[twriter]

I have found a way to make my battery level guage recaibrate consistently.

From the information on this site and ellsewhere, the battery guage isn't always a true indicator of actual battery level and every so often recalibrates itself. This recalibration seems always to be in the negative: it recognizes that it's registering more than actual battery power available and sets itself down to zero (but never the opposite).

If I travel about 30 miles or more along a fairly flat freeway (and not into any headwind) at constant freeway speeds, I'll get the battery guage up to full. Every time I get it up to full, and I mean every time, this way, within about the next 5 miles or so of driving, the battery guage performs this reset.

If I spend most of my time driving short trips around the city, I'll never get the guage much above half way. And during all this time, the guage never resets this way. This happens only after I've spend significant time on a freeway and get the guage registering full that way. This is so consistent, I can predict it happening. I just can't figure out why it works this way all the time.

I'm driving one of the early 2000 models and have put more than 60,000 miles on it.

 

[Rick]

Sounds like possibly the behavior of a deteriorating battery pack. It could also be the MCM, I believe there is an update. Also, have you had the ECU replaced? I've heard of someone locally who had ~50,000 miles on his and was getting one or two recalibrations on the way to work each morning. He finally got a new battery pack.

 

[Will M]

For a gas guage, they just have a float that sits on top of the fuel and mechanical linkage measuring how high the float is floating. Even doing this, the meter can be imperfect. For the electrical charge guage, the job is a lot harder.

They can easily measure voltage, but that is a bad indicator of State of Charge (SOC). There is a "hysterisis" effect between the state of charge and the voltage. In other words, if the SOC changes, then settles at a new level, the voltage will lag a while before it changes to reflect this change in SOC. Voltage also has other short-term modifiers than SOC. Voltage can go up and down quite a bit just by varying the load.

It would be better to measure the amperage flow. If electricity were water, voltage would be water pressure, while amperage would be volume. Unfortunately, for systems like this, you can't really measure amperage directly. Instead, you put a very small resistor (to not use up much electricity) and measure the voltage just ahead and just behind this resistor. There's a slight voltage drop as the electricity flows through this resistor.

The system measures this voltage drop and mathematically calculates the amperage from there. Complex systems have multiple points of potential failure. If the resistor becomes damaged and changes its resistance, or if either of the two volt meters becomes inaccurate, then you don't get good measurement of amperage.

Meanwhile, measuring the amperage accurately is only one piece of determining the state of charge. You also need a computer with memory to record the flow of amps over time, because amps alone won't give you SOC. You are actually measuring "amp hours".

The batteries are supposed to hold a certain number of amp hours of charge. You measure how many amp hours that go in and how many amp hours go out, and factor in an efficiency ratio. In other words, because charging the battery also generates some heat, that is a loss in energy. Discharging the battery also gives off some heat, so that is more energy lost.

Given that, you can only expect to get back maybe 90% of what you put into a battery. So, to determine the SOC, you have to start with the battery in some known state (fully charged or fully discharged), then count all the amp hours that go in and all the amp hours that go out, factor in the 90% (or whatever) efficiency and then you have a rough guess of what the SOC is.

But if the batteries become damaged or worn out, then the efficiency part of the math will be wrong and the system will repeatedly find the batteries at a voltage level it associates with being fully charged or fully discharged when all the math says the battery OUGHT to be at some other SOC.

That's when you get a recalibration. The computer says the battery ought to be at, say 50% charge, but suddenly, it notices the voltage level down to a point that shouldn't happen unless the battery is fully discharged, so it says, "Okay, let's throw out our previous 50% guess and start back at zero."

It could also see a voltage level up to something that should be impossible unless the batteries are fully charged, when the math told it the batteries should be at 50%. Again, it throws out the old guess and starts over with a recalibration.

So, the resistor could be bad, one or both of the volt meters could be bad, the wiring between the resistor and either of the volt meters could be bad, the RAM in the computer that records the amp hours could be bad, the batteries could be at some efficiency level that doesn't match the programmed math, the programmed math might be imperfect when factoring in something like heat that might be different from what the programmer was expecting, or a wire somewhere in the system might be bent enough to effect a voltage level or an electronic signal or a connector might have a poor connection because a bolt is loose or some corrosion has built up...

Meanwhile, sitting in the driver's seat, we just expect all this to work. We expect the SOC guage to be as consistent and accurate as a gas guage (which we expect to be more perfect than it is). We are clueless at how remarkable it is that these things ever work at all.

 

[Armin]

stick with the facts, Will!

Hey, Will,

I agree with most of what you say, but I think I need to correct you a little here...

Unfortunately, for systems like this, you can't really measure amperage directly.

True.

Instead, you put a very small resistor (to not use up much electricity) and measure the voltage just ahead and just behind this resistor. There's a slight voltage drop as the electricity flows through this resistor.

Nooooo!

While what you describe is one way of measuring current, the Insight doesn't use it.

The Insight uses a hall-effect current sensor. In fact, several of them in different places. It basically measures the magnetic field around a current-carrying conductor and calculates the current from that. But the same basics are true for this as for any measurement: there are some errors induced. We don't know how accurate the current sensor really is. (or the voltage sensor or anything else...)

 

[Armin]

Every time I get it up to full, and I mean every time, this way, within about the next 5 miles or so of driving, the battery guage performs this reset.

Twriter,

first I got really excited reading the subject line of thir thread. But looking at it, it seems you just have a very severe cas of the recal problem. Extremely severe, I might say. If mine was this bad, I'd be at the Honda dealership right now, holding the Service Manager hostage until someone resolves the problem Just kidding.

But seriously, if I understand what you are saying, your SoC gauge never ever shows full (19 or 20 bars) for more than a few miles. You are perpetually recovering from the last recal. Most of the time, unless you use the procedure you described, the next recal happens before you recover from the previous one.

What is your driving pattern? city/highway? Trip length? speeds? aggressiveness? lmpg?

 

[twriter]

My driving pattern, when I've been working, has been something like this:

Communte, M-F, typically 30-50 miles. Both my home and my work are typically close to freeways, or expressways, so I do just a small amount of city-street driving at each end. I typically have a choice of a slightly shorter, very flat freeway, or a bit longer, rather hilly freeway.

I take the car out only occasionally on a weeknight after I get home. On weekends, there can be up to 3 or 4 trips of city driving. Rarely do I go anywhere of distance on a weekend.

Going to or from work along the flat freeway, I'll get the battery guage up to full eventually, unless I'm driving into a fairly strong and consistent headwind. If I take the hilly freeway, I'll typically get the battery guage to full by the time I get to work, but not quite full for the return trip home.

If I've gotten it to full getting home by taking the flat freeway. if I go out for a short trip around the city next, either that night or the following weekend day, that trip will result in a recal of the battery guage within a few miles. If, I were to do this pattern every day, this would happen every day. If, on the other hand, my next trip is back on the flat freeway going to work, no recal happens. If I go back along the hilly freeway, I'll usually see a recal going up the first major hill.

These recals seem to happen only when I get the battery guage to full, and then have to use my Insight over hilly terrain, putting some "strain" on the battery.

OTOH, if I drive solely around the city (which happens a lot in recent months because I've had no job to commute to), I never get the battery guage up to full.

This can't be an age issue; it's been like this every since I purchase my Insight with 17K miles on it.

 

[Will M]

My understanding of how this works is that the system tries to estimage the state of charge by measuring how much amperage has gone into or out of the battery since the last known full charge or full discharge. There is a known maximum and minimum voltage for any given state of charge (SOC). If the system notes that your voltage is higher or lower than it ought to be for your current SOC, then it knows that the current SOC estimate is wrong and it recalibrates.

If yours recalibrates often, then either the SOC is not being accurately calculated, or something unusual about your 144volt electric system is yielding an unusual voltage for a given SOC, or at least one of the sensors is inaccurate. Sensors could either screw up the SOC calculation or falsely trigger a recalibration.

Remember that the guage can't MEASURE your SOC. It can only estimate it through calculation, estimating amps in, amps out, battery efficiency and temperature. It is a very complex guess and recalibration is a normal part of how the system works. Frequent recalibration, however, isn't all that common, so either your driving conditions are somehow different from most people here, or there's some minor malfunction in the system that would probably be quite challenging for a mechanic to hunt down.