Recals Reconsidered

[nemystic]

Here's another interesting observation that leaves me with more questions than answers.

Today, after quite a few days without use, Sputnik was pressed into service as there was too much rain for the bicycle.

On a familiar 20 mile RT, with 18 miles of hilly terrain (some large enough to necessitate down-shifting), at least as many stops, and "putzy" drivers in front of me, there was heavy use of the IMA. The resulting FE was 92mpg, and the battery pack temperature rose from 77 deg. to 93 deg. F.

The strange thing was that the SoC indicator NEVER MOVED from 90% throughout the entire trip. (That's 2 bars from the top, 18 if the total is 20).

I've done this route many times, most often with less putzy drivers ahead, and with less increase in battery pack temperature. In most cases, the iSoC moves at least +/- 2 bars in the course of that round trip.

I've seen this "stubborn SoC indication" phenomenon many times, and my best guess as to the common denominator would be that it's usually after the car hasn't been used for at least 3 or 4 days, and/or if the ambient temperature of the car is stable for a prolonged period.

Could it be that the enigmatic SoC indication algorithm has an elapsed time component, or otherwise a threshold of activation?

My speculative thoughts about this are that after long periods of inactivity and stable ambient temperatures, the battery pack cells probably equilibrate to a consistent temperature.

Today's observations seem to confirm that the Insight's SoC indicator is not based exclusively on integrated current to and from the battery pack.
Armin's sub-pack voltage measurements would seem to indicate that voltage differentials are not a major factor.
Perhaps battery temperature measurements are a significant factor in the Insight's SoC algorithm?

There's more than a rumor that Toyota hybrids have a routine that periodically "conditions" the NiMH battery packs without noticeable impact on the car's performance during the process.

As more pieces of this "puzzle" come together, a more interesting picture should eventually be revealed.


As a clarification, I was informed that the Insight's stock assist level indicator doesn't necessarily always show the real current output. One case is during sustained, rapid acceleration in second gear. I haven't noticed this, probably because I don't do WOT driving.

[Mike Dabrowski 2000]

I was watching the battery voltage today, while I drove a rather aggressive highway route. An example: The voltage with no charge or assist was at 153, I had turned off assist and regen to let the pack equilibrate. The SOC was about 60%. I put 50A into the pack, and the voltage rose to 168v, then after a delay with no additional regen, I asked for 100A of assist. the voltage dropped to 134. The pack internal resistance at 100A, can be measured buy looking at the voltage change under known load. Each group of 12 cells is monitored with the voltage tap points. under those high current conditions, the voltage difference between the subpacks will be greatest. The time to recovery voltage after a high current condition or heavy recal , Temp rise with a heavy regen or assist cycle, also holds information that could be used to detect battery soc and health.
IamIan and I swapped the packs from my silver Insight to my red one, this morning. We were hoping to find this 151K pack was having recals so I could do some further testing. I got both an IMA code, (temp sensor problem), and a check engine light.
Tried resets and other test, but finally had to put the old pack back in. We had charged my old pack right up to the top before pulling it. When we first started the car, the voltage was at 168V. The reset caused the SOC guage to be at the bottom, and the BCM had to try to charge the pack to see where the soc was. The voltage got to 183V before limiting of regen was full, and the soc rose in a positive recal.
Armin has gathered some interesting numbers during his testing. Sorry to hear that nothing concrete jumped out of the data.

[james]

I'd be interested in adding a more accurate current meter, if it could be done fairly simply with parts I can get at Radio Shack or Jameco. (I'm not up to designing my own circuit boards.) Also battery temps: one thing I have noticed since installing YKW is that the fans are always on when I shut down after driving a bit, and sometimes even when I start. I don't remember ever hearing them before.

Of course this is summer, with temperatures hitting around 100.

[Mike Dabrowski 2000]

James,
The output of the current sensor has about a .02V/ amp. It is powered by a +- 12V signal from the BCM. The output is be bipolar for assist or regen, around chassis ground.
The last partially lit YKW display led is brightness modulated so you can see the current by looking at the last led brightness.
A high impedance DVM connected to chassis ground and the current sensor output (see BCM photo). will give you all the accuracy you should need.

[Mike Dabrowski 2000]

The pack from the Silver Insight was put in my car. Several IMA codes were set, and it was determined that there was some serious issues with the pack. Ian and I decided to pull all the subpacks and do a subpack by subpack capacity test.We are determined to get to the bottom of the IMA pack problems, and look for an alternative pack design that will hopefully eliminate the problems
The results will be put on the new webpage: Mikes Saturday Hybrid Electric Workshop.
http://www.99mpg.com/workshops/mikessaturdayhybri/
Any local hybrid fans are welcome to join in the fun.

[b1shmu63]

Mike, it looks like you have some great projects lined up for the future and sounds like you are takling some of the deeper Insight issues. I wish I lived in the neighborhood.

[Mike Dabrowski 2000]

Kip,
I wish you did as well, as I know you would be a great help with the projects.

[nemystic]

I broke recent tradition and burned over a gallon of fuel today, travelling to a remarkably scenic area for a spectacular bicycle tour. ('Perfect summer weather in New England today.)
'90.7 trip mpg with no recals, and one unsuccessful attempt at an overcharge. Heavy use of the battery pack seemed to cause some thermal related charge limiting. Usually, this is temporary, but the best opportunities were lost.

Project # 7 looks really interesting.

A little over a year ago, a work colleague who's not only a robotics expert (as many of us in this workplace are required to be) but also much better educated in advanced motion control, looked at some of the hybrid car technology and modification initiatives I shared with him, and he suggested that it should be possible to simply "dial in" (select) a fuel economy value (mpg), like setting the speed on a cruise control.
As this feature in a car might seem to be inconceivable, I'd suggest that
the primary barrier to its realization is energy storage capacity and the related costs.
Stay tuned.

[DaveW]

Mike I really wish I were closer to you, I have an idea for the battery replacement, just not sure I want to fund the operation to test. If you need any help testing the sub packs I have some of the best battery matchers out. Keep us posted.

Dave Whitesel

[Mike Dabrowski 2000]

Dave,
Ok now I am curious, what type of battery balancers do you have?And do you want to share your idea for a replacement pack?