Tested: The Most Efficient Way To Accelerate With An Automatic (Toyota Matrix)

[Arragonis]

Originally Posted by mechman600

Are they not the same? I cannot find any published information about display vs data sampling rates.

I assumed by your original argument that the data sampling rate also must be quite slow if it is inaccurate during fast acceleration.

The sampling rate will be much faster than the display. The process will be something like :

counter = 1
:Loop
Take Snapshot
Update Totals
Increase Counter
if Count > Interval Setting then
update display
reset counter = 1
Goto :Loop

The SG2 is good as an instant readout and an estimator of AVG for every day use - I love mine! But not sure it is accurate enough to draw a conclusion here. See MetroMPG's quote from LL earlier.

Best test would be top the tank up, do some slow runs, top it again - repeat a few times. Then repeat it using faster acceleration and compare the amount of fuel used. It would need a large, flat area, a can of fuel, a measuring jug and a some time and patience. It would be very valuable to all though.

I'd give it a go but I can't get Hermann to "brim".

[PaleMelanesian]

In my experience, with a manual transmission, the best approach is moderate-to-high throttle and LOW rpm. As low as you can manage. I shift at points where I land in the next gear at 1300-1500 rpm. That means I'm shifting at or below 2000 rpm.

The trouble is an automatic won't allow that combination. It holds a lower gear to higher rpm instead of allowing such open throttle. I believe the test results here, for an automatic, but my experience doesn't agree for a manual.

[MetroMPG]

Originally Posted by PaleMelanesian

The trouble is an automatic won't allow that combination. It holds a lower gear to higher rpm instead of allowing such open throttle.

Depends on the automatic, of course!

Just today I drove an older Audi A4 3.0L automatic with a "manual mode" that would allow me to short shift into top gear at almost half the road speed compared to the programming in automatic mode.

And it would hold that lower RPM under high engine load, unless I literally floored the accelerator to activate a downshift.

Updated this post: http://ecomodder.com/forum/showthrea...ing-21149.html

[Arragonis]

I do this all the time with Hermann but he tends to upshift quickly on a light throttle anyway so it makes maybe 2-300 rpm difference in most gears. He will hold a gear for about 5-10% speed under the up-shift point except on hills when he wants to downshift all the time and I have to prod him back up again.

Possibly going to look at whether the induction system is gummed up, but that is for another thread...

[E.Roy]

mechman, when you get a chance, can you please read your TPS value on SG2 for zero throttle?

My 2000 Corolla 4spd shows TPS value of 9 on SG2. I have a theory about adjusting my TPS to lower this value, hoping that the ECU/TCU will upshift earlier at higher throttle openings, or that it will shift at the min speed required despite having a higher throttle opening.

I had a '90 Camry that I left the TPS disconnected and the ECU shifted the car extremely early.

Edit: It appears my Corolla has no adjustment for the TPS

[JRMichler]

Originally Posted by E.Roy

I had a '90 Camry that I left the TPS disconnected and the ECU shifted the car extremely early.

Sounds like an easy way to add an early upshift switch.

[mechman600]

Originally Posted by E.Roy

mechman, when you get a chance, can you please read your TPS value on SG2 for zero throttle?

TPS = 15% @ idle
TPS = 80% @ WOT

LOD = 8-13% @ DFCO (depending on RPM)
LOD = 80% @ WOT (any RPM)

I have a theory about adjusting my TPS to lower this value, hoping that the ECU/TCU will upshift earlier at higher throttle openings, or that it will shift at the min speed required despite having a higher throttle opening.

In theory all you would need to do is create a small (metered) vacuum leak. Your throttle would be more closed at the same load to maintain the same MAF/o2 sensor readings. Not sure what the trans goes by as far as downshift logic goes, but if it goes by throttle plate angle, that would do it. If it goes by algorithms based off multiple sensor readings, it would't work.

[E.Roy]

^^^ The small vacuum leak would raise your idle though, which should consume more fuel. Thanks for posting your TPS and LOD values.

I should clarify when I said I left the TPS connector off on the '90 Camry it was by accident, and the car would shift absurdly early, but it did let me realize the ECU relied on TPS voltage to control shift points.

[euromodder]

Originally Posted by PaleMelanesian

The trouble is an automatic won't allow that combination. It holds a lower gear to higher rpm instead of allowing such open throttle.

And if you briefly lift the throttle , get it to shift up, then get on the gas again (staying short of kick down) ?

Friend of mine is driving his 3L BMW X3 diesel with 8 speed ZF auto to 5 L/100km , that's 47mpg, taking it easy but not really hypermiling it .

The ZF shifts up very quickly if you don't really floor it - and with a 3L TD engine with heaps of torque, low rpm is no issue at all.



Did you go along on the Audi X-country run with CleanMPG BTW ?

[PaleMelanesian]

Originally Posted by euromodder

And if you briefly lift the throttle , get it to shift up, then get on the gas again (staying short of kick down) ?

Indeed. That's a useful technique. I especially use that right near the top gear lockup point, 45 mph in my Odyssey.

Friend of mine is driving his 3L BMW X3 diesel with 8 speed ZF auto to 5 L/100km , that's 47mpg, taking it easy but not really hypermiling it .

The ZF shifts up very quickly if you don't really floor it - and with a 3L TD engine with heaps of torque, low rpm is no issue at all.

Did you go along on the Audi X-country run with CleanMPG BTW ?

Wanted to, but couldn't for family medical reasons. At least I got some good sleep instead.