Air temp sensor location

Hey everyone! I've got a Link G4+ plug-and-play ECU on my version 9 STI and I'm looking to put an air temp sensor in the manifold so that my ECU can compensate for the cooler air temps my water injection will bring.

I have a Bosch open element sensor that I would like to use and I'm uncertain of placement. I'm thinking that I would either like to put it in the floor of the plenum sticking straight up so that I did air flow across it coming out of the throttle body, but I need to make sure that I have space underneath the manifold for me to put the sensor there. I would also need to have a bung welded on for that location which will cut down my clearance a little bit as well.

My alternate location is in one of the runners which seems like it would always have good airflow across it and it's closer to the port so it should be a more accurate air temperature. However, I'm worried that it might cause turbulence which may affect fueling to that cylinder. Do you guys have any thoughts or comments? I don't daily drive the car but I would like to have the ability to if needed, so I would like to avoid difficult heat soaked hot starts, which I do think the plastic sensor will help cut down on versus the metal GM sensor.

I have included pictures have both my proposed locations

I'm open to any input! Thanks!

The image uploader doesn't seem to work, so here are some links for the pictures I took

https://drive.google.com/file/d/1pQMTF77ts3Lf9lIoy9a4Jie4tHulovN5/view?usp=drivesdk

https://drive.google.com/file/d/1opsO1nXJpXZ7-jxnvbgBy-jlz_Osoip5/view?usp=drivesdk

Plenum is preferable to runners. It's a more stable temperature.

Any reason why? Just curious, I feel like the runner will always have air flowing across a sensor if I place it there. But there is less volume, so I could see it heating the air faster

Put one sensor in each runner and you will see that they vary significantly from runner to runner..

Cory your first link is accessible but the second tells me I have to ask permission, presumably from you, to view it.

Sorry about that! I just updated the permissions

https://drive.google.com/file/d/1opsO1nXJpXZ7-jxnvbgBy-jlz_Osoip5/view?usp=drivesdk

There's always multiple view points on where is the best place to mount the IAT sensor so you need to consider opinions and reasons and then make your mind up about what you feel is best in your application. What we need to consider is what we are trying to measure which is the temperature of the air entering the cylinder. On face value it may seem that the inlet runner is the ideal location, however you'll find that it suffers badly from heat soak when the airspeed is low (idle and cruise). The picture you've posted also shows it will block a reasonable volume of the inlet runner which probably is less than ideal. Given your two options I'd suggest the plenum floor, however you're still going to find the sensor reading susceptible to heat soak, particularly with an alloy plenum.

Personally I prefer to mount the sensor in the intake plumbing pre throttle body. This makes the sensor a little less susceptible to heat soak and personally I find I get a more stable tune. You do however need to consider that at low airflow there is going to be some heat transfer from the inlet plenum and head into the inlet air so hopefully you can get the impression that accurately modelling charge air temp is tricky at best. There's considerably more potential to improve this with ecus that offer a charge temp estimate table where you can bias the charge temp towards the IAT or ECT sensor value based on rpm and load.

Thanks for the information Andre that is very helpful!

My follow-up question for that is regarding my water injection setup. I was planning on having the air temperature sensor Downstream from the water injector, and I had planned on putting the water injector on the outlet of the intercooler (STI with a top mount intercooler) I'm afraid that since I only have a couple inches of space I will either need to place the injector after the air temp sensor and ignore any cooling effects I see, or I will need to place the temp sensor right after the injector and risk overcooling the sensor due to close proximity and getting a false reading.

Do you have any thoughts on that? It is a bit of a tricky situation unfortunate :(

If I may just pipe in on the IAT sensor placement. GMH (Holden) fitted the IAT in the inlet runner on the venerable injected V8 from the VL Group A right through to the VS (1987-1996). When they swapped from MAP to MAF in 1997 they moved the IAT to the inlet tract inside the air filter housing. Later after the demise of the Holden V8 and the introduction of the Gen III and then Gen IV V8 it become part of the MAF. Now that background is out of the way in my experience on Holden V8s they are easier to tune when the IAT is in the inlet runner itself. Having said that the IATs Holden used are small compared to what you have in the picture.

If I'm not mistaken, Holden used a GM sensor correct?

I've been back and forth on using them, they seem to be the tried-and-true method for standalone ecus, but the sensor is also a brass casing and I was thinking that might lead to more heat soak.

Yes Holden used GM sensors. Holden is/was GM. The sensor in the manifold is brass, the sensor on the intake pipe/air filter housing is plastic and depending on where they are fitted they can, and do, melt. I, personally, wouldn't fit a plastic sensor in any part of the manifold unless you somehow stopped heat soak from the block to the manifold.

Has anyone here used a fast response sensor like the T1 Race potted glass bead thermistor mounted right near and inlet port on an ECU such as a G4+ with charge temp estimate table? If so, what differences have you found in how the charge temp table ends up compared to the traditional pre-throttle IAT position?

I've got a couple of the T1 sensors and plan to mount one into the TGV delete on my EJ275 to compare the readings between that and my pre-throttle sensor, then try each one individually with the charge temp table.

I'm at this same stage with my build. My setup is a twin turbo V6 with twin intake plenums. My plan was to use two of the Bosch composite TMAP sensors ( http://www.nzefi.com/product/bosch-3-bar-t-map-sensor-with-weld-on-boss-connector ) bolted directly to my manifold, one in each plenum, where the sensor probe would be pretty much sticking out right at the entrance of #3&4 runners on each plenum as my runner bell mouths are quite close to the manifold walls.

The bosch sensor body is advertised as to be made of a composite material to resist heatsoak and can be mounted directly to the manifold.

Would doing this still not be recommended with the composite sensor body? Haven't found much feedback of people in the aftermarket using the TMAP sensor , other than that factory cars uses it to mount directly to manifolds, and most people I see who complain of heatsoak with direct manifold mount are using the normal metal bodied IAT sensors.

I see lots of ECUs now have a feature that uses the coolant temp sensor around idle and low rpm (charge air temp) as mentioned by Andre earlier in the post, as a more accurate air temp figure at the intake valve as the air velocity in the intake isn't high and if the IAT sensor is mounted in the cooler piping, the temps of the air will be higher by the time it goes into the combustion chamber as it will pick up some heat in the manifold. With the composite sensor mounted directly to the manifold and at the entrance of the runner I would think it would be the best of both worlds as in the sensor would read the more accurate higher temps at idle and low rpm and will also read the correct temps entering the runner at the higher rpms as air rushes past the sensor to enter the runner.

This is all my speculations and would like some more expertise thoughts and opinions on this method of those who may have more hands on experience before

What ECU are you using? Can it utilize two of those sensors?

I'm using an AEM Infinity. Has loads of GPIOs to use these sensors. One would be to use during running and the other would be for data logging purposes.