Air Temperature Sensor - Improving Engine Performance?

[boxercup]

Greetings,

This thread's time has finally come full circle derived from many posts in many related tuning threads.

Ben has often suggested this mod and I assure you that it in theory it works!

Let's see how it works for our BMWs!

Looks like there are some very interested able techies here:

You Know Who You Are And Your Talents Are Appreciated!

Can it tune out the flat-spot in Andrew's BMW? Yes in theory it can.

What will it cost at other RPMs - I am sure that we will see?

Remember, Race bikes are tuned for every race track and they run through 7 - 10 maps to make a full lap!

On the track it is about lap times.

It is different on the road with our BMWs; it is about the enthusiast being satisfied with the way they perceive that their bikes are running.

The Levels of Self - Fulfillment* are key here:

*Self - Fulfillment

-noun
Fulfillment of one's aspirations, hopes, etc. through one's own efforts

------
---

So for the techies:

Ben wanted to know about Ignition and the effect a resistor has on it.

Whenever you change the mixture on a Motronic and the BMS-K using the IAT port, the ECU will call up mapping that has Injection and Ignition values that are always Co-Related values. The converse is true.

Yes, if richer mapping is called up it will have the related enhanced timing value. (Easy to check!)

All On-Board mapping is co-related when the IAT port is used.

On-Board dynamic mapping is always better than any tuning device that employs static mapping. (Piggy Back Tuners)

------
---

Oh!

Andrew (Did He have The Beer Yet) Thanks!

BTW, another tidbit is the delay of the IAT effect. It is much quicker than reported above!

Andrew can proof that!

Let's see how this goes!

[Boybiker3]

Originally Posted by AndrewA

Ben, it has been reported that the ECU can take 30 seconds (or longer) to register changes at the Temp Sensor, so having a pot that can be adjustable on the fly wont really help (particularly in the early going).
snip

1. Any time you shave you get back triple in improved subjective measurement. Lord knows there's enough bad subjective measurement.

2. Betcha nobody knows what the update period is.

3. Even if the update is around 30 seconds, the average delay is 15 seconds and any time you hit under a few seconds, your butt will know (or won't know).

Why not just do it the way I suggested since there are no "costs" to my way and there are no advantages to the way you want to do it? But, your choice.

[Boybiker3]

Originally Posted by boxercup

snip

Ben wanted to know about Ignition and the effect a resistor has on it.

Whenever you change the mixture on a Motronic and the BMS-K using the IAT port, the ECU will call up mapping that has Injection and Ignition values that are always Co-Related values. The converse is true.

snip

Gosh, I may not be using the word "map" correctly, but golly, I thought the Motronic 2.4 had only six or eight maps corresponding to the color cat code plugs used in different parts of the world and that is that. OK.. there's limp-home mode, and you can call that a map too, I guess.

Funny, it is also my perhaps badly mistaken notion that ambient air temperature is handled as a fairly minor (and fairly cold) trim factor. Maybe I am reading Brad Black all wrong. Can happen to me, I suppose. Maybe I should really call "trim factor" a "map" too.

But do correct me if there really are all kinds more maps in there that I don't know about.

[boxercup]

Originally Posted by Boybiker3

Gosh, I may not be using the word "map" correctly, but golly, I thought the Motronic 2.4 had only six or eight maps corresponding to the color cat code plugs used in different parts of the world and that is that. OK.. there's limp-home mode, and you can call that a map too, I guess.

Yes they are all On-Board maps! The 2.4 uses a cat "Jumper" that selects the resident map. As we all know, Model variants of the 2.2 Motronic with air filter and catless performance exhaust work well for some by pulling the o2 sensor and engaging the default mapping.

Funny, it is also my perhaps badly mistaken notion that ambient air temperature is handled as a fairly minor (and fairly cold) trim factor. Maybe I am reading Brad Black all wrong. Can happen to me, I suppose. Maybe I should really call "trim factor" a "map" too.

Trim variants will call up specific mapping. Mapping is not trim!

But do correct me if there really are all kinds more maps in there that I don't know about.

Seems like you already know that there are a few maps in the Motronic.

Give yourself credit!

The BMS-K does not use a jumper to select mapping. It is the preloaded model specific firmware loaded into the BMS-K by BMW.

The BMs-K ECU is quite the data-logger!

Best!

[OwenM]

The temp sensor on my bike reads about 7k ohms at 20 degrees C and 17k ohms at 0 degrees C. The thermistor is very non linear and I expect something like 40k ohms or more will be needed to simulate -25C.
I have just replaced my 10k pot with a 100k pot and I have marked temperature points for reference. The pot is mounted in the fairing for easy access while on the move and has a 6.8k fixed resistor in series so that cant set a temperature higher than 20C. This setup should give me enough adjustment to take the ECU as rich as it will go and with maximum advance just so I can see what it does.
The temp sensor is not in circuit at the moment, I’ll connect it when I work out what is needed.

It should be easy to see the responses time to changes in input with the GS-911 as it gives real time feedback from the ECU.

Since my RT does not run lean enough to cause significant loss of power so I expect most of the improvements I get will be due to more advanced timing, which in my experience can make very obvious improvements in torque and throttle response, it may be worth going a little too rich in order to get that. When cruising (closed loop) the ECU should tune out the richness.

[boxercup]

There will always be things that are beyond our grasp, that are impossible to explore. But the fundamental laws, I believe are knowable and finite. And the coming years in tuning could be the most exciting of all as we explore new tuning methodology coming to us with the new generations of ECUs with User Switchable and Floating Automatic Mapping <>Multi Maps, Traction Control, Electric OH Valves & Electronic Cam Timing adjust-ability and other emerging technologies.

The ECU is now a really a MCU!

We are not at the end, but at the beginning of a new era in tuning.

But whatever we find, there will always be new horizons continually awaiting and challenging us!

-Thanks to Michio Kaku!

[throttlemeister]

Originally Posted by Boybiker3

1. Any time you shave you get back triple in improved subjective measurement. Lord knows there's enough bad subjective measurement.

Well, that's a great one. The king (or idiot) of subjective measurements, false assumptions and twisted words complaining about subjective measurements and semantics.

[Boybiker3]

Originally Posted by OwenM

snip
This setup should give me enough adjustment to take the ECU as rich as it will go and with maximum advance just so I can see what it does.
snip

How do you note the spark advance?

Goggle brings up at least two pdf Bosch ambient air temp spec sheets on page one. Any way for you to link us to the R1100 and/or R1200 ones - they might be different? I wouldn't ever doubt your resistance measurements, but those values seem high for a circuit design in an automotive environment - just my intuition.

Originally Posted by throttlemeister

Well, that's a great one. The king (or idiot) of subjective measurements, false assumptions and twisted words complaining about subjective measurements and semantics.

Like to give us an example or two?

[OwenM]

Originally Posted by Boybiker3

How do you note the spark advance?

Without a dyno, collecting timing data is not really practical. Who cares what it is anyway, the proof of the pudding is in the eating.

Originally Posted by Boybiker3

Goggle brings up at least two pdf Bosch ambient air temp spec sheets on page one.

Why should we care about the specs of the sensor, we are overriding it anyway.

[roger albert]

Quote:

Originally Posted by throttlemeister View Post
Well, that's a great one. The king (or idiot) of subjective measurements, false assumptions and twisted words complaining about subjective measurements and semantics.

Yes, funny, eh? I see he's finding some of his new words from the google now.
It's a great place to parrot words from. Try some wikis too ben.

[boxercup]

Originally Posted by Boybiker3

How do you note the spark advance?

Ben, you know how!

Goggle brings up at least two pdf Bosch ambient air temp spec sheets on page one. Any way for you to link us to the R1100 and/or R1200 ones - they might be different? I wouldn't ever doubt your resistance measurements, but those values seem high for a circuit design in an automotive environment - just my intuition.



Like to give us an example or two?

Ben,

You are part of the team. Don't leave yourself open for criticism.

Be real - Reverse the trend or stand by the side and watch!

[phoenixtexoma]

Robert: You mentioned variable cam timing. I recently helped overhaul an old Cushman Highlander for a friend which uses a centrifugal clutch and had an idea. Why couldn't that concept be used to create variable cam timing, say from +9 to -6 depending on RPMs? I'm a fan of Ockham's Razor -- the simpler, the better. Just a weird thought.

[boxercup]

Originally Posted by OwenM

It should be easy to see the responses time to changes in input with the GS-911 as it gives real time feedback from the ECU.

Yes!

Since my RT does not run lean enough to cause significant loss of power

It runs lean in the lower RPMS

so I expect most of the improvements I get will be due to more advanced timing, which in my experience can make very obvious improvements in torque and throttle response, it may be worth going a little too rich in order to get that.

Too much fuel will reduce HP gained by more aggressive advance.

There is a fine line that you will reach in tuning that when crossed the HP will go down!

Balance is key!

When cruising (closed loop) the ECU should tune out the richness.

Will it?

However, let say it does! - It will not counter the long term effect of a rich mixture that offers fouled plugs and poor combustion efficiency.

Go for it!

[cageyar]

Originally Posted by phoenixtexoma

Robert: You mentioned variable cam timing. I recently helped overhaul an old Cushman Highlander for a friend which uses a centrifugal clutch and had an idea. Why couldn't that concept be used to create variable cam timing, say from +9 to -6 depending on RPMs? I'm a fan of Ockham's Razor -- the simpler, the better. Just a weird thought.

Honda already beat you to the punch-they developed the origins of variable cam timing back in the 80's, first on motorcycles which they often use as a test bed for technology, then automobiles became the primary application: VTEC.

Honda Worldwide | Technology Picture Book | VTEC

[OwenM]

Adding fuel to my standard RT made no significant deference, so it was obviously not running excessively lean.

I have no intention of going colder with the air temp spoof than absolutely necessary, if it’s too rich I will know. Obviously balance is the key, but that’s not difficult to establish, it’s only an engine not rocket science.

If the feedback loop does not correct the A/F ratio during cruise conditions it’s not working real well. Its conceivable that mixture could be so rich that the feedback system is unable to correct it but that’s hardly likely under the circumstances.

[boxercup]

Originally Posted by phoenixtexoma

Robert: You mentioned variable cam timing. I recently helped overhaul an old Cushman Highlander for a friend which uses a centrifugal clutch and had an idea. Why couldn't that concept be used to create variable cam timing, say from +9 to -6 depending on RPMs? I'm a fan of Ockham's Razor -- the simpler, the better. Just a weird thought.

Good thought phoenixtexoma!

However, in working with the R Series <> 850 - 1100 - 1150 - and the 1200 <> There is not enough space and the technology would add excess weight to the rotating mass of the cams.

We can't recreate the egg here!

------
---

Very interesting that you mention, "Occam's razor"!

To a degree, we all employ that principal. It works!

"Less is more" and "Just the Facts" come to mind.

I live by that principal and it's derivatives.

We should all look at deeply and use this principal in this thread!

Nice Phoenix!

[jduke]

It just seems a waste of talent for one of you guys not to at least TRY an FRK at the same time you're out doing a dozen runs with the resistors.
Come on, please, one of you guys take Robert up on his return policy and test his solution to what you guys have worked on and let the rest of us know what you think.
You've done all this work, we've kept up with your posts, don't stop this close to the finish line.

[OwenM]

If others want to do that they are free to, but I will not be spending any time on such an exercise.
If people have a problem with the typical boxer mid range flat spot I suggest they install a foam air filter like the Unifilter, I have used them on the R12S and RT and they smoothed out the mid range on both, I suspect its due to a damping effect on resonances in the intake tract.

[Bob Hancock]

Originally Posted by OwenM

If people have a problem with the typical boxer mid range flat spot I suggest they install a foam air filter like the Unifilter, I have used them on the R12S and RT and they smoothed out the mid range on both, I suspect its due to a damping effect on resonances in the intake tract.

I tried that on two bikes...didn't work.

[OwenM]

Didn't work at all or didn't work enough?
The Unifilter was a definite improvement over the BMC filter I was using on the R12S, but the RT never had a significant problem with the standard filter to start with so the Unifilter made little differences.