Hey Mahjik thanks a lot for that information... I think I will drill into the stock elbow...we'll see. On a side note I'm curious if it's ok to just use regular 18-gauge wire. I found this great article on AIT for other cars and they said they needed to buy a special wiring extender otherwise the sensor could read innacurately?


http://www.autospeed.co.nz/cms/article.html?&A=0360&P=2

 

Even though this won't happen because the air is moving way to fast to have any significant amount of heat transfer throught the UIM, if the computer thought the air was cooler than it actually was, you'd get MORE fuel. This is a win-win mod, and there is data backing it in the PFC section.

 

They are talking aboute extending "Thermocouples". I believe all you are extending on the FD is just standard electrical connections (someone correct me if I'm wrong). In the later case, standard 18-Gauge would be fine.

 

So someone has measured the temps inside the UIM from other areas other than just the standard AIT location? If so, I'm curious to see the difference as that's what I'm referring to, not temps from the IC piping. We've all see those posted.

 

Mazda did allow for it to some degree. The correction map squirts more gas than needed when hot, reducing the impact of heat soak. Check my post:

https://www.rx7club.com/forum/showth...hreadid=137398

At very low airflows (idle or low hp cruise), the stock location will reflect air actual temps. With a moved sensor, the idle control and closed loop control should avoid overrich conditions. Actually may improve 'tip-in' response.

BUT on a cool/cold day, at wot with very big IC, esp fmic, you will be at the 'leaner' end of the correction map with stock ecu, and a PFC is needed, esp with other mods. This is true with sensor at stock place, and to a lesser degree at the moved positions.

 

On a 90 degree day, i've stopped at the store, come back out after a few minutes, and upon restarting (and for at least 5-10 mins after even if i get on the highway) seen temps on the order of 70-80 C on the PFC contoller.

Sorry, but theres no way you'll convince me that 90 degree air that was just sucked into my intake, was not subjected to significant pressurization (didn't get into the boost), ran through my intercooler, and into my UIM became heated 70 degrees (70 C = 158F) in a matter of SECONDS to travel the entire intake tract... that HAS to be an innacurate measurement...

I do agree that the PFC actually ADDS fuel when the temps go up past a certain point (contary to popular thought)... i'd guess on my car thats about 55-60 C based on the fact that it runs like crap after that point.

I'm less worried about a lean condition than i am an over-rich one in summer/around town driving... once i see 60 C, i might as well take it home.

 

the sensor stayed hot 5-10 minutues in your example because the uim and the metal part of the sensor stayed at that temp. at idle, and low hp (high vacuum), air takes much more time to flow thru the still hot manifolds and ports, and sensor temps will reflect slow moving air temps into the chambers. once you go to modest power, flow increases and ingested air will be much cooler than the soaked sensor readings.

 

You know what, all of this talk brought me back to my original point. It may be the fact for aftermarket ECU's that the AIT should be moved to relate "actual" air temperatures, BUT why for the stock ECU? Wouldn't Mazda tune the ECU with the AIT in place and a wideband O2 for the exhaust? I mean did Mazda really care if the stock sensor had accurate readings? As long as they were PREDICTABLE readings then it should be fine. The only disadvantage to everything is the heatsoak, but how many of us know the code inside the stock ECU and how it runs the vehicle? Don't you think after producing this car for almost 10 years they would've done something about this? (who knows, we know what they did about other problems :P) I'd like to do this modification and I NEED to install my sensor in the next day or so, but with everything else being stock I don't know if moving the sensor is a good idea at this point?

 

You need to install your sensor in the next day? Are you replacing a failed one? If so, I would recommend just replacing it in the stock location and verify you fixed your problem. Then if you want, do the modification at a later time. No sense in creating more than one problem all at once.

Now, if your current AIT is just fine and you are just thinking of doing the mod for the heck of it.... I would still say wait... Most of the guys that do it are using a PFC or have something they can monitor their temps.

 

Hey Mahjik, good call. I was thinking about doing the mod but I'm having 2nd thoughts. Plus I don't have anything to monitor the temp right now anyway. We'll see what happens The car's in the garage :P

 

Why doesn't someone buy a second IAT sensor and install it in the elbow. Monitor those temps and then compare it with the stock IAT sensor location. Then you can look at both temps at the same time. This would help us determine if the stock location of the IAT sensor really has a problem with heatsoak and also how fast it reacts compared to the elbow location. It sounds like it is a good mod though and I will probably do it. I have noticed that my intake temps are very high after a hot start. It takes more then a few minutes to come down in temp and that is at 40 plus mph. If you are driving slower it takes even longer. I do not go WOT untill the temps come down some. Which means I have back out of many stop light races and freeway onramps.

 

Ok, it sounds like you just volunteered.

Guys, if you look at the data-logit default values, you will see that the most fuel taken out (80C) is about 1-1.5%. This is not that much unless your car is tuned to the ragged edge.

 

I think you missed the point of the discussion. There isn't a question of the temp difference from the elbow to the UIM. The question is, since the UIM is so hot verses the elbow (especially the stock elbow), how is that effecting the air temps in the UIM? Would moving the IAT from the bottom of the UIM to the top provide better and more accurate readings?

There are already people who have compared results of the stock IAT location verses moving it to the IC piping. Just search through the PFC forum.

 

Well after all of this time I should update this too...I ended-up installing the sensor in the elbow basically as a test. I figured if it didn't work I could always switch it back and it was a good project. The only problem is, I don't have a PowerFC to watch the temps so I just have to hope it's doing something. Surprisingly after this install the car feels a lot smoother through the 3k rpm range, not as much stuttering. The hardest part of the installation was soldering new wires to the old ones (because the old ones were brittle and very short). I just ran the new wires through an extra piece of vacuum line I had so it matches the rest of the engine bay. I left the old sensor in the stock location so that there would not be any leaks. So far so good