Quick Question regarding sensors

[24OZ]

Guys,

 

I still have a lot of reading to do on MS, and I hope you can help me out as I am about to put my engine back in the car and havent been able to find this out, yet.

 

All i have left to do is to bolt on the intake manifold, whilst it's off do any sensors for MS need to be taped in, as this would be the ideal opportunity to do it.

 

Thanks.

[mobythevan]

It would be nice to have the intake air temp sensor tapped in to the intake manifold. There is some arguement about heat soaking problems, but the general consensus is to put it in the intake manifold.

[24OZ]

Thanks for the reply Moby.

 

I was thinking about putting the IAT into the J pipe or intercooler pipe (no IC just yet though) just before the throttle body.

 

Will do a search and see why others dont like that idea in favour of tapping into the manifold instead.

[z-ya]

The closer the IAT sensor is to the cylinder intake port, the more accurate the measurement. IFf you mount it upstream of the TB, the MS will get incorrect IAT sensor readings. Drill and tap a 3/8NPT hole in the intake manifold, and be done with it.

[BRAAP]

Sorry guys, but I’m going against the grain on this and here is why, with data logged MAT.

 

The purpose of the IAT is to measure the temp of the incoming air that the engine ingests so the ECU can compensate for air density due to temperature changes, (it doesn’t not take into account air density due to humidity, it takes a Hot wire MAF to do that).

 

I personally feel the intake manifold is NOT the place for the IAT. Mine is in the manifold currently and it heat soaks BIG TIME!

 

The air that reaches the combustion chamber is not the same temp as the intake manifold itself, though the slower the air flow through the intake, the lower the Delta “T” between the air and manifold itself, (Delta “T” is temperature differential), as the temperature of intake manifold will have an influence on the temp of the air that is going into the engine as the air has to pass through that heat soaked intake manifold. Also of note, the temp of the intake manifold will be greatly affected by the amount of air flowing through it. i.e. at idle, not much air flow, it will heat up, but at WOT, lots of air flow to cool off the intake manifold, and vice versa. So yeah, it is a good idea to have the ECU be able to adjust the AFR for what affect the intake manifold has on heating the air entering the engine. In mounting the IAT in the manifold itself, the ECU now is reading mostly the temp of the intake manifold itself, not so much the actual temp of the air the engine is ingesting. I feel that mounting the IAT further upstream would be the best compromise. How much upstream? That will be vehicle dependant and you’ll have to data log that yourself.

 

My best guess for my N/A application, (in boosted applications, you want to measure the temp of the ‘boosted” air so placement will be different, but the same theory still applies), is to install the IAT in my intake tube ahead of the throttle valve, but yet still in the engine bay. I think that would be the best compromise as the IAT will still heat soak, though to a lesser degree and hopefully more consistent to what the actual air temp is going into my cylinders. I still have to get to that before I sell off the intake system and the engine out of the race car.

 

 

Here is a data log approx 40 seconds in length with my IAT mounted in my intake manifold. Note the temp swings and how those swings follow the throttle position.

 

Top graph is “TP”, aka Throttle Position and “MAT” aka Manifold Air Temp The second graph is just “MAT” only. Note that the MAT max was 144 degrees, min was 99 degrees. Also, you can see the temp drop steadily while the throttle was WOT. The IAT sensor was registering “HEAT SOAK” and then would cool down gradually while at WOT. This particular data log is 1st gear, 2nd gear, 3rd gear WOT blast, then a quick 5th shift for a brief part throttle cruise!

 

I do agree that intake manifold heat soak does affect the air density of the air the engine is breathing, but not to the degree my IAT was registering.

 

 

[rztmartini]

what if you isolated the IAT by making a larger hole then threading the IAT in a material that will not be affected as much by the heat soak (maybe delrin or another composite material) then threading this into the larger hole? might work...shine some light through any holes in this theory...

[z-ya]

The IAT sensor is an open air sensor, so it is already isolated conductively from the aluminum intake manifold.

 

Paul,

 

Is your fuel system connected like this:

 

pump->fuel rail->regulator->tank

 

It looks from the picture that the fuel rail is a dead end. If this is the case, vapor lock in the fuel rail is probably what is causing heat soak. Also, you are not running a heat shield with that intake manifold. I would either wrap your header, or put the stock intake manifold heat shield back on (or both). Yea, it doesn't look as pretty, but it sure is functional.

 

Placing the IAT sensor somewhere else besides the intake manifold is just a band aid attempt to fix a heat soak problem. When air moving through the system, heat soak is not significant, so having the most accurate measurement of IAT will be in the intake manifold. If you located it somewhere ahead of the TB, the IAT compensation algorithm with not be be getting the actual IAT. Also, if you want the IAT compensation function to have any chance in minimizing a heat soak issue, it must know the temperature of the air in the intake manifold.

 

On my supercharged car the IAT reaches 280F on the track, and I have no heat soak problems at all. The IAT sensor is in the manifold. I'm using a stock fuel rail (no dead end), and I wrapped the header. I can come off the track, wait 5 minutes, then start it back up without any symptoms of heat soak.

[BRAAP]

I wasn’t referring to “fuel” heat soak as in the fuel rail, (I haven’t had any issues wit the dead end fuel rail.) I was referring to heat soaking of the intake manifold itself and that heat soaking affecting the temp of the air going into the engine, not fuel.

 

 

I like the idea of thermally insulating the IAT from the intake manifold. Good idea.

[z-ya]

Paul,

 

Again, try to insulate the intake from the header, this should fix your problem.

 

I'm not a big fan of dead end fuel rails in EFI applications. With the fuel rail in series with the fuel flow, you will always have cool fuel in the fuel rail. The only fuel that can possible heat up is the small amount in the injector. In race (WOT) type applicaitons, the fuel isn't in there for long.

[RTz]

Looking at the bottom (white) trace, the IAT is dropping SLOWLY (with the throttle opened). About 40 degrees over a 10 second period. If that was a REAL reflection of air temp, it would drop virtually instantly when the throttle is opened. At least, thats how I see it. Looks like the sensor itself is heat soaking.

[z-ya]

Ron,

 

Almost every time I drive my turbo car, I watch in real time (Wolf3D), the intake air temperature (IAT) change dramatically when the throttle plate opens. But the system takes time to respond. It will take seconds for the intake to cool, and the sensor also has a reaction time. The sensor does not heat soak, it varies electrical resistance with temperature, that's it. As long as the temperature is within the temperature operating range of the sensor (which it is), the temperature it reports (in the form of resistance) to the MS is accurate.

 

A 144 deg F inlet temperature is actually quite cool (the graph is in degrees Fahrenheit, right?). If it was Celsius, that woudl be 291 degrees F. Like I said in a previous post, my track car runs temps that high without any heat soak problems.

 

Paul, you are using an open air sensor, right? If you are using a coolant sensor for IAT, the heat of the intake can conduct into the sensor giving false readings.

[BRAAP]

Paul,

 

Again, try to insulate the intake from the header, this should fix your problem.

 

I'm not a big fan of dead end fuel rails in EFI applications. With the fuel rail in series with the fuel flow, you will always have cool fuel in the fuel rail. The only fuel that can possible heat up is the small amount in the injector. In race (WOT) type applicaitons, the fuel isn't in there for long.

 

If I insulate the headers, yes I agree that would help reduce the heat that the intake absorbs, probably to a great degree. The issue still remains that the intake manifold will still heat soak just from being bolted to the hot cylinder head and the intake and the exhaust flanges are in such close proximity to each other, the intake manifold will still absorb a fair amount of heat which will still alter the IAT readings from what the actual inlet air temp is, (I refer you to the “data log” pictured above).

 

Now this heat soak “issue” that we are talking about is only related to the IAT itself, not the fuel system or the injectors etc.

 

I do feel that based on my findings in having my IAT in the intake manifold like it is, that ceramic coating the intake manifold is a good idea, including the manifold mating surface.

 

 

Well, as for my “fuel”, I don’t really have any issues with it being a dead end set up. The car starts up great, even after making a hard run, then letting it idle with the hood closed in 100 degree outside temps, i.e. a thorough hot soak of the engine bay, the engine will fire right up, no hiccups etc. I feel that the dead end fuel rail is a simpler installation, visually cleaner, and it functions no different than the flow through design, but this is just my opinion. FWIW, quite a few of the current production cars are using a similar system but they are running dead end from the rear of the car, mine is only dead end from the firewall.

 

I get the impression that you don’t like warm fuel? I would agree that “cool fuel” does help absorb heat from the intake air charge as it is being squirted into the air stream, but there are arguments for preheating the fuel as it will atomize better when it is squirted. As for which way produces the most power or most crisp throttle response? Dunno, I don’t have enough evidence to support it either way other than I feel hot or cold, the engine will run fine. Too hot might affect how well the injectors work and how long their longevity, but that is another issue altogether. Being as it is high pressure EFI, 40 PSI, I have yet to realize any heat related issue regarding my fuel system. Only that my IAT is readings are affected by the fact that my intake manifold is absorbing lots of heat that my Inlet Air Temp sensor is recording even though my actual inlet air temp isn’t that hot, i.e. erroneous IAT readings due to IAT placement in the manifold.

[BRAAP]

Yes, it is an open air sensor, not the H2O sensor. What I find interesting is that the IAT installed in the AFM of the factory EFI Z cars will actually read almost instantly any changes in temp. I stuck one in the fridge and it instantly measured the change, removed and again instant changes. I then proceeded to use that sensor for a budget tire pyrometer with my DMM to get and idea of temp differences across the tread, not absolute temps. At any rate, I would like to think the GM open air IAT should also record very quickly any changes to air temp. Might have to play with that a bit to verify it.

[BRAAP]

Talking about the response time of the IAT sensor was really started to dig at me so I went out to the shop and data logged the IAT removed from the intake manifold to see just how fast it would respond to air temp changes using a heat gun and compressed air blowing across the sensor. As I mentioned previously I had hoped that this “open air GM IAT” would have a quick response time and it does.

 

Ron Tyler helped as he handled the heating and cooling chores while I monitored the MS gauges and data-logging…

 

We also found some other interesting info as well as the Temp rate of change. The body of the IAT senor itself will heat soak….

 

Ambient Temp in the shop was 55 degrees F, initial readings when we powered up MS on the IAT gauge in MS confirmed that 55 degrees. Then while Ron held the IAT by the wire pigtail, he used the heat gun to blow across the IAT and the gauge screamed from 55 degrees to 150+ in a matter of couple seconds. He then removed the heat gun, and the temp reading held steady, no change up or down. He then grabbed the sensor around the metal base, at this time the gauge started to read a nice slow steady decline in temp. Then he grabbed the pigtail again, the temp stabilized right around, 120 or so. Then he blew shop air across the probe and it rapidly dropped till he removed the air at 55 degrees reading. Then we started the data log, added heat, paused, added cooling, stopped the datalog.

 

This pic shows the datalog from 215 seconds to 240 seconds time span, (Time line is at the bottom in aqua color. Each aqua bar along the bottom is 2.1 second intervals, (don’t ask me why it is at 2.1 second interavals, it just is). Starting temp was 55 degrees. The vertical aqua bar is at the point which Ron immediately removed the heat gun, 163 degrees. Heat started at approx 217 seconds, heat removed at 221 seconds, a temp rise of 108 degrees in only 4 seconds. Not instant, but rather quick. From 221 seconds till 231 seconds, Ron just held the sensor by the pigtail, 10 seconds, temp change was only 1 degree from 163 to 164 back to 163, even though the sensor was held in 55 degree open still air. Then he blew shop air across the sensor at 231 seconds till 236 seconds bringing the temp reading back down to 54 degrees. The two steps on the down slope are where the air stream missed the air temp sensor.

 

Conclusions;

 

1) The GM IAT is capable of reading changes in air temp rather quickly.

 

2) The GM IAT will heat soak within itself and remain at whatever temp it last saw for quite some time, unless another strong outside influence is presented, i.e. air flow across the sensor itself or heat being transferred through the body of the IAT.

 

3) Whatever object that is directly in touch with the body of the IAT, (what it is threaded into), definitely has a strong influence on what the IAT reads.

 

 

Based on what my previously posted datalog shows in regards to the rate of IAT change on the running engine, I am now quite convinced that whatever the IAT is mounted in/on will have an affect on its reading and that reading is not the true temp of the air being ingested into the engine. That would mean that the ECU is not adjusting the AFR in direct correlation to what the actual air temp is, i.e. the fuel correction being applied to the fuel map is not being properly compensated for its density due to temp.

 

The real question is just how important or how much affect this has on the Air Fuel Ratio? Not sure, but it will have some affect and we should at least be aware of that and try to optimize sensor placement in our custom EFI systems based on that.

 

 

[z-ya]

Great work Ron and Paul. I no idea that even the open air sensor would be effected so much my the temperature of the material it was threaded into. I wonder of there are other sensors available that are less effected by this problem.

 

I am still confused as to why you are having this heat soaking problem, when I'm running the same setup, and I have no signs of it at all. Infact, the JCR intake manifold is less than 1/2" away from the header. I do have the header wrapped though. Still, the intake hits over 280deg F easily.

 

Best thing to donow is move the sensor for a test.

 

Can you explain in more detail what the circumstances are that cause oyur heat soaking problem?

[z-ya]

One other thing, did you use teflon tape on the IAT sensor threads? I did, and I used a fair amount as the person who originally threaded the hole for 3/8" NPT went a little too deep with the tap. From what I remember, Teflon is an excellent insulator, both electrically and thermally.

[woldson]

though that was extensive research when you say heat soak at what temp/ or dose it not change anymore, at what temp. I can say this that even though it was a hand and slow change, shop air kicks butt ie cooling welds. sooooo if manifold is at 230 and air is at 150 across sensor me thinks there is enough isolation to see a temp of 150 to 170??? at the sensor. now it will depend on the speed at witch the air is moving across the sensor. if you can measure manifold temp and look at sensor temp under different conditions you see the diff. without a supercharger or turbo your air intake should be very cool. mabey wrap the intake? Is your fan mabey pushing air out instead of in? This is just my 2 cents but i have worked with cooling quite a bit in computers. also you would have to ceramic coat inside and out side of intake manifold.

[RTz]

The issue I see, using an isulator, is that it could possibly work against you, especially in a street driven car.

 

For example, Pauls manifold heats up under cruise when the airflow (through the manifold) is low for a long period. Its heating up mostly from exhaust (as you mentioned earlier, Pete) and to a lesser extent from contact with the head, and mabey a touch from reversion. This heat soaks the manifold, and in turn, heat soaks the IAT's mount. When he goes WOT, this "air cools" the manifold, in turn, cooling the IAT's mount. If you used an insulator (teflon tape, delrin, etc) then its an insulator both ways... drive for an hour on the freeway and the INSULATOR is going to heat soak... and it will take even longer to cool down at WOT.

 

In my opinion, there is no perfect place for an IAT. Just before the point of fuel contact would probably be ideal. Readily available sensors don't allow us to do that. So, every location is a compromise. The compromise that I choose to make is just upstream of the TB, mounted in some sort of plastic. Obviously, that won't work in your SC car.

[z-ya]

Ron,

 

I could mount it right in front of the TB (where we put the dyno air temp sensor), but I am not having any problems with it in the intake, so its staying where it is.

[24OZ]

Great research guys.

 

Someone gave me a tip to fit it in the third runner of the Intake manifold. There is already a hole there for the stock BOV and you just need to retap it.

 

Which brings me onto my next question, i know i need a 3/8" TAP.. does it matter how many threads per inch it is? I saw one on ebay 3/8" TAP No. 18.. would this one work?

 

Thanks.