s2000Junky

You can program the e-manage to do that for you. When you hit your designated hot IAT target, you can pull what ever timing you want. Better then detonating your motor. You can make fuel corrections in this scenario as well.

s2konroids

I'm pretty sure it does (i've recall reading it somewhere too), I also think it would be odd/not good to not too pull timing once you reach a preset temperature determined by the IAT temperature. I cannot remember for sure but when using an OBD datalogger and log generated by the GEMU i *think* it did when goes over 45c/113F - all tuned and used on a dyno.

dibi

I just sent an email to Jeff Evans with some questions concerning my upcoming tune. This should help answer your question

Question
-How do you calculate/adjust for the change in outside air temperatures throughout the year when tuning a boosted car? (Since it is Feb and the air is much denser/colderthan it will be in the summer ) Will I need a retune when it’s warmer?

Answer
The air temp compensation maps account for difference. The air temp sensor should be the GM fast response and moved into the turbo inlet. This will more accurately account for ambient temps.

s2konroids

Thanks good info for folk, I already knew that depending on ECU (has different table/meaning for the same thing) personally I have only tweaked a flashpro, tuned a GEMU and tuned a DTA S60 full standalone.

s2000Junky

I have mine set to start pulling timing at 156F

I see 80-125F average operating range from cruising to sitting in traffic. I set it to pull timing at a lower IAT before and found it was reflected on my data log, so it does work.

mase1

Thats not the best solution, especially for a car that sees a variation of intake temperatures (ie. Road Racing).

Ideally, I like to have them in the intake stream, just before the throttle body, but it really varies on the application and vehicle. For most street cars, we usually leave them in the intake manifold.

Then comes the ECU parameter traits. I've asked Devon at AEM to see if we can get the software to have some type of load based IAT correction, as that would be helpful for idle/start up heatsoaking issues. The cars will run lean if you have parked your car for a few then restart due to the heatsoak/atomization/absorption. so most people will notice right after they restart it runs lean for a bit. Well on a normal application, as intake charge temps raise, you want to pull fuel out due to less dense air. If you dont have a load based temp correction as well, this obviously wont work.

Being that I do a tremendous amount of road racing work, I've gotten a lot of experience tuning for these broad spectrum details. It takes us a good 15-20 minutes of hard track time before the engine temps/cooling temps really start to settle in. I can sit on the dyno a long time, but it never quite is the same as hard road course conditions.

The heat rejection of the intercooler system really plays a huge part, and depending upon the setup, the efficiency of the intercooler setup plays an upmost important role when it comes to intake temperatures. If you have the intake temperature sensor at the turbo inlet, yes its reading ambient temperatures, but thats not telling you whats actually going into the motor.

IE. Prime example: we constantly track two higher whp NSX's that are singled turbo charged. On the street say 80 degree ambient, you might be lucky to get the intake temperature about 100 degrees if you're really leaning on it. after 20 minutes of track time, it will sit usually around 135ish once its reached full temperature and heat rejection. we have to pull a good amount of fuel out between 'street' intake temps and 'track' intake temps. So if the sensor was at the turbo inlet, we would never been able to fine tune the engine correctly for 'track conditions'

for 75% of you guys that will never see these track conditions, it doesnt play as big of a role. We can combat the lean condition right after start with some extra start decay. As for the IGN timing correction, we do typically pull timing, but the factors really depend on the engine constraints. IE. compression, octane, overall temps we will be seeing. ETC ETC.

and the last FYI, of all the Grand-AM cars I work with that are FI, we use intake temp sensors before the intercooler and then right after, we derive our correction factor on a number of things, but majority is the post intercooler temp sensor before the throttle body in the Bosch MS4 ECU.

s2000Junky

good stuff

dsddcd

Quick Answer
The best place I found was in the post intercooler before the throttle body and out of the air stream of hot air from the radiator.

Short Answer
Lets establish what the temperature sensor does first. It is placed in the air stream be it intake or manifold in order to trim the fuel and ignition. As the temperature rises the air density does also so less dense air mean less fuel to maintain a 14:7/1.0 or stoichiometric burn. Additionally when air temperature rises the cylinder also heats up and timing must be pulled to prevent detonation. The hotter the cylinder the faster the combustion happens in the cylinder.

As we know the turbo heats the air through both the heat of compression and conduction, then passes through the intercooler to get cooled back down but it can never get back to ambient temperature w/o a intercooler sprayer or air to water with a chilled water reservoir. This entire process happens VERY quickly when accelerating in fact it would be pretty common to see temperatures off the turbo of 200-300F or more depending on the boost pressure. So with an intake mounted sensor it may be hard to measure the temperature so quickly do to the heat soak from conduction of the engine. Secondly you want to try and keep the sensor from seeing heat soak from the ambient air.

Now for the sensor, as mentioned above it does need to be VERY fast acting. I would quantify this as able to rise 20-30 degrees in the air stream in a matter of a second or two. Unfortunately I have yet to find such a sensor so I chose the GM screw in air intake sensor commonly used on the older models.

So now that we have a sensor how do we use it. Traditionally there was only one sensor and a simple table to enact fuel and boost trims based on the temperature. This works great for N/A cars but for supercharge is not the most efficient way to do things. To be done correctly there really needs to be a delta trim function. By delta I mean the difference between temperature A & B over a fixed amount of time. A perfect example would be if I were boosting and saw a 10F rise in temperature over a 2 second period I would know that the intake temperature was quite high. The problem is there is not a sensor that I know of that can read the actual temperature so all we can so is assume that if the sensor we have chosen see a large rise in a short amount of time there has to be a big temperature differential between it and the air. This same type of control is used by the throttle but is quite complex.

Now that was said, what does moving the sensor father up the line do? Well it adds a little bit of horse power, that’s it. With a more conservative tune you will always account for the extra heat and pull timing. The reality is for most us we could see a bit of extra power as the intercooler efficiency increases but this would be part time and may only get 15-20WHP.

So the moral, get as close to the actual temperature as you can and play it safe.

dibi

this is a great thread!

spectacle

I've done some more research and this topic goes a lot further than I originally thought. I've uncovered some facts I'd like to share.

On the OEM IAT sensor - it is slow by DESIGN:

And here's the money maker:


Bottom line - the stock IAT is slow to begin with, and will give you even worse, inaccurate readings sitting in the intake manifold where it is exposed to heat soak. Moving the IAT sensor to the charge piping is the best compromise between getting an accurate temperature reading, albeit a slow one. In practical terms, it means sitting at a light and during light loads, the temperature logged or shown by a factory based ECU will be accurate, but during WOT pulls the temperature reading will be at least, in the best case scenerio, about 10 secs old.

Why is this important? Here's another quote:


Remember, the S2k from the factory is a N/A car whose air temps don't change as fast as a boosted car (hence the slow spec'd facotry IAT sensor). But when you introduce a turbo/SC and intercoolers in the equation which can swing the temperature readings very fast, its easy to see why the stock IAT sensor wouldn't be up to the task of doing what they were originally intended - compensate the fuel tables for the change in temperature. This is why boosted cars tuned in the summer run different in the winter, and vice versa.

So how do you fix this? Well, if you're like me and you have a factory based ECU (KPro), at the moment you can't. Hondata does not have a scalar in the software for inclusion of a faster reading IAT sensor like the GM IAT. If you wire the GM IAT, or any other IAT sensor to a factory based ECU, the temperature readings you get will be WRONG. Why? From the same thread linked above:

If you're on AEM or another standolone with provisions for a faster reacting IAT sensor, I'd highly recommend moving it to your charge piping and using something like the GM IAT or this: http://www.t1racedevelopment.com/ind...mp-sensor.html

All quotes posted by Legion on Honda-Tech, pulled from this thread: http://honda-tech.com/showthread.php?t=2240305