ok, for some reason AEM dropped the ball on the S2000's intake air sensor calibration table. they plug in the same values as all the older honda IAT sensors and it's not right at all for the S2000. we got a calibration table off of the AEM forums that people were using which in a small range was accurate, but overall not right either.

camuman and I got tired of not really knowing what's going with the temps so we decided to do a little experiment we extended the wiring to the IAT sensor so we could read the temps of some heated water/coolant side by side with a temp sensor. since we didn't have a way to start cold and heat the water up in the garage, we started with the coolant near boiling and cooled it with ice. kinda funny, part way through we ran out of ice, good thing there were some icey flavors in the freezer that nobody likes

it was a cold night (for floridians anyway) and by the time we got from the stove to the garage and set up the rig the first reading was at the 178 degree break point, we eventually got the mix down to the 36 degree break point. so we now have an accurate temp table from 178 degrees to 36 degrees. we graphed the results below, the difference between the real results and AEM's is pretty aweful.

sensor out, wires extended...


tools...


rig is working...


the rig in all it's glory


the temp probe next to the IAT...


adjusting the sensor calibration table as we go...


yeah so we ran out of ice, good thing we had flavored ice things in lemon and orange, nobody likes those flavors anyway


almost as low as we could get the mix...


ha ha, little water, little coolant, some ice, and a bunch of flavored ice


here's the result, blue line is what we ended up with, red is the one that was off of the AEM forums (it was actually ok in a small range) the green is AEM's, WAY off for the S2000's sensor...

 

wow...wtf was AEM thinking? looks like the AEM forum did it backwards.

good job. love the use of kitchen supplies, ice pops, and Craftsman meter :D

 

Haha hell yeah! I can't read the graph on my blackberry but it looks like they messed up quite well. Ill be sure to use these values on my startup cali.

 

Dam, I want some popcorn shrimp now!

 

On a serious note, it would have been really nice to test more then one sensor in the same conditions. I mean obviously this is perfectly accurate for your sensor. But with another sensor you could have really dialed in on a good table for everyone. I'm sure all the sensors are going to have some forum of variance.

Great Work

 

ha ha, where were you the other night! that would be ideal, when I get the AEM set up in my S2000 we'll have to compare the sensors and see if there's difference. I'm sure there's +/- 1 or 2 degrees from sensor to sensor.

 

now to fix the coolant temp table, that base one isn't right either. again there's one floating around on the forums but we don't know if it's 100% right.

 

just a little update, it's pretty interesting to see the difference and to see how much the intake manifold heats up the air going into the motor. here's a couple of things I noticed on my bone stock motor (just installed the AEM EMS).

using a temp probe once the car is heat soaked the manifold goes from about 125 degrees near the throttle body and the first runner to about 145 degrees near the firewall. so no matter how effective your intercooling system is the air is basically passing through a 135 degree (average) oven on it's way into the combustion chamber. yes the sensor on the intake pipe before the throttle body will show a lower number, but that's before going through the "oven".

we noticed similar results on camuman's car even after the hondata gasket, it just took longer to get heat soaked. I noticed the same thing on my supercharged and intercooled CRX with the hondata gasket. I'm going to test and see what the under hood temps are on the S since I think that's the culprit. in my CRX the temps never went under 130-140 degrees near the manifold once it was heat soaked even on the highway.

even though the manifold is so hot the IAT sensor will get below that temp when cruising and especially full throttle (all-motor) when fresh air is going through the manifold. cruising I'll see about 130-140 degrees light throttle, floor it and the temps start to drop. camuman's car all motor on the dyno and the temps would drop about 15-20 degrees during a dyno pull.

engine braking starts raising the intake temps. I believe it's from the strong vacuum in the manifold and reversion from the combustion chamber back into the intake manifold. I also believe this (along with the under hood temps) is what starts to heat soak the manifold even with the hondata gasket.

the manifold heat soaking is inevitable. all-motor or staying out of boost and cruising the temps will drop under cruise conditions but come to a light and there's not enough volume of fresh / cool air to combat the intake manifold. if you've still got the stock air box put your hand in the intake vent at idle, you'll hardly feel any air flow. now blip the throttle and rev it up, you'll feel a good amount of airflow, that's what it takes to start to cool the manifold.

also the sensor does not react slowly. on my all motor setup it will start to lower when you go full throttle and start to rise when engine braking. looking at AEM logs just looking at the air temp I can tell when I got on or off the throttle. withing a half second of flooring it on the dyno the intake temps started to drop and within a half second they started to go back up. part of the delay is just the increments the sensor works in, it's in steps, not a smooth line. on the all-motor dyno the temps started at 140 degrees and ended at 125, but instead of a smooth line to see the changes instantly it stepped down in 1.8 degree increments. example, 140 - 138.2 - 136.4 - 134.6 - 132.8 - etc. within two seconds of letting off the throttle on the dyno and engine braking the temps shot back up from 125 degrees at the end of the run to 145 degrees! 5 degrees hotter than when we started and 20 degrees in 2 seconds! when the sensor was out of the car for our tests you could blow on it and watch the temps raise and then drop back to ambient. I'm sure there's a limit to how fast it can change but it's at least 10 degrees per second

on my supercharged CRX there was an air to water intercooler built into the manifold. on the dyno the intake temps would start at around ambient (because the hood was open) and end up around 30 degrees hotter. at the end of the run instead of the temps shooting up like the all-motor setup, they would hold at the peak about 5 seconds then start to decline again. in this case I believe that the intercooler is doing it's job and keeping the temps from going up during engine braking like it does with all-motor. turbo car with a front mount and the little bit of air going into the motor will be cool fresh air and I'm sure it'll show the same kinda thing. Tony mentioned seeing the temps hold for 10-12 seconds after a run then drop on a turbo car. since that's on a front mount car that makes sense to me after looking at the other results, only a little bit of that cool air from the front mount is going to get into the motor and mix with the hot air in the manifold.

there's the argument as to why honda moved the sensor on the DBW S2000's out of the intake manifold to the intake pipe and that it must have been "wrong" before. I don't think it's because one is "right" or the other is "wrong" they're just different. if you go back in history honda has moved that sensor between the two locations a lot of times. the 99-00 civic Si's had the sensor in the intake, then the S came along and they moved it again to the manifold. my personal guess is that it comes down to cost. just looking at how much they changed on the manifold for the 06+ S2000's it's gotta save a lot of machining and cut down on costs (looks better too ) to have it the way it is now. either one will show you the temps raise or lower under load and let you know if your intercooling is keeping up under full throttle conditions. my only argument would be that the intake sensor on the manifold is that much closer to the intake runners and that much closer to the temps the actual combustion chamber will see after passing through the "oven". and it's been shown that light load conditions lead to much higher intake temps so tuning light load and compensating for changes in intake temps will be easier / more accurate with the temp sensor on the manifold.

anyway, just showing what I've noticed on my S so far that also pretty much lines up with the results from my CRX, just sharing info

 

Will adjusting the settings, affect your tune?

 

these settings shouldn't change any part of your tune with the exception of cold startup. with the base map air temp corrections it will add fuel under cold start conditions to richen up the AF ratio until the car warms up. once it's warmed up there's really nothing that the air temp effects tune wise unless the car starts to overheat, then it'll start adding fuel again. that's going by the base map, your tuner may have changed that but I doubt it, it's not something you mess with too much tune wise besides cold start up.

if you go to fuel -> advanced fuel -> fuel trims -> fuel trims you'll see all the different fuel trim tables. one of them will be the air temp fuel correction table, by default is adds a little fuel starting below 86 degrees but leaves the rest alone. ultimately cold start is more based off your coolant temp than air temp. we did this more to see what's really going into the motor than to effect the tune.

 

So you are saying we shouldn't touch the IAT sensor cal table below 86 degrees?

 

nah, I'd replace the whole sensor cal table, then it's showing the right temps across the whole band. if your tuner didn't change the base map settings on the air temp fuel trim then it's not changing a whole lot. at 68 degrees it's adding .78% fuel, at 50 degrees it's adding 1.56%, at 32 degrees it's adding 2.34%, it's not a lot of change, I wouldn't worry about it. plus the intake temps are not in that range very often.

 

i just installed my ems on my bone stock ap1.

i am using your table and ill be checking the coolant temp sensor soon.

i have not done this but an easy way to check the coolant temp sensor and ait, they should be within a few deg when the car has been sitting over night and things are nice and cool.

 

Thats a good point.

I've recal'd both my iat and coolant temp tables so far.

 

i checked yesterday and today, after the car has been sitting a while the cts and iat read the same. this is with your ait cal and the aem supplied cts cal

its crazy how crappy the 01 ecu ran this car... i can actually drive it below 3000 rpms now and no bucking in stop-and-go traffic