

Total pack voltage down to 175.62V, with a spread of 0.26V.

We can see that temperature is depressing things significantly, particularly with Tap #4. It's also possible that more and more cells are starting to see dV?
 

·
Moderator
Joined
·
6,916 Posts
Discussion Starter #136
The 17.xx readings are with the trickle charger operating. Only the 16.xx readings in post #132 were no load.. or input, rather. Just thought it would be interesting to see what they settled in at after resting for 6 hours before resuming charging again.

But yes, being able to watch the taps under load is a strategic goal... Off to look through more of the AMP catalog. ;)
 

·
Administrator
Joined
·
10,596 Posts
Don't you have a blown BCM Eli?

You could salvage the tap connector from that and make up your own switch box gizmo to a meter.

I assume the lower V taps relate to the cells higher in the case as they will be slightly warmer from the rising heat of cells underneath I assume and so have lower voltages.
 

·
Moderator
Joined
·
6,916 Posts
Discussion Starter #139
That's pretty neat Mike, giving me ideas. ;)

Here's some graphs of things:

Grid Charge 1 Subpack Voltages

Grid Charge 1 Battery Temp

Grid Charge 1 Total Battery Voltage

Total charge time is around ~22 hours, with something like 7Ah inputted.

I wish I would have started doing this from the beginning, that would have been very interesting. But I started like 9 hours in. Oh well.

It seems to have leveled off.. input is less than 200mA now. That's as good as she's going to get I guess. :)
 

·
Moderator
Joined
·
6,916 Posts
Discussion Starter #140
It was an MCM that I fried. :( That's a good idea at least though. Maybe Ron has a blown BCM he'd sell me cheap.
 
121 - 140 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