Desertdog (Carl) has discovered a new technique for the FEH/MMH which he calls the "Opposite of the Fake Shift". What ever the name, this is an important find for hypermilers such as myself. CarlD first posted this on CleanMPG.com here:
http://www.cleanmpg.com/forums/showthread.php?t=17097

Carl is our expert on the SGII and many other subjects and I'm very proud of his knowledgable post.

GaryG

 

I've noticed this for a long time.
And I've described it long before the SG had SOC ability.

Pressing the accelerator hard, even for a spilt second, takes the car out of EV. You all know this.

When the car is in motion, the car can use the forward momentum to in effect, "push start" the engine. When the car is moving slowly, or is stopped, the car must use battery power to spin the generator/starter motor to start the engine.

Using the car's momentum to start the car, obviously will use little, if any additional battery power.

From what I'm reading, if the SOC hits, or is below 40.0 %, the car chooses for some reason to use battery power to start the engine.

A) we need to find out if this is always true, or only true in certain cases, and
B) what are the gains of forcing the ICE on before SOC 40% and
C) why

I'll start looking more closely at this effect.
I've ingnored it for the most part over my nearly 4 years of ownership and brushed it off as insignificat to MPG. But then again, I've only had a SG with SOC for less than 1 year.
-John

 

John, it took me some getting use too, but I'm getting better the more I use this technique. You don't press the accelerator hard, so don't confuse the members here. It's a quick small blip that causes the restart and it's very light. If you save .3 - .4% SoC every time you have a restart, it all adds to your EV time and MPG savings after just a few times. In fact, I'm getting angry with myself when I let the engine restart because of a low SoC now. This is the best technique for saving fuel I've seen in a few years now. I may have some ideas to even improve it in the '09, but I still need more testing.

GaryG

 

Today, I had great results using the Reverse Fake Shift almost 100% of the time. It appears you must be in the decel mode for the blip on the accelerator to work like a charm. Also, the folks that have the '09 FEH/MMH can also go 100% Fuel-Cut (DFSO) when the SoC gets down to as low as 40% by shifting to "L". This will also stop the battery SoC drop and begin regen right away. Remember, the '09 automatically starts the engine anytime you shift to "L", so I just verified no loss in SoC on a restart using this technique.

Using this new technique today, I took a beginning Nav Sys Tank reading with 412 miles from 50.5mpg to 51.1mpg ending with 452 miles. If we are correct that the Nav Sys reads 4% high, this would give my new '09 FEH a 49mpg tank with pumped gas using E10.

GaryG

 

As everybody should know, the ICE can't be started solely using vehicle momentum. This is basic physics of the planetary gearset. During EV the generator is free wheeling and to spin the ICE requires counter torque from the generator. The "blip" method seems to minimize the total amount of electrical power needed to spin the ICE up. It takes some practice, as too deep or too long of a blip doesn't result in the desired no/minimal SoC loss. Typically when done correctly I save about 0.2 to 0.3% SoC for every EV-charge cycle, which is good for a few percent MPG gain. It can also be used to turn the ICE on during a P&G using DGE which normally works best when you have SoC of 53%.

The SoC drop from ICE starting is noticeably more if you've had the battery harness work done. If you have your foot on the pedal when SoC drops below 40%, the Soc can drop as low as 39.3% or so from the ICE spin-up. If you are coasting in D, the SoC drop is usually only about .3% or so depending on vehicle speed.

The blip is a minor technique but done well it does have measurable benefits.

 

As not too many people know, but DesertDog should know, there is a device called a "generator brake".

If the generator brake is applied, the engine can be spun by the car's forward motion, a mechanical force vs. electrical force. Indeed, some form of energy must be used to apply such a generator brake, but presumably, this method is very efficient.

Now, I'm saying this can be done.
I'm saying it is very likely this is being done.
I've yet to find positive proof it is being done.

I'd be very interested to hear the other theories of why this "reverse fake shift" makes a difference.... not just that it does. Thank you.

 

My thoughts were, during regen in "D" or "L" the traction motor (MG2) provides voltage with no conversion losses directly to the generator to spin the motor for a start-up. There is no need to drain the battery for this voltage. The PCM then provides spark and fuel injection for engine acceleration. While in EV, the '09 spins in fuel-cut when you shift to "L" while in decel. The '09 is much different in that it gets a run-up in RPM (2,000 or higher) every time you decel in "L". Earlier models had a run-up only when the battery would not except a charge during a decel in "L".

The quick blip during decel in "D" causes the PCM to start-up the engine and since MG2 is providing regen power, that power is use to spin MG1 in the start mode. Carl is exactly right regarging the larger battery drain during a start-up while accelerating. I've been getting these SoC readings in the 39.3% range and thought the '09 was programmed to just drop the SoC further. The reason I was constantly seeing these big drains was I was using cruise control in EV so much. Cruise control causes slight acceleration during a start-up to maintain speed and this was the problem. I still use CC in EV, but when the SoC gets down to ~40.8% I shift to "N" and coast till I get down near 40% (traffic permitting) and pop it in "D" for the quick blip in decel. Works like a charm and my MPG average is shooting up now. Shifting to "N" cancels CC for those that didn't know.

Since the '09 has 100% fuel-cut in "L", and we now know the restart doesn't drain the battery by a shifting to "L", this is a good technique for stoplights. I've been using this for weeks and didn't understand how the battery was charging so fast. It was because there was no drain on the battery at the restart.

GaryG

 

Generator speed/torque control is used for virtually all ICE starts out of EV. The generator brake GPSman refers to is used primarily to fix the final drive ratio of the eCVT during steady state cruise and would never be engaged when the generator is spinning @5000 RPM or so. It would be next to impossible to attain acceptable NVH standards if the vehicle's momentum was used to start the ICE in this manner. If you want proof of this for yourself, code up the generator torque xgauge and total desired torque xgauge and watch what happens during a normal ICE restart. What the blip method does is to very briefly trigger an ICE restart due to torque demanded and then the demanded torque immediately goes back to zero.
That is why the blip works, as the vehicle does not have to maintain the desired speed/acceleration while spinning the ICE with the generator. In general, providing a smoother ICE start out of EV requires more torque from the generator (as well as precise speed control) and hence more battery power. A very brief spike in desired torque minimizes this and saves a small amount of battery power.


Posts #2 and 6 in this thread are why I posted this on CMPG first and not here. GPSman did have a nice CYA disclaimer in post #6, though.

 

Post 2 and 6 where mild compared to others in the past, but was one reason I stopped posting for 9 months here. I think you know how I feel about new techniques and my desire in passing the information on too everyone with the FEH/MMH.

While this technique only saves a small amount of battery power, it's increasing my mileage average a great deal. I only expected to increase my 7th tank from 48.06mpg to about 48.5mpg, but this technique has brought my average up to 49mpg in only 50 miles and I may hit a 50mpg tank in the next 200 miles.

Your discovery is big time to me and I can't wait to start a new tank using it throughout for a final tank MPG average. Our '09 FEH/MMH operates in the 40% to 42% SoC range in the city and when the SoC drops to 39.3%, it really hurts MPG.

I have not disagreed with anything you've posted and enjoy reading them. Sorry I posted this information here, but I ask you to spread the word on CleanMPG.com first. I was just too excited when I saw the results.

GaryG

 

All good info, both here, there, and elsewhere.
Just remember, there is not only one way to do things...
Especially, with this car!


( and it's good for people to post the differences between older and newer model years - not too many people own one of each )

thanks Gary!