TOPIC: Interesting Find_Throttle Flucuations_Negative Trims

I found out something very interesting lately, maybe it's a known thing, I don't know.

What I noticed while driving was:
When I repeatedly flucuate the gas pedal quickly in a pumping action, about 5 times per second (say from a throttle angle of 16-26)
I notice the Short Term Fuel trims go very Negative and stay that way even when going up a hill!!!!!

I've done this several times while watching the the results on my ScanGaugeII.
As long as I keep doing this, the Long Term Fuel Trim has gone from a +5 down to a -3 in about 15min. of driving this way.

Can anyone help me explain what could be causing this?

In my limited understanding I've come up with some theories:

1. Perhaps the throttle plate angle opening & closings are creating bursts of extra Vacuum
further assisting the groove's toroidal waveform like a smoker puffing out smoke rings?

2. Maybe the computer can't respond quick enough to cancel out the groove's effect?


1999 Honda Civic DX
1.6 Ltr
4 speed Automatic

JV so when you do this and LTFT goes neg, does it hold this or go back positive? What do you see at next startup?

Tracy G

that is what its suppose to do, remember anything a computerized system sees will automatically adjust twice as much to compensate,when you gas it the ecu will take away, with the carb type a vacuum leak was lean ecu type will be too rich happy grooving dan

Thanks Dan, as always a good insight to it!

This may give a new technique to save gas, JV, did you do the "flutter" under low load, like flat road, or under more load. This is a whole new area to explore...

I say we need more input on this. JV and anyone else, esp. w/ a scanguage installed- investigate this further to establish trends in results.

Seems to me it's best for road driving instead of in town. We need to see how situation-specific this may be. JV- can ya give more details, how long did you have to "flutter" the throttle, under which load conditions, and how long did it persist. Might this be a new way to teach ECU to cut fuel, even during ECU re-learn after Grooving...? I'm intrigue'd! I can't do it on my cars, so please, let's see if we can get a handle on this one, everyone. We need to see if different cars respond like JV's did. More Input needed!

Tracy G

Sorry for getting back so late, I don't have access to the net much anymore.

Yes, when my LTFT goes negative, it stays there the next time I restart.
The problem is that it takes a lot of inconvenient driving to accomplish this
and it's dangerous to do while driving, that's why I can't keep it up for long.

This happens under ANY load condition!
It takes a little bit of playing around to get it to happen, got to find the right rhythm.
Perhaps my ECU is different than normal?
I don't know. That's all I can really say about it.

OK thanks, JV that's encouraging! Maybe sometime you might try to further detail this fascinating info.
Such as- how much flutter was needed, like how many times did you do it to see the FT's go neg, was it just 2 or 3 times in a row, how long each time, etc. I'd be doing it myself if I could...

I've never heard of this before. Again--anyone w/ a scanguage installed, if you try this please give some documentation on this unusual computer "training trick"

Another question JV- since FT's went neg., did you have any resulting driveability/power issues resulting? thanks brother!

Tracy G

Ok, The fluttering has an immediate affect on the STFT(Short Term Fuel Trims)!
I'm talking within a few seconds, maybe less.

What I notice is, once the STFT readings go Negative, I keep the same rhythm of fluttering as long as my foot allows me. It's Not very easy or convenient to maintain But, if you keep consistent for a long time, the LTFT will drop as an expected result.

Now the problem I see with this is that you're training the ECU to get better fuel economy under these Very Unnatural conditions. Once you return to normal driving practices, the Fuel Trims go back up immediately. So my feeling is that this is probably just a cheap trick & likely nothing sustainable.

And again I stress, this is NOT safe to do while driving! It requires too much concentration and could lead to a greater risk of getting into an accident!

I forgot to mention earlier, another reason I was getting my LTFT's down quicker was:
as my STFT's spiked to a negative value, I'd immediately turn off the engine to transfer that data into the LTFT. I'd coast down hill with my momentum, turn the engine back on, continue fluttering the gas peddle until I got a deep negative STFT, then turn the engine off immediately.
I'd keep doing that until I felt it was too dangerous or too tired.

As long as I was doing this, the engine felt noticeably more efficient/powerful!
As I mentioned before, I could be doing this up a steep hill with the same results! Load was NO issue at all! In fact it's much safer while going up a hill as it's easier to slow down if you had to.

I don't know what else to say about it. It just takes a fast consistent touch.

thank you very much, JV. That's a goof writeup on this unusual technique.

You're right about it's potential dangers, this isn't for just anyone to try. Requires concentration, patience, and a Scanguage in an OBDII vehicle plus the right road situation (like little/no traffic.) My test route between Fernley and Fallon NV could be a place to try it, though it's very flat.

I'm still intrigued with it. It's interesting what you say on driveability, how it feels better...

big challenge is- how to have the ECU maintain the neg. trims, for better efficiency and better power/response...? Makes me wonder if those who make replacement chips/plugin tuners/reflash computers, if they know of this trick. Wonder if it can be simulated thru data stream manipulation, that's wayy out of my ballpark.

I also wonder- can this work on any OBDII vwhicle, or did the Groove enhance it's effects? HHMMM...

Lastly I wonder- about incorporating this technique during a computer re-learn procedure after Grooving? You say FT's reverse back to normal as soon as normal driving is resumed. Well I think I'll try it next time I do a re-learn anyway.

thanks to you JV, this is tantalizing!

Tracy G

Yea the engine feels good while applying this trick however, you notice a slight surge obviously as you do this, you flutter fast but very subtle. If you go too extreme with it, it doesn't seem to work. There must be a sweet-spot, perhaps it has something to do with harmonics?

I also wondered whether or not 'mod chips' could emulate this signal to get MPG gains.

Anyways,
After thinking more about this, and with Dan's input on the matter,
the rapid change might be throwing more carbon than usual at the O2 Sensors, triggering a lean command from the ECU. If this is the case, then perhaps this trick might lead to shorter sensor life. Not good.

It may be another trick to add to 'Hyper Miling' for those obsessed with getting all the MPG they can get, regardless of the cost.

Makes perfect sense to me.

ECU is obviously delivering fuel based on max throttle plate angle; engine vacuum is drawing more air past the MAF.

Modulating the throttle like that is modulating the air flow.

The wider the throttle plate opening, the more air; the more air, the more fuel. The more unburnt fuel at the o2 sensors, the harder the fuel trim.

My guess: definitely the Groove would enhance the effects; they'd be some big puffballs of air rolling down the intake manifold along with a bunch of smaller ones, like the ones Ron talked about in front of the whiteboard in that video. vortexes, vortices...
As in another air handling discussion we've had recently, Tracy...the trick is to keep the air moving, it seems.

just a casual observation, not exactly on topic, but-

I saw some info recently, don't remember the source- but what the computer is doing is measuring air input upstream of the combustion chamber, and then the results of the combustion (O2 sensors.)
It adds fuel based on combo of this info. There are usually 3 inputs regarding the air- IAT (temp.)
TPS (throttle angle) and then either MAF or MAP.

In speaking with another G-Man recently, who we haven't heard from here in a long time- he said it can be good to re-locate the MAP sensor's vacuum connection to further away from the throttle body, to lessen the Groove's effect upon it. There's also the Mapster. IAT can be addressed w/ resistors, or even just attenuating it's input by partially shielding it from the input airstream.

So there are various methods to "fool" the ECU, and I'm intrigue'd still about the throttle flutter, it would be great to be able to shift FT's on a longer-term basis.

Tracy G

I'd be interested in what priority/order sensor input the ECU makes its fuel delivery calculations based on...This is good. The logic tree is coming together.

MAF/IAT (rate of flow divided by density?) would determine MASS of air prior to throttle plate, TPS tells computer how much of that is sneaking past the throttle plate for fuel delivery based on fuel map, all reckoned by o2 sensors post burn/bang/push.

Here's a webpage I found when investigating fuel trim: Fuel Trim Info - Ross-Tech Wiki
I'm assuming injector pulse duration in milliseconds is because fuel rails operate at (constant?) pressure, and fuel trims (additive (to fuel map??) for long trim, multiplicative for short trim) are in % of standard duration.