SOLVED- Unexpected ICE Turns On

[Crota]

Thank you for all the hard work here. It is nice to know why my car starts the ICE when I've driven less than 1 mile and did not apply more than what would be 3,000 rpm of "power" from the accelerator. Early on reviewers said this car is a car designed by Engineers for Engineers. I have to agree with that description.

 

[bfd]

You might also search Toyota patents for more on this…

 

[insightman]

You might also search Toyota patents for more on this…

I'm sure Honda's engineers were charged with the very difficult task of avoiding hybrid car technology covered by any of Toyota's patents.

 

[AnthonyW]

Well now I have a wrinkle in this whole theory. Before I go into that, I have to define what you are looking at. See the picture at the bottom of the post. That is a Scangauge test unit from Scangauge that I am working on with them to crack the Clarity's secret PID's. Here are what the values represent:

MXS - Behind the scenes State of Charge measurement of the battery which I believe is the true SOC.
HBV - voltage of entire battery pack.
STC - Regular SOC that you see on your Hondalink app. Just has a decimal here.
DCT - Temp of the DC to DC converter.

See this grid to see to see how they correlate:

*Since our battery is a 84s2p, I divided HBV by 84 to get the voltage per cell.

I have a free ChargePoint charger within walking distance of my work. When I picked it up the MXS was 95.9 and HBV was 342. I only charge to full twice a week. Other days I try to keep the MXS between 90% to 40% (92% to 35% on HondaLink). When I charge to full, MXS is always 96% but last week it was 95.8% and, today it was 95.9%. These are true OCV measurements because once the charge was all done the car sat for about 2 hours before I picked it up (that is what is pictured). It is about 0.3 miles to drive it back to work. No hills or declines, just a parking lot, an intersection and my work parking lot. Backed into a parking space, put it in park but left the car on to listen to Sirius while I put up my front window shade. All of a sudden the ICE kicked on. Scared me half to death. Because of the way I charge, I have not had an unexpected ICE turn on since the first month I purchased the car (April). I look at the Energy Flow on the display and see the green power bar flowing from the engine to the battery. I think to myself, Honda should fix this, the green bar should be going to the ICE not from.

Then I look at my Scangauge and that is what blows my mind. Driving the 0.3 miles had reduced the MXS to 95.5%, STC/HL to 99.2% and HBV stayed at 342. But when the ICE kicked on the values started to increase. It ran for about a minute until the readings of 96%/100%/343 as shown in the pic below. Once MXS hit 96% the engine shut off. The HBC went up to 343 which I had never seen before. I looked and say the "Ready" indicator was still on so the car was still on.

Has anyone had the ICE come on after a full charge but not moving. While I still believe in this post's explanation of unexpected ICE turn ons, there must be some other reasons the ICE kicks on. Remember I was in Park and there was no regenerative braking or excess energy being created at the time. I can think of 2 possible reasons:

1-the car was recalibrating both the state of charge percentages
and/or
2-the car was doing some active cell balancing.

We don't know exactly when and how the car balances the cells. Our best guess so far was that it happens at the top end of the charging cycle just before the charge turns off because the battery is full. What I experienced seems to indicate we have other things going on. Welcome your thoughts.

 

[Emanuel Green]

For those who are obsessed with the ice not coming on, why not just get an ev. In the big picture it's beneficial for the ice to run periodically. If not the oil will oxidize in engine before it looses viscosity. Also gas will go stale in the tank. Like it or not the ice compliments battery longevity by sharing the load and reducing the discharge rate of the smaller ev battery used in a phev especially when driving more aggressively. I personally like the way the my car performs and I use the different drive modes regularly.

Many people are concerned about the range of current EVs. Especially for individuals and one-car families, they need (or at least want) a car that can make out of town trips.

Personal anecdote: at the time I bought my Clarity, the only EVs available in my state that could go over 100 miles on a charge were the Chevy Bolt and the various Teslas. The latter cost upwards of $50k, so that was out of my price range.

I did consider a Bolt, but decided I liked the Clarity better as a car. And I think it was a good decision: the battery is big enough for 95% of my driving, and I have no worries at all about taking longer trips. With the Bolt, even though the battery is quite sizeable, I wouldn't be able to go more than 100 or so miles from home and still have range to return. Public charging infrastructure is simply not at a point where it's viable yet. (There is one CCS charger in my city, and it's been broken for almost a year now.)

 

[insightman]

What I experienced seems to indicate we have other things going on. Welcome your thoughts.

Clearly, Honda's been using their cellular connection to monitor your meddling with their PID's. Somewhere, a bunch of Honda coders giggled as one of them responded by punching in a code to activate the engine in your stationary Clarity and exclaiming, "Watch this!"

 

[Carro con enchufe]

I think the title of this thread should now say “still unsolved”

 

[M.M.]

Fantastic detective work and probably as close as we'll get to a solid answer unless a Honda tech spells out the algorithm. This has been my one of my two primary guesses since discussing it here a few months back and talking with some colleagues, but it was pure speculation.

I'm wondering how the audible click and stop in regenerative braking you can feel at 14mph fits in, though (still can't believe there's no hysteresis on that). If the car was using the mechanical transmission for braking I figured it was due to the gear ratio, but if it's the starter that's providing the resistance there's no reason it would be speed dependent.

Peripheral question, as I haven't been on here for a while: Has someone sussed out the logic the car uses to start the ICE on a mostly-full battery for this energy dump? I've been able to reliably reproduce the behavior all summer on the slight hill near my house if I try, and have probably 75% of the logic worked out, but it's driving me nuts that I can't quite pin it down definitively and my shame has kept me away.

 

[M.M.]

Has anyone had the ICE come on after a full charge but not moving. While I still believe in this post's explanation of unexpected ICE turn ons, there must be some other reasons the ICE kicks on. Remember I was in Park and there was no regenerative braking or excess energy being created at the time. I can think of 2 possible reasons:

1-the car was recalibrating both the state of charge percentages
and/or
2-the car was doing some active cell balancing.View attachment 2491

I've had mine come on multiple times while sitting in park either listening to music or fiddling with settings. I've seen it with a full charge, as you did, partial charge more than once, and empty battery.

The empty battery case makes sense, of course--I tested and confirmed that the ICE will cycle on when the car hits about 1% SOC and off at about 4% SOC. But that does not explain why it would turn on when parked and "idling" for an extended period like we've observed, either with full or partial charge. My experience seems to rule out both recalibration and cell balancing, since it also happens (in similar situations) when well below 100%.

 

[insightman]

but if it's the starter that's providing the resistance

The traction motor, not the starter motor/generator, provides the regenerative braking resistance--it's the motor that's connected to the wheels.

Has someone sussed out the logic the car uses to start the ICE on a mostly-full battery for this energy dump?

You're in the right thread to see the earlier speculation about this conundrum. My theory is that the Clarity PHEV's starter motor/generator reverses the normal direction of its torque while operating as a motor to resist the running engine during regenerative braking when the battery is full. The Accord Hybrid and new Insight simply have the starter motor/generator spin the shut-off engine to use up the excess electricity that can't go to the fully charged battery, but Honda may be reticent about simply spinning the Clarity's engine without warming up the oil. Especially in cold weather.

 

[M.M.]

The traction motor, not the starter motor/generator, provides the regenerative braking resistance--it's the motor that's connected to the wheels.

Am I confused about the way the electric motors are arranged, or did I just word my comment badly?

The way the patent is drawn, MG1 is connected to the ICE and MG2 is connected to the wheels, which would make MG1 the ICE starter/generator and MG2 the drive/regen motor per both the third diagram you posted and logic. MG2 provides resistance to the wheels to slow the car, then shunts that energy to MG1, where some of it is in turn used to push against the ICE (same electrical action as when it's acting as a starter, I think; basically doing engine compression braking with two intermediate electric motors). When I said "the starter that's providing resistance" I meant that MG1 is using the ICE compression for additional resistance so it can dump more energy than is otherwise possible through just electric motor inefficiency.

Am I getting something backwards here, or do I have the right idea?

It's an interesting contrast to what the Volt does, where its drivetrain allows it to essentially pit the generator against the traction motor to avoid starting the ICE, but the Clarity's transmission isn't configured for that if it would even be allowable patent-wise.

You're in the right thread to see the earlier speculation about this conundrum. My theory is that the Clarity PHEV's starter motor/generator reverses the normal direction of its torque while operating as a motor to resist the running engine during regenerative braking when the battery is full. The Accord Hybrid and new Insight simply have the starter motor/generator spin the shut-off engine to use up the excess electricity that can't go to the fully charged battery, but Honda may be reticent about simply spinning the Clarity's engine without warming up the oil. Especially in cold weather.

Everything you say sounds entirely correct, this patent clearly explains all of the behavior I've observed (except the disconnect below 15mph), and letting the engine run for lubrication sounds like the kind of decision an engineer would reasonably make. The only other way the Clarity could be using the ICE to slow the car would be with the mechanical transmission, but the combination of the fact that it's apparently not geared in a way that would function, and this patent describing the other option--which you managed to discover and were so kind to share--has me more or less 100% convinced.

That said, I was actually asking about the other conundrum--not how the Clarity dumps energy with a full battery, but when it decides to dump energy. It's totally understandable that this mode would start after a long downhill drive on a full battery, but some folks here (I believe KentuckyKen in particular had been trying to figure it out?) had noticed that it can be repeatably kicked in after three or four blocks of driving on a very slight downhill. I'm wondering if someone already sussed out the exact logic for that decision while I wasn't paying attention, or if I should resurrect an old thread or start a new one to dump as much as I've figured out in the past few months of experimenting.

 

[insightman]

same electrical action as when it's acting as a starter, I think; basically doing engine compression braking with two intermediate electric motors

That's what the patent indicates and that's what the i-MMD Accord Hybrid and Insight do, but they never start their engines to dispense with excess electricity.

Am I getting something backwards here, or do I have the right idea?

You have the right idea. I misinterpreted what you meant by resistance.

this patent clearly explains all of the behavior I've observed

The patent never mentions the engine starting or how the starter motor/generator operates after starting the engine. The starter motor/generator twists one way to start the engine, but after starting the engine, is it twisting the same way, using up the excess electricity from the traction motor by trying to speed up the engine? Or does it somehow try to twist the other way, trying to slow down the engine? Is there a newer patent describing the Clarity's operation?

If the battery is fully charged, there's no difference between going down a long incline or pulling on the minus paddle as you come to a stop right after you leave your garage, right?

I'm wondering if someone already sussed out the exact logic for that decision while I wasn't paying attention, or if I should resurrect an old thread or start a new one to dump as much as I've figured out in the past few months of experimenting.

Any and all speculation and experimentation results are welcome here. The often mysterious decisions made by the computers governing the Clarity's engine operation are the most interesting topic of discussion for me (and, unfortunately, for @bobcubsfan, whose engine lights up more often than all the other posters' engines combined)

 

[dalbeck]

This is all very interesting stuff, and in general interesting to see other users experiences so far in comparison to my own. I have to say since owning the vehicle for 2 weeks now and putting just shy of 1400 miles on it, the car performs better than expected. I have averaged 55 miles on battery, and this could be bad on my part but I have always charged the battery to 100% at home with my L2 ChargePoint, and while at work on the L1 charger I get about 70% charge before my trip home. With that being said I have yet to have the ICE surprise me. Actually, the battery surprises me more often when it reads 0 but the car still shows the EV light. I do a good amount of distance driving outside of work and frequent the HV mode so maybe that is why I am having more consistent and expected results with the drive because the ICE runs often enough. Either way, I just let the car do what it needs to, so far everything seems to operate correctly and it been the best drive of any car I have owned to date.

Eyeballing a Model X for possible next vehicle in a few years or something similar to it, but definitely would like to get a full battery car next to compliment the Clarity.

 

[insightman]

Some worriers don't believe Honda has left a large enough buffer at the top end to protect the Clarity's batteries. Others have speculated that battery balancing takes place only when the battery is fully charged. The official word on page 466 of the Owners Manual makes it clear that you should always fully recharge the battery:

To help extend the lifespan of the battery, it is recommended
that you fully charge the battery each time prior to driving.

 

[AnthonyW]

So I got to observe this again yesterday while driving. Just as a reminder I am using MXS which I believe is the true SOC reading. It ranges from 96% to 20%. HBV is the voltage reading. 342 is the voltage of the entire pack when full. Divide that by 84 and you get 4.07 volts per cell.

My battery was full and when I leave my house there is a hill that I have to go down immediately. I apply my brakes and notice that MXS goes up to 96.3 and the voltage goes up to 343. The ICE turns on and immediately the voltage and MXS start to decline slightly to 342 and 95.8 while I am sitting at a stop sign. The green bar that goes directly from the ICE to the battery on the infotainment is going the wrong direction, but the voltage and state of charge are definitely going down. I drive about 3/4 of a mile to a red light and stop and I am down to 94.5. The ICE is now putting charge back into the battery. My guess is that it is giving back to the battery while it is getting up to operating temperature and now that there is capacity available for the charge to go. There are no cars at the intersection so I let my green light turn red and watch the MXS go up a point to 95.5 and then the ICE turns off. I quickly change the gauge to the engine coolant reading and it is 159F.

So my belief is that the folks who pull over and shut their car off after the ICE turns on are not letting the engine get up to temp and not getting back the charge that the engine is giving while it is on.

 

[insightman]

The ICE turns on and immediately the voltage and MXS start to decline slightly to 342 and 95.8 while I am sitting at a stop sign. The green bar that goes directly from the ICE to the battery on the infotainment is going the wrong direction, but the voltage and state of charge are definitely going down.

So what is your conclusion about how the Clarity is using up the excess energy? Do you think Honda believes it would be too confusing to show the green bar going from the battery to the engine?

It's interesting to learn that after depleting the battery, the engine immediately restores the SOC to full. What is the cycle-time for this discharge/recharge? Does it take place over and over without the engine shutting off when going down a long mountain descent? Or does the discharging continue on the descent and when the need for discharging is ended the engine continues to run until it restores the SOC to full?

Also, I got over my "engine anxiety" months ago when I learned the engine wants to run long enough to warm up. So I stopped pulling over and turning off the car the instant the engine icon lights up. However, I've gotten much better at driving carefully to keep the engine quiet.

 

[Viking79]

ScanGauge needs to update their max SoC equation, as it definitely doesn't charge that high, it is only charging to around 85% if what you are telling me is correct for full charge voltage of around 4.05 V, as I have pointed out before you can't take 4.05/4.20 (96.4%) as the relationship isn't linear and has a non zero intercept, meaning an SoC of 0% corresponds to around 2.7 to 3 V.

My point here is that with the car at 100% charge it has plenty of room to accept at least a small amount of charge as the SoC is still only around 85%, but regen braking can be on order of 50 kW, so they might figure DC fast charging at 85% will shorten the battery life or maybe could cause issue if not all cells have finished balancing (although they shouldn't be far off from each other).

I think you are right that the engine is being started to deal with the excess regen energy, especially immediately after charging (maybe more heat in system, slightly higher battery voltages, etc).

Edit: I re-read your post and it is interesting the charge is still going up, maybe simply a bug that it starts the engine? Or maybe it is able to deal with some excess power in extreme regen cases (like you live on top of a mountain), so it starts the engine preemptively?

 

[PHEV Newbie]

Some worriers don't believe Honda has left a large enough buffer at the top end to protect the Clarity's batteries. Others have speculated that battery balancing takes place only when the battery is fully charged. The official word on page 466 of the Owners Manual makes it clear that you should always fully recharge the battery:

To help extend the lifespan of the battery, it is recommended
that you fully charge the battery each time prior to driving.

While that is what Honda recommends, I believe they suggest that so the typical driver avoids the most damaging effect (other than heat) on Li ion batteries and that's the depth of discharge (DoD). The worst you can do is fully charge and then fully discharge (even with buffers at top and bottom). The best is if you can keep it within a narrow range around 50%. I know that's not practical for most drivers but there are some on the forum who only drive 10 miles a day. For them, they should charge to about 70%, drive their 10 miles, and then recharge to 70% again. If that person charges to full on Monday and then drives until depleted on Friday, and then charging to full again, repeatedly, that's the most damaging pattern for his needs. Then again, such a person would put so few miles on their battery that it'll still last a very long time. Study after study regarding battery longevity show that repeatedly wide DoD usage dramatically shortens the life of Li ion batteries. I doubt that the Clarity battery is any different. Otherwise, Honda would rightfully tout the benefits of such a revolutionary battery.

 

[insightman]

My point here is that with the car at 100% charge it has plenty of room to accept at least a small amount of charge...

So did Honda set two limits? One limit for plug-in charging and a second, higher limit for dealing with regenerative braking? If the second limit is safe for the batteries then why wouldn't Honda extend the car's battery range by using that higher limit for plug-in charging, too?

Regardless of how many limits there are, I wish I could learn exactly how the Clarity PHEV actually burns gasoline to burn off excess electricity to implement the limit(s). Is the starter motor/generator running in "reverse" as it works to slow the running engine? What kind of fuel-efficiency hit does the car take when it's using this pseudo engine-braking technique to slow the car going down a mountain?

Implementing a regen recharge limit is so much simpler (and likely more fuel-efficient) in Honda's non-plug-in i-MMD hybrid cars (as described in the patent). These cars never have to worry about warming up their engines so they can simply shut off the engine's fuel injection system and close its valves to provide resistance to the normal, "forward" torque of the starter motor/generator.

 

[bobcubsfan]

Wow, this is amazing. I am applauding from the sidelines.

There is still one baffling thing: , even if the engine is used as a friction load (and even if burning gas happens, and is part of the plan to provide even more “load”):

In this particular case, when triggered in this way, there is no reason whatsoever for the ICE to keep running after the braking stops, and even after thencar stops (as bobscunsfan and others have reported).

I am developing a suspicion that the engineers who designed this fancy braking, and the engineers who programmed the “once ICE starts, keep running it until it warms up” didn’t quite work out the communication. So when the ICE is triggered by braking on a full battery, it invokes the “keep ICE running and burning gas until it warms up” path, even though that’s not necessary here.

[This would be a true glitch]—but 100% speculation .

I have said all along that there is an error in the software. This seems like a case of left-hand right-hand not communicating.