I am researching the feasibility of changing my intake air temp sensor at present so starting a thread here to record what information I can find out. Specifically I want to identify a *high speed* sensor that is compatible with the Power FC. I will conduct some testing along the way to try and ascertain what is the best sensor to go with and if there is any difference between various sensors.

The hypothesis is that by increasing the response rate of the IAT signal to the PFC, it will be possible to 'improve' the state of engine tune in both power and drive ability under different conditions.

One issue is what sensor is the fastest one that will work with the PFC??

The easiest sensor to make work is the factory GenIII sensor, although some testing I did with Spidey showed that it was not particularly fast to adjust.
(photo pending)

The most common one to use seems to be a GM MAT sensor, and is sold as the 'default choice by' ATS, Berk, and GTS Chris. My testing has shown this to be about the same as the factory sensor

_KUNENA_BBCODE_HIDEIMG

The sensor recommended by TDP is Bosch part number
0 280 130 085 or 0 280 130 039
[img]http://www.linkecu.com/products/InstallationAccessories/Sensors/boschairtempsensor[/img]
http://www.linkecu.com/products/InstallationAccessories/Sensors/intakeairtempsensors

And I have also found what is coined as a 'high speed' version of the GM sensor:
[img]http://www.bmotorsports.com/shop/product_thumb.php?img=images/iat-sensor2.jpg&w=300&h=267&wx=300&hx=300[/img]
http://www.bmotorsports.com/shop/product_info.php/cPath/129_142/products_id/400

So this will be a bit of a work in progress but wanted to make a start at it to see what I can find out..

Contributions welcome, once they're on track! (or just karma click if you like my stuff ;) )


Further reading:
http://zilvia.net/f/tech-talk/302418-intake-air-temp-ambient-temp.html

Here's some testing that I did on it last year:

Ok just an update on this as you have got me curious.. :devil:

I was at the dyno on Saturday pushing 19psi, and the intake temp only recorded 27c during a 3rd gear pull up to 7,200rpm. Basically I think this is :bsflag:

So I decided that I would do a little test in controlled conditions. Step 1, using a secondary IAT (ok, it's a cheapo but it gives a comparable baseline), I put it in vicinity of the intake pipe for a few minutes to let it localise to the temperature:


_KUNENA_BBCODE_HIDEIMG

As you can see, there is a bit of a difference, I figured that the engine based IAT (reading on the PFC controller) should be colder as it's physically mounted to the engine which is metal and the car hadn't been started that day.

Then I took a hair drier and shot it at my el cheapo temp sensor and the temperature went up over a period of about 30 seconds.


_KUNENA_BBCODE_HIDEIMG

Next I put the same hair drier in the inlet manifold intake, at the same pressure for 30 seconds and the temp did not adjust at all. So then I put it on hotter and more pressure for about 2 minutes and it went up by only 1 degree!


_KUNENA_BBCODE_HIDEIMG

So basically I think I have a problem, there could be an explanation however, so I will check with ATS on what they think should happen (seeing as I bought my PFC kit there).

Will keep you posted.

After this, I sourced a second MAT sensor from ATS however I did not follow up with testing it. I will be pulling the pipes on the engine at some stage in the next few weeks and will replicate the test with the new sensor to see what I can find out.

My feeling is though that this sensor is the wrong type or way too slow, as Robbie suggested that it was 'very hard' to get a good performing map due to the slowness of the sensor output.

Very good research - will def help anyone looking to map their 2.

KC

Courtesy of the all knowing Ishii in Japan, here is a chart showing the temperature sensor characteristics for a GenIII 3sgte:

_KUNENA_BBCODE_HIDEIMG

Ishii tells me that the coolant and air temp sensors are more or less the same, so this should be the resistance range we need to accommodate:

_KUNENA_BBCODE_HIDEIMG

And here is what I found on the Bosch high speed air intake temperature sensor. It's not exactly on, but it looks pretty darn close. So hopefully it could be sufficient for use with the Apexi Power FC...

_KUNENA_BBCODE_HIDEIMG

According to this website: www.allferrariparts.com/search.php?description=true

That Bosch sensor is used on the Ferrari 355,456 and 550.. so I guess hopefully it should be ok for high performance applications

Following all this research Quandry (very deserving of a karma click might I add) I assume you will be putting the temp reader on the pipe from the intercooler / chargecooler to the turbo and not on the intake pipe as heat soak will change the temp.

Now some mussing from me, if you are going down this road will it end in a FPR or variable resistor to control the injected fuel volume on each injector, wideband lamda on each cylinder's downpipe. That will then, all be used to extrapolate the fuel volume injected based on Lambda/turbo intake temperatures + some heat soak figures for each cylinder. If you go that far do you then need to calculate the height above sea level and get a reading for the air density and work that into the calculation as well.

Ok now that I have confused myself, is there a ECU out there that would be capable of doing all that work and driving/ controlling the variables at furious speed to ensure there is no knock, knock, BOOM as at this stage your sailing to close to the wind.

All the above post is green light thinking, and pays no respect to the known laws of physics, thermodynamics, fluid dynamics or common sense. And is possibly the reason I don't have a turbo as the boost would blow my mind senseless.

Give me a spark, twin 45 webers / air and I am happy out, but I really like your thinking on this item.

slipsliderg wrote:
Following all this research Quandry (very deserving of a karma click might I add) I assume you will be putting the temp reader on the pipe from the intercooler / chargecooler to the turbo and not on the intake pipe as heat soak will change the temp.

Correct


Now some mussing from me, if you are going down this road will it end in a FPR or variable resistor to control the injected fuel volume on each injector, wideband lamda on each cylinder's downpipe. That will then, all be used to extrapolate the fuel volume injected based on Lambda/turbo intake temperatures + some heat soak figures for each cylinder. If you go that far do you then need to calculate the height above sea level and get a reading for the air density and work that into the calculation as well.

PFC does all of this already


Ok now that I have confused myself, is there a ECU out there that would be capable of doing all that work and driving/ controlling the variables at furious speed to ensure there is no knock, knock, BOOM as at this stage your sailing to close to the wind.

All the above post is green light thinking, and pays no respect to the known laws of physics, thermodynamics, fluid dynamics or common sense. And is possibly the reason I don't have a turbo as the boost would blow my mind senseless.

Give me a spark, twin 45 webers / air and I am happy out, but I really like your thinking on this item.

I'm pursuing a faster IAT to eliminate the possibility of random knock on a hot day. Having said that, the tune I have is fairly safe so it's hopefully ok.

Did you know: your engine is 'knocking' to some degree all the time. It's the intensity of the knock that you need to watch. When dyno tuning a car what a tuner is looking for is:

a) that the A/F ratio is at peak torque level for the load/rpm cell
b) that the ignition timing has been raised up to the point where no extra torque is generated, and then backed off a couple of points

Then EFI sensors such as water temp/air temp/air pressure etc are used as correction factors to ensure that if environmentals change (i.e. you drive up a hill, or drive up a hill on a hot day, or drive up a hill on a hot day and your rad springs a leak) that the ECU will throw in corrections to compensate and dull down your performance to safe levels.

That is why I need a faster AIT sensor, as we're having to run a *safe* map because the one that is in there is not fast enough to cope with the upgrades and be able to adjust the tune if something goes wrong..

I'm using GM part number 25036751 (It's rebadged AEM, the 'fast' GM one), imo it's plenty quick enough - as I never even take timing off until > 40 degrees. It never gets anywhere near that usually - I have a much larger intercooler though.

Located - here:


_KUNENA_BBCODE_HIDEIMG

The stock location is shit, and useless at the same time.

Ok I just bought one of these:


_KUNENA_BBCODE_HIDEIMG

Celica GT4 RIAT Inlet Air TemperatureSensor

From this ebay seller: http://myworld.ebay.co.uk/fennedevelopments/

What can I say, the guy has a pretty good pitch. And if he's right that sensor has the correct resistance for the 3sgte...

This auction is for a new RIAT sensor, a rapidly responding version of the standard Toyota Celica GT4 (ST205 model) inlet air temperature sensor. The RIAT sensor replaces the standard Toyota sensor and is supplied with a new harness to connect with the engine loom. Heatshrink is supplied to make a professional installation.



This product:

*
eliminates heat soak- a weakness of the standard sensor
*
reduces the chance of detonation damaging your engine
*
improves economy
*
improves transient torque and power
*
improves your, or your tuner's, ability to map your engine closely - leading to further power and economy gains.

This sensor was developed for use on RX7's to cure a problem with the standard Mazda sensor. The same problem can be found on Toyota Celica GT4's and the fix is the same - replace the sensor with a modern open element type sensor.The sensor requires a thread adapter, supplied but not shown.



The sensor has the following resistance values:



Temperature Resistance

deg C Ohms

-10 9397

20 2500

80 323



The sensor can be fitted by a home mechanic and requires only the original sensor to be removed which is accessed from the top of the engine. No special tools are required, just a socket set and a soldering iron.



Read on for futher details...



What does the inlet air sensor do?



It sits in the airflow passing through the plenum, just downstream of the throttle bodies. This is the best place to measure the temperature of the intake air before it hits the injectors and the combustion chambers. The air temperature and its pressure are used, by the ECU, to calculate its density – and therefore the mass flow rate of air entering the engine. The correct amount of fuel is then added to burn with the air and create the combustion pressure that generates the engine torque. It is easy to understand that the ECU needs accurate data to determine the fueling, and that’s where the RIAT comes into play.

When the air is cold, it is dense, the ECU adds more fuel and the engine makes more torque, and more power. Cooling is achieved with chargecoolers, intercoolers and water injection methods. During stationary periods, radiated head from the radiator, engine block and turbo units heat the intake air, intercooler/chargecooler and eventually the whole intake tract. When this occurs to a standard sensor, it absorbs this heat and retains it for minutes after the rest of the system has cooled down. It is quite possible for the vehicle to complete a lap of a race track, or a 1/4mile run before the sensor accurately reflects the system temperature. During this period, the ECU has been fuelling the car incorrectly, thinking the air is hot – a lean condition that could lead to detonation. Tuners, and Toyota, compensate for this by running on the rich side of what is safe, which leads to poor economy and sooty exhaust during acceleration.

The RIAT will enable tuning closer to optimum, giving you greater fuel economy, power, and allowing real time management of water injection trigger points. It will also allow you to ‘see’ peak temperatures that may have been masked by the slower responding sensor. Do not be surprised if your car surges away from heat-soak situations as the ECU responds the inrush of cold air and fuels accordingly.

Finally, the sensor is an automotive grade component, designed specifically for this purpose. It will not fail and carries a lifetime guarantee. The harness is manufactured to suit, and is again automotive grade with the highest temperature resistance available. It will not melt in the heat of your engine bay. You will receive an installation document that explains how the component assembles to your vehicle, and how it is possible to check the calibration of the new sensor against the old, to ensure you are not risking your pride and joy, FENNE DEVELOPMENTS understands how important your investment is to you.





What is the problem with the standard sensor?



Two years ago I found that the inlet temperatures I was seeing (Apexi ECU with commander) with my RX7 FD3 (series 3 1992 onwards) did not tie in with what I knew to be happening. The inlet temperature should have fluctuated as I accelerated through the gears as the turbos passed through their range of efficiency. I was running stock twins, at 1.1 BAR, so they were running above their normal operating range. On inspection I found the standard air temperature sensor to be heavily shielded with a large time constant. This means that any change in temperature will not register quickly with the ECU. If you have wondered why your water injection, or a sudden increase in speed (and intercooler efficiency), does not translate into variations in inlet temperature, then the standard inlet temperature sensor is the reason.



What are the gains?



The graph (picture) shows the differences between the two sensor types when hot and cold air are applied. The RIAT sensor (in blue) can be seen to respond very quickly. Registering the initial step increase in temperature. At 20,000msec the Mazda Sensor is lagging the RIAT sensor by 15°C and fails to reach the system temperature before the heat source is removed. At 140,000msec the step inputs become larger and faster, simulating accelerating and braking around a race track. The RIAT sensor responds quickly, recording peaks of 45°C and lows of 10°C. The standard Mazda sensor is heatsoaked at this point and records a peak of 34°C and a low of 31°C. A clear demonstration of the benefits to be had by moving to the RIAT sensor.



Because the RIAT sensor responds so much faster, the car can be mapped leaner. Why? Because your ECU will now see the surge of cold air as the water injection kicks in or the air starts to flow externally through your intercooler, it is therefore no longer necessary to add a blanket rich map, to protect the car from heatsoak events. The RIAT sensor just does not heatsoak. At all. Why is heatsoak dangerous? When the standard sensor heatsoaks, the ECU thinks the air is 70°C. It thinks the air is hot and thin, with not much oxygen. As you accelerate away from heatsoak, joining a motorway for instance, the boost triggers the water injection and the intercooler starts working efficiently. Inlet charge temperature drops to 5°C. The ECU doesn’t know this yet as the sensor is still reading 69°C, 68°C, 67°C slowly catching up with reality. The engine is now breathing cold, dense, oxygen rich air. This causes a lean condition taking you closer to the detonation limit and potentially overheating your turbine blades in your turbo. Of cause this doesn’t really happen, your experienced tuner will have mapped your car rich because he knows that when he maps engines lean- they detonate and break rotor tips. So your engines safe, but drinks fuel and your friends tell you she blows black smoke during hard acceleration.



Is it safe?



Yes. UK tuner, Dragon Performance, has tested the unit and recommends it, commenting that “the difference is like night and day. The logs show it better than I can explain, the speed and accuracy is in another league, highly recommended.”



This sensor is manufactured by an OE supplier, has been used extensively for 5 years and has NEVER failed in service. Beware of cheap copies. These may not have been developed for the demands of the intake manifold system. Contact with fuel vapour, high temperature and vibration may age them prematurely.



Is there a fuel saving to be had?



I have calculated the theoretical possible fuel economy increase. I have used a generic temperature correction curve and assumed a worst case heat soak of 40 degrees C. To explain, the temperature correction curve is that which the ECU uses to determine how much fuel to put in, dependant on incoming air temperature. i.e hotter air=less fuel.

The heat soak is from my own experience, I have seen 70°C on a cool day, so 40°C is quite conservative. Others may have seen more or less?

The reduction in fuel for a 40°C increase in inlet air temperature is 6%. It can be argued therefore, that to guard against a lean condition, a 'safe' map would be at least 6% rich due to the sensor lag.



To conclude, a 6% reduction in the over-rich condition should be possible.

The fuel saving works out at £13.50/1000miles (20mpg, £4.5/gallon).



Why does this sensor make mapping easier?



When mapping a car you are altering the ignition timing and the fuelling at each load point. You examine the data output in four ways - knock, AFR (Air fuel ratio - lean/rich condition), EGT (Exhaust Gas Temperature), combustion pressure. With this information you can adjust the timing and fuelling to suit your requirements of power, economy, and drivability.

When rolling road mapping, the demands on the intercooling system are extreme as airflow may be down on normal road/track conditions. This can cause the inlet air to run hot and cold depending on your load state.

If the inlet air sensor is not following the actual inlet air temperature, the ECU will not be able to match the fuelling correctly and this will cause rich/lean conditions. This will actually make it more difficult to map, as you will need to wait longer at each load point for the sensor to 'catch up'. Straining your engine for longer than necessary, and increasing the risk of detonation.

It is actually easier to map a car with a faster acting inlet temperature sensor, as it is for any sensor. The better your input, the faster the ECU can respond and the steadier your outputs will be. Its important to understand that the ECU is using the inlet temperature to adjust the fueling automatically, all the time the engine is running. You will not be able to 'see' any difference unless you specifically look at air inlet temperature.



Its also important to note that if your car does heat soak, the ECU steps up the fuelling (at about 60-70°C) to reduce the detonation risk. This impairs fuel economy and makes the car sluggish, and the risk of plugs fouling is increased. This sensor eliminates the heatsoak at source. You may still see high temperatures, this will be due to an overstressed or poorly constructed intercooling assembly.



Is a re-map required?



The sensor is safe to use without a remap. It is an automotive grade unit and each sensor has it's calibration checked. There is a small risk that your own sensor has drifted out of tolerance and when your car was mapped, the fuelling was adjusted to suit this 'drift'.

The risk of damage is very small for two reasons:

1. A 40°C 'drift' is worth half an air fuel ratio. That means that if your sensor was reading 40°C cold, then replacing it with a correct sensor will cause your car to run 6% leaner (half and air/fuel ratio). If your sensor was this far out, your tuner would (should) have noticed as your inlet temp would have been -20°C in an ambient environment and replaced your standard sensor. If your car has not been remapped, the new sensor will correct any problems with your existing sensor.



2. Included in the instructions with the product is a simple method to check the calibration of your sensor. You will require a volt/resistance meter to do this, or have a laptop link/commander so that you can read what the ECU is 'seeing'.



This is a great product and those of you with Apexi Commanders and laptop links will see the benefit immediately. Don't be dismayed if you see new peak temperatures - they were there all along. Now that you are aware you can target them with upgrades to intercooling. If your car is due a mapping session, fitment of the sensor will allow your map to be leaner without risking the engine. Saving you fuel and giving you a punchy, tight response.

The RIAT sensor will be dispatched within 48hrs, postage is included in the auction price. The RIAT sensor is guaranteed for life. Should the sensor fail, simply return the unit and a replacement will be dispatched.

could you not just adjust your IAT calibration table and use any good sensor ?

Unfortunately not, the PFC does not have an IAT calibration table... it's hard coded.

Can't find much info on the web for the RIAT version ??

I emailed the guy to see if he has anymore - if not, I might get the GM version.

That sucks that you cant alter that data, but the AEM is the other end of the scale where there is TOO much adjustment!

If you want to get AITs down, I can get genuine AEM WI kit very cheap from Latvia.

Can get most AEM kit at a good discount in fact!

@Quandry - did you fit it ? working okay ?

Guy said he wll have them back in stock Friday

Spidey wrote:
@Quandry - did you fit it ? working okay ?

Guy said he wll have them back in stock Friday

Lol did I fit it? hasn't turned up in the post yet....

I have to see what fitting it has also, might be plug in for your manifold but unlikely to work like that on mine...

8 days / 6 working days .... maybe not too bad.

He said they were due in Friday but he has them listed already

So, fingers crossed you get it soon. Never checked the fitting - hopefully it will fit your as well.

Sensor turned up in the post today. Looks pretty good. It has an adapter piece on it that I guess makes it plug and play for the GenIII. By the look of it, it's possible it could be the right size for my GM sensor also.. Came with connectors, wiring and heatshield. Very nice. I'll run some tests on it this weekend with a multi meter.

Cool - he said they would be back in stock today, so must order mine

Speaking of ebay orders, when I got my custom plates, there was one white plate and one yellow plate (UK style - ugh). So I had to go back to the guy and order another white plate so that my car doesn't look like azzz with a blue/yellow plate on the back.

I never understood yellow plates. Does anybody know what (if there is) the reasoning behind it? I know its the legal UK spec, but why yellow? You could have a gorgeous paintjob and then this big dirty yellow plate. Looks sh1te! White is much more unassuming and subtle and tends to go with more colours.....


EDIT: Just noticed that I got a promotion! Senior boarder now WHOOOHOOOO, must happen after 450 posts. I was expecting something might happen at 500...cool.

I once had non standard plates on my car - they were literally 5mm shorter font.

The peelers saw me one day, and they literally HANDBRAKE TURNED in the road and spun the wheels creating a cloud of smoke to follow me and pull me over to tell me my plates were illegal. I honestly thought they were going to pull a gun on me they were that worked up.

They go buck mad over plates up here, buck mad - all because their precious ANPR can only read standard fonts.

Yeah its a joke the way they carry on. When I first started driving, my very first car, I put fancy plates on. They were perfectly normal and legal in all ways bar the font was italic, ie, leaded over a bit to the side. A bike cop sitting opposite me at red lights at a junction, saw me, saw my indicator and pissed off up the road I was going to go, ready for me. So, lights turned green, I went around the corner and there he was, bike parked SIDEWAYS across the road and the fat prick doing jumping jacks!!!

I pulled straight in, not a prob, never a days bother with the police in my entire life and the guy went totally, absolutely, donkey fucking mental on me. I swear I thought he was going to beat the living shit out of me the way he was going on. What on earth happened to that guy that day to abuse me like that over a poxy reg plate. BIG DEAL!! give me a slap on the wrist and ill stick the other ones back on, but I was so shaken I had to turn around and go back home. He screamed in my face for 20 mins and told me I was going to court, going to pay massive fines, lose the car, my licence. Read me the riot act. I was only young and driving a few weeks. So, never since have I bothered wasting money on plates. BOLLOCKS!! Oh thinking about it now makes me so mad...


EDIT: Quandry, I apologise, just realised we totally hijacked your thread..

No more on this, this topic is for Quandry's Air sensor.....sorry

Looks like good feedback from JimGTS on IMOC

IAT feedback

Yep.

I did some testing between my high speed and GM senor (using a multi meter) and the difference was about half a second in initial response time. However I would like to see what the difference is when you change about 20-30 degrees (i.e. a hard pull in 5th from cruise to lunatic speed)

_KUNENA_BBCODE_HIDEIMG

Note: after writing up all this stuff/testing etc... I found there was a problem with the pinning on my ecu. I have since rectified that and found the GM sensor to be accurate and fast.

I'm off to the dyno again soon so will try to do some monitoring on intake temps under load then.

I'm fairly sure that's the same GM sensor that is used with my Hydra Nemesis.