"Proper" intake air temp (IAT) sensor location....

[Brandnew]

I'm having an argument about the proper location of the intake air temp (IAT) sensor with a fellow car fan. I'm sure I'm right but I want to check with you guys to make me feel better (or worse?).

Okay, the car in question has an IAT sensor right at the throttle body. The throttle body on this car is attached right to the intake manifold. Okay, that works fine.

Now, as I've posted previously, there is a kit that allows us to put a Roots supercharger on a new custom intake manifold. Here's the problem. The kit does NOT relocate the IAT sensor. With the SC kit the IAT sensor is left on the throttle body and thus never reads the "SC heated" intake air temperature.

First off, this is bad right?? A lot of the people are having LTFT issues and I think that this sensor location could be (at least part of) the cause. Probable or not??

Second, I'm recommended moving the location of the IAT sensor to the intake manifold. They state that this isn't really necessary since you won't gain anything since the LTFT compensates for it. They also state that putting the IAT sensor in the intake manifold will create more problems becasue the sensor will "heatsoak" from the IM heat and cause false high intake readings.

---I think the IAT sensor is there for a reason; to provide as accurate inputs as possible. Why have it if the ECU has to compensate for it??

---I have another car with the IAT sensor in the intake manifold with no problems so I don't agree with them. What is everyones take on IM mounting of IAT sensors?? Is "heatsoak" a concern or not??


On a slightly related note, there is a discussion that the "tuner" has compensated for this improper mounting location and temperture reading. People are saying that the tuner adds "X" amount of heat to the reading to get the fuel calculations correct. The problem is the "tuner" doesn't measure boost so the can't know when the SC is actually increasing the air temperature.

--Can the ECU even be programmed to do this (it's a newer Motronic system)??? People are saying that this is "probably" done by matrix based tables to modify the temperature. I don't think you can use the ECU maps to modify an input, can you?

I personally don't think the tuner has done anything to the tmperautures. I think they just let the ECU compensate at it's whim (i.e. O2 sensors).

Bascially, I think the kit needs a lot of work to be properly impletmented to this car. Just checking if my way of thinking is correct.

Sorry for the long post. Thanks in advance!!

 

[turbododge]

I have spent a lot of time messing with the IAT location on my twin turbo 340 with a Speedpro system. If you are running a speed density system without a MAS sensor, the location seems to be very critical, and can have a severe effect on the tuning. I had the it right after the throttle body on an aluminum adapter to the manifold. When I would tune the mixture at 70 degrees, and then try to run at 90 degrees ambient, everything would be off again. I put it in many other spots with similar results. We chased the tuning for a long time before Speedpro activated a temp compensation algorythm that I can program. It holds tune at all temps very well now. The problem comes from having the IAT at different locations than from when the original program was written, and no two engines react the same. If you are too far away from the port, the air can pick up more heat than the algorythm says it should. If you are too close, it does not pick up as much. Throw on top of that the fact that colder air will pick up more heat than warmer air, and it gets a mess. I would certainly think that any good system should have a temp compensation that you can program.

 

[Brandnew]

Thanks for the reply. Well that makes me feel better knowing that they can at least do something with IAT input. Now, if they did??

I should have stated I'm running a mass air system, with MAF et al.

 

[Andy330hp]

Don't most MAF sensors have built-in air temperature sensors?

On a true MAF system, my understanding is that it will always measure the correct mass flow rate, so IAT sensors do not do much for fueling compensation of "denser charges".

What they tell you is that, when the intake is really cold, you'll have puddling, and so you need to compensate.

When the intake is really hot, you'll have thinner wall film, and therefore compensate accordingly.

When the intake is really hot, you must run less timing to avoid knock.

I think it is these concerns and not "mass flow rate" that the IAT needs to address. I don't have much experience with MAF, this is from what I've read and what I theorize

 

[Brandnew]

Thanks for the reply.

A MAF will measure the correct mass flow...near the air box. But when you take the air right before the intake and run it through a supercharger you change the mass.

I'm just wondering if (and to what extent) they use the IAT sensor to do some last minute adjustments. And if so, to what extent can this signal be modified. If the air comes in a 85F and hits the intake at 200F then you clearly need to change something to compensate.

The MAF on this car does not have a air temperature sensor.

 

[patprimmer]

Brad

You do not change the mass of the air. You increase the density by pumping the same mass into a smaller volume at a higher pressure. A byproduct of this is to increase the temperature which further increases the pressure, but the mass is the same as it was as it passwd the MAF.

The supercharger (vs NA) will draw more air through the MAF, but it will do that by increasing velocity at that point. The MAF will sense the extra velocity.

Regards
pat

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[ElwoodGreen]

Brandnew,

If you changed the mass of air internally in your supercharger, then you did something nuclear. The supercharger can only change the air DENSITY. The "mass air flow" sensor should do exactly what it says, no matter where it is in the system. It should tell the ECU how much mass of air per unit of time is entering the engine. To do that accurately, the ECU needs to know air velocity, temperature, pressure, and cross-sectional area at the point of measurement.

Manifold air temperatures, as mentioned by Andy330hp, are probably used for other purposes. However, since your original installation is right at the throttle body, that makes a nice machined cross-sectional area for flow rate calculations. I have no experience with your system, so you'll have to find out exactly what your IAT is supposed to measure, and why. Only then will you know where you need to put it in the modified system.

 

[Brandnew]

Thanks!!!

That's answers it. I was thinking along the lines of density not mass. DOH! If you think in terms of mass it has to stay the same.

Thanks again!

 

[Brandnew]

Just as an update. I did a little more research and the IAT's in this instance are used for timing control and knock regulation.

 

[Andy330hp]

you're welcome

 

[schmidtj86]

I would have to agree that the sensor should be in the hotter manifold, because of what you just said, Spark control. The hotter charge will definitely have more tendancy to knock, and the ECU should be reading the actual intake charge temperature to prevent that.

The fuel trim getting messed up may be related, depending on the type of mass air flow sensor. How much change are you seeing? Where is the change being seen? The fuel trim may be changing due to the longer track that the the air goes through from the MAF to the port, making the fuel trim learn during the transient. Could also be that the original calibration was compensating for reverse pulses, and the SC blocks those out. Just guessing though.

 

[automatic2]

The sensation of temperature also contributes to correcting a somewhat nonlinear mass reading.

 

[Brandnew]

Schmidt,

Just to answer your question the fuel trim is off by 15-20%. It's actually running rich (i.e. -16%). Not everyone has the problem but those that have it can't seem to get it to go away.

 

[albertfastsubaru]

The IAT does alter the timing and fuel mixture in most cars. It also can be coordinated with the engine temperature sensor by the computer to determine if this is a "cold start" and the computer should run in the "warm-up" mode. If you do not have the ability to force the computer to use a custom curve and it is currently not responding correctly with timing and fuel you could buy or create a device to alter the output of the IAT sensor to make the vehicle think the air is a different temperature. Since the IAT is a resistor at heart and most of them think it is colder when the resistance increases that is the easier thing to do (add more resistance- you could even have a pot and make it seem colder). This type of thing is already available and is used on non-mod cars to tweak performance a little (and burn more gas).

If you want the computer to think the air is hotter you need to decrease the resistance. You could replace your IAT with one that had less resistance (thinks it is warmer) for the same temp and then get something like above to dial-in the number. You could also set up a pot in parallel to the current sensor and then dial-in "new" temps. This is a bit more complex since it is not linear : 1/Rt = 1/Ra + 1/Rb
Rt= is total

Hope that is helpful information.

 

[Andy330hp]

Yeah, then you could buy a tornado for your intake and a display that shows "A/F" based on only your narrowband EGO

Keep in mind that something such as suggested above should not be viewed as "permanent". MAking your ECU think it's cold starting will also alter your spark advance, cranking fuel, transient fueling, shift schedule, idle A/F, and will probably reduce economy and maybe overheat your cat if you go too far.

For what its worth

 

[albertfastsubaru]

Cold start is only when the engine temp = intake air temp. You would never want that as a permanent condition and wonce the engine is warm (over 180) this would never be the case.

For the purpose described here the intake air temperature "adjuster" could be used to change the computers view of intake air temperature to make it more inline with the actual temperature (hotter than the IAT before the supercharger thinks it is).

Example:

IAT reports 80 (real)
Supercharger compresses the air
Actual IAT is now 115

If the computer is dealing with this incorrectly setting the engine parameters (too much fuel and spark advance) based on the IAT something needs adjusting. Ideally (as mentioned above) you would have temperature compensation that you could program on the ECU. Lacking that, if you could tell the ECU a temperature to make it work correctly you could achieve the same results. Playing with the overall resistance of the IAT sensor would allow you to change the ECU view of the temperature.

You could replace the IAT with a known resistance for certain conditions, have an adjustable resistance and just dial-in something, add resistance to the current IAT sensor to make the ECU think it is colder but still track with outdoor air temp, or reduce resistance to make the ECU believe the air is hotter (and reducing resistance is the most difficult since it involve either a parallel resistor and non-linear adjustments or finding a replacement sensor that fits your setup AND has a lower resistance to start with).

In this case of the original requestor he would like need to tell the ECU the air is hotter than the current IAT is reading. He has the more difficult problem (decrease resistance read by the ECU on the IAT circuit).

A few possible possible answers for the original person:

1. relocate the sensor to somewhere hotter (like after the supercharger) and then use a pot to adjust the too hot reading down to the "cooler" reading you need.

(This is the best choice as it would know when the SC was cranking. Heat soaking could be an issue but should be dimished when there is a large airflow)

2. Find a temperature sensor that has less resistance than your current one at the same temperature (and add a pot if needed) - would just be some offset but not know when the SC was really working hard and making the air hotter.

(1 and 2 would track temperature for you with your offset since it still has a real sensor)

3. Replace the IAT completely with a pot and dial-in your own IAT for the current conditions (this is not perfect and would be something you might do for race or fairly static conditions)


Ideally your ECU would allow you to change the response curves to make this work without this "magic." If it does not you can try these things.

 

[Andy330hp]

Actually, maybe cold start will be the least of his worries

As stated above, IAT, while it does affect fueling somewhat, on a MAF system it does not affect fueling nearly as much as it does advance. So, he will have to tell the ECU that it's WAY hotter to get just a little fuel out, and in the process have his timing retarded a ton. If he has a cat, it may not survive this experiment.

The problem is, by modifying tables you are only affecting one output. By changing an input, you are affecting a RANGE of outputs.

No matter which way he goes, doing what you described should not be viewed as a permanent solution. It may be used as a tuning aid AT MOST. If the air temp trick works for correcting the fuel, he should then have a chip burned with a new IAT compensation table, so that he can get the correct advance again.

The person who originally posted needs to find out what is really causing the excessive fueling, whether it be a faulty MAF sensor, an air leak before the oxygen sensor, etc. I could see how the superchager has changed his VE: Assuming he didn't change cams, the supercharger could be blowing a lot more air through the cylinder and out the exhaust with high cam overlap, hence the same MAF reading would correspond to less "consumed" air and thus an overly rich condition. It could also be something completely different. Do some more research

 

[patprimmer]

Andy,

I would have thought that some fuel as well as air might blow across the chamber and out the exhaust during cam overlap, having no effect on final mixture, but having said that, blown motors like wide lobe centres (like in the region of 116 deg) so as not to waste to much air that you had to pump into the motor.

Narrow lobe centres on a supercharged motor do help to cool the exhaust valve.

If the function of this sensor is to provide the ECU with data to correct for cold start, why would it not work if the sensor was simply left hanging near the air intake, or left in the inlet system, downstream of the throttle, but upstream of the blower. If the throttle plate was downstream of the blower, it would get real complicated and beyound the scope of someone without the resources to reprogram and remap the ECU

eng-tips, by professional engineers foe professional engineers

Regards
pat

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips Fora.

 

[Andy330hp]

Perhaps I should simplify my position

The drastic overfueling experienced by Brandnew is almost certainly not the sole result of IAT sensor position

On modern OEM ECU's, it is hard to say with certainty (unless one of you here actually helped develop it) that we are aware of ALL the parts of engine management that use the IAT sensor input

Haphazardly altering the IAT input, while possibly providing a hint as to which direction one must ultimately go with their tuning, is not a good idea becuase of all the unseen effects associated with it.

The correct solution is not to attempt to overcompensate by telling the ECU it's 20 degrees in the intake when its actually 120. It's to figure out WHY the engine is getting such incorrect fueling in the first place.

I did not mean to turn this into a discussion about how a MAF sensor system responds to cam changes, for as noted above I do not have enough experience with MAF systems to feel confident. I was offering a possible starting point only, not a certain solution.

 

[hotrodwrx]

Just to answer your question the fuel trim is off by 15-20%. It's actually running rich (i.e. -16%). Not everyone has the problem but those that have it can't seem to get it to go away.

Is this fueling error across the board at all rpms, and boost pressures, or does it change with boost? If so it could be as simple as skewing the fuel pressure slightly. If the static fuel pressure or the absolute pressure across the injector (fuel pressure - manifold pressure) is not what the ECU expects, it will calculate the wrong fuel injector pulse width.

Most turbocharged, supercharged cars have a raising rate fuel pressure regulator so the delta P across the injector remains constant regardless of boost. If it is running a fixed fuel pressure then as boost increases the relative pressure across the injector drops and fuel flow decreases. Some of the after market supercharger kits use a rising rate fuel pressure regulator that increases fuel pressure at 2:1 or some other compensation rate, sometimes using this as a poor mans way to modify fuel flow to compensate for the increased boost pressure.

Some fuel pumps have problems keeping up flow at higher fuel pressures developed in supercharged systems and have non-linear fuel pressure (can't maintain both flow and pressure at high boost settings).

You might want to measure your fuel pressure with respect to boost and confirm it is correct.

Larry