Air temp sensor and driveability

[Zmanco]

Disclaimer: I am NOT looking to reopen the debate on which location provides the most accurate intake air temp readings.

 

As I've been tuning with megalogviewer, I found that sometimes it would make a lot of changes to my VE tables and other times just a few. The big changes seemed to happen when I would tune when cool, and then do another session later when hot out. Same thing when reversed.

 

I also noticed that I had more lean-surging and hesitation when driving during the day in stop/go traffic than the evening when it's cooler. Eventually I came up with the hypothesis that the IAT is reading higher than it should on hot days and perhaps leaning things out more than it should (hot air is less dense than cool air and hence needs less fuel - all else being equal). This is with the sensor mounted in the plexiglass spacer just behind the throttle body.

 

As an experiment, I moved it to the center of the intake manifold and have found that I'm getting much more consistent driveability from a given tune (VE table). My guess is that the air temp readings, while probably skewed high due to heat soak, are on a relative basis to each other more accurate. In other words, on a hot day, the IAT readings are raised more in line with the actual temp of the air going into the cylinders.

 

I plan to leave the sensor in the center of the intake manifold as on my car it seems to give me the best compromise for driveability. If it was a dedicated track car, where part throttle didn't matter so much, I'd probably go back to just behind the TB.

 

BTW, this is with the stock intake manifold from an '80 (with the webbing between runners), a ceramic coated header below it, and the GM sensor.

[z-ya]

I think either place you had it should work OK. Maybe because you have it in the middle, you are getting more of an average temperature across the intake manifold. I also usually mount them close to the middle of the intake. Racing or not, that is a good spot for it.

[Zmanco]

When I drove with gauge on the laptop showing the intake air temp, I noticed that in low speed driving on a warm day the temp seemed to average higher than when mounted in the plastic spacer. (Note: this is very subjective as I did not do back to back drives under the same conditions with the sensors in each position.)

 

My theory is that when the VE table was tuned properly for a cooler day, the resulting much higher temp on the warm days caused MS to lean things out too much. Having the sensor in the middle of the intake manifold seems to reduce the difference in temp between the cool and warm days (probably by reading higher on the cool days), resulting in a mixture that is closer to optimum for both conditions.

 

At least that is what I think is going on.

[WizardBlack]

I've put standalones in a few cars in my day. My personal preference is the outlet endtank of a front mount right next to where the pipe is connected. This keeps it out of the engine bay and close to the flowing air.

HOWEVER, considering your bold statement at the top, I'd say anywhere where you get consistent heat is going to give you consistent readings. Either somewhere where the heat is pretty constant (intake manifold) or non-existant by relative standards (FMIC) should work equally well. As long as you tune it for the location and it heats up with warmer air, it really doesn't matter.

[BRAAP]

Disclaimer: I am NOT looking to reopen the debate on which location provides the most accurate intake air temp readings.......

 

 

[trippintl0]

By the way, I've been reading this thread - I have the same problem. Car runs fine until it gets good and warm, and then when I get to a stoplight, it wants to run lean, because sensor is reading like 155F at idle. I decided I would make a little mod in the .ASM code to discount the air temp sensor if TPS is < a certain value. >> BUT! << as I was digging through the code, I realized there's already a provision for this! I have MSNS-Extra Hi Def. code. The option is called, "Coolant-Related Air Density", under the "Advanced" menu in MegaTune. It was created for this exact problem. Basically, if the rpms are less than a certain value, it will NOT derive the air density from the air temp, but from a Mixture of air temp and coolant temp. (you can control how much weight it puts on Coolant temp vs. Air temp., and the RPM's that it will activate and deactivate at) As soon as I put it on, I had perfect idle, even when hot; Perfect hot-restarts, etc. And I still have driveability when the temp sensor is heat soaked. It's wonderful!

 

(note: yes, I realize this is a "Band-Aid" fix! I don't know why my Air Temp sensor heat soaks so bad. But this is working good until I can have time to fix the real issue, since this is my daily driver!)

[Zmanco]

Interesting feature - I had seen it before but didn't realize it as a way to deal with this issue. Would you post some screen shots of your settings? It's not immediately obvious to me what values to begin with. I'd rather start with yours than reinvent the wheel from scratch

[proxlamus©]

hmm ive noticed this too.. but never took the time to look at it

 

post up some screen shots!

[cygnusx1]

As wizardblack said, it's not so much the location of the sensor as it is the way you program it to deal with the numbers. Ideally, you want to measure the air temperature at the intake valve. In practicality, you measure the air in a more convenient location and interpolate with the programming.

 

Stay away from areas that experience fluctuating pressure changes. ie. the throttle plate. An open element sensor will follow actual air temp more closely through temp transistions.

 

If the car runs poorly at one temp or another, you are better off reprograming your corrections curve. Don't fish around for another sensor location until you have tried the recurve.

[z-ya]

Editing the firmware is a band-aid type fix. The problem is that each time you upgrade the firmware, you need to remember to get the source for the new version, insert your changes, compile, and then download. Then you need to test it all over again to be sure your changes didn't screw something up.

 

Although I have not had any problems mounting the sensor in the plenum and running the default code, Ron Tyler seemed to have discovered that thermal isolation of the open air sensors are not all that great. This would be the only reason not to mount it in the plenum. I'm wondering if you install it where the cold start valve usually is, but mount the IAT in a material that will do a better job of thermal Isolation. Use the two screws that would normally hold the cold start valve to attached the fiberglass, lexan, or other material you have the IAT mounted in.

 

If you mount it in the IC plumbing, or another intake pipe that is not thermally connected to the engine (thermally conductive), you fix the thermal isolation problem. But now the air temperature reading will be incorrect, especially at idle when there is very little air flow through the intake manifold. In a race engine where you are either WOT or idle, it doesn't really matter where you mount it. But if you sit in traffic a lot, mounting it where the most accurate temperature measurement can be made is very important.

 

Sorry if I have opened up this can of worms all over again...

[RTz]

Some observations and opinion...

 

Early GM MEFI used a central plenum location.

 

GM abandoned that location years ago. Most GM IAT's seem to be either in the airbox or the intake tube upstream of the TB.

 

All locations are a compromise.

 

I'll agree that the closer to the valve, the more potentially accurate the reading, but its not realistic with current technology.

 

If your tuning is such that you can keep you're AFR's consistent with varying temperatures, you've won, regardless of location.

 

Transitional pressure changes shouldn't be an issue if the filter settings are correct.

 

The insulation idea, used in conjunction with a Bosch style sensor, might backfire... if its heat soaked long enough to bring the insulation 'up to temp', it will take that much longer for it to return.

 

I was looking at the Audi 1.8T IAT sensor and it seams very well isolated. Might have to snuggle up with it...

 

[olie05]

There is a plastic open element GM iat sensor. I came upon it in the intake pipe of an early 90's Buick. Seems like this would be the most accurate solution, regardless of placement. (and without having to find the curve on the Audi one.)

 

Note:not my picture, just hosted on HZ for posterity.

[RTz]

There is a plastic open element GM iat sensor. I came upon it in the intake pipe of an early 90's Buick. Seems like this would be the most accurate solution, regardless of placement. (and without having to find the curve on the Audi one.)

 

That's the sensor I frequently run across. I wouldn't be afraid to try it. Though not boost friendly due to mount style.

 

The Audi IAT curve wouldn't be that hard to come by methinks.

[z-ya]

Thanks Ron, all excellent points.

 

How does the Audi sensor make a seal if there is no pipe thread? Do you need to machine a hole, then drill and tap a hole next to it for the mounting screw?

[Zmanco]

I'm wondering if you install it where the cold start valve usually is, but mount the IAT in a material that will do a better job of thermal Isolation. Use the two screws that would normally hold the cold start valve to attached the fiberglass, lexan, or other material you have the IAT mounted in.

 

....

 

Sorry if I have opened up this can of worms all over again...

I started with the standard GM sensor in the lexan spacer behind the TB. Just as Ron Tyler said, once heat soaked, it takes longer to dissapate the heat.

 

So does anyone have some screen captures (or just list the values) they used for the "Coolant-Related Air Density" settings?

[RTz]

How does the Audi sensor make a seal if there is no pipe thread? Do you need to machine a hole, then drill and tap a hole next to it for the mounting screw?

 

I'm told there is an O-ring at the mating surface. I've been eyeballing that sensor for a week or so, and since I have no self-control, I ordered one. Should be here in a couple day's. I'll post pictures when it arrives for confirmation.

[trippintl0]

Sorry I didn't post pics, I don't get on here much, and I have dial up right now. But here's a couple. Disclaimer: I have no idea if these settings are "Ideal". Of course this is a band-aid fix anyways, but regardless. I just played around with these settings for about 10 minutes, and the driveability got so much better, I haven't touched it since then.

 

 

 

 

good luck, hope this helps somebody else like it helped me!!

[WizardBlack]

There is a plastic open element GM iat sensor. I came upon it in the intake pipe of an early 90's Buick. Seems like this would be the most accurate solution, regardless of placement. (and without having to find the curve on the Audi one.)

 

Note:not my picture, just hosted on HZ for posterity.

 

That is slip in style and will only slip into the intake pipe (pre turbo). If you need one for a boosted car it should be threaded to go after the turbo and intercooler.

[RTz]

Bosch Part# 0 280 130 085.

 

Indeed there is an O-ring on the shaft.

 

Shaft dia., .468". I suspect a .484" hole (31/64ths) would be 'just right'.

 

Mounting hole, .260"

 

Temp range, -40 to +130 C.

 

.

.

 

 

 

 

Also, the calibration curve...

 

[X64v]

Mr. rontyler,

 

Any news as to the Audi IAT sensor? Heat soak tests, usage, and so on?