IAT sensor location Vs tune accuracy?
 

Okay I'm getting my ali sensor bung welded in tomorrow so I'm removing the intercooler outlet pipe tonight and need to let the welder know where exactly to weld the bung for the sensor.

As mentioned in loads of threads on here, the GM IAT sensor is prone to heatsoak if mounted too close to the throttle body. So I've thought about maybe relocating it to the area of pipe exiting the intercooler outlet in a similar fashion to Braineack's set-up.

The question is, will there be any disadvantage to having my sensor located in position B - will it be an accurate indication of the air temps entering the cylinders with regards to dialling in an accurate fuel map or should I go with position A and risk heat-soaking when hot-restarting?
Will my tune accuracy be less accurate in position B over position A? I can't imagine temperature differences between the two loactions would be more than a degree (Celcius) or two at most.

Attachment 196373

The main problem for me is that the TDR intercooler pipe goes right over the top of the rad and also sits quite close the chunky radiator top hose. Switch off the engine and all that radiator heat rises directly into the intercooler pipe. This means that in position A, the sensor will be prone to alot of heatsoaking if left for 10mins after a good hard run. I'd guess with an around and behind the rad intercooler set-up, the risk of heatsoaking the IAT sensor would be less than that of the TDR design.

Any thoughts anyone before I get it welded up?

The air is moving to fast to measure a significant difference between point A and B. B is less prone to heatsoak, but not by much. The IC gets a lot of radiation from the radiator, so it'll heat up just as well.
I have my GM sensor in A and it heats up to 45°C in matter of minutes.
I'm gonna replace mine with a plastic sensor with low mass hoping that it heats up less.

When I first installed my system I put it in the bottom of the crossover pipe about 3" from the upper radiator hose and got epic heatsoak.

I would recommend getting it out of the engine bay if possible.

i put mine on exactly where your A is and i have that mp62 set up. no heatsoak problem, read ~ outside air temp when just cursing (showing how nice that tdr kit works)

So if there is no significant difference in temps between point A and B and given the extremely short pipe run on the TDR set-up, would welding the aluminium bung into the wall of the intercooler outlet be the best option? i.e. keeps the sensor outside of the engine bay but also provides accurate intake air temps that aren't artificially too low?

The outlet wall of the TDR IC is probably only a few more inches away from position B too.

If I were to do it again, I'd install mine as low as possible in the endtank.

B or bust.

Those location would be the equivalent of my MS metering >30°F above ambient after being parked for a few minutes or just 5°F above ambient.

Thanks for the replies.

I've got my 3/8npt aluminium bung and from the info you guys have provided and also from searching other threads, I'm going to have the bung welded into the intercooler endtank.

I let the engine bay heatsoak today and left it for 20 minutes before popping the bonnet and feeling the pipes. Both positions A and B were identical in temps, probably because it's such a short run of pipe. I then felt the intercooler endtank and it was a helluva lot cooler.

Your experience and testing of different locations has saved me a load of hassle.

Cheers guys!

Here is where I had it located:

Attachment 196330

Attachment 196331

I got it all set-up tonight and took it for a run. Engine starting and both coolant and IAT sensor display 23degC which is nice. I idled the car up to a coolant temp of 85degC and my IAT went up to 25degC which is cool. A quick blat and IATs rise progressively to 28degC. When I calm down, the IATs roll off towards a near ambient temperature again.

I then heat-soaked the engine on the drive and went in for some tea. I came out 20 mins later and the IAT sensor heat-soaked slightly to 36degC. I started her up, and idle AFRs were lean (~18:1), so applied some IAT related correction around 30-45degC until my AFRs dropped back to a sane figure. I then applied the IAT related correction to start to reduce at 500rpm and end at 2500rpm. I'm guessing this is the correct method of tuning heat-soak for hot re-starts? The second air hits the FMIC, the temps are going to drop off rapidly anyway.

The issue is not the reading while the engine has been running, the issue is that when the engine is turned off while hot, then restarted a few minutes later, the sensor will show much higher temps.

FWIW the way the AEM does it, it has 2 tables:

1) % Enrichment vs. coolant temp, and
2) for for how long (time-after-start), vs coolant temp.

The idea is that it will enrich the fuel for X seconds after startup, depending on coolant temp, and by a % that also depends on coolant temp.

It's not perfect, and I also want to relocate the sensor away from 'A' above.

I like the intercooler end tank idea but for the hassle ...

It seems lots of other EMS developers have similar ways of dealing with heat-soak. I think the GM sensor is the main culprit regarding the MS. I wish MS1 had a time delay after engine cranking feature (similar to the way ASE can be tuned). That way, on track IATs that briefly peak abnormally high will mean you don't enter your higher temp bins which solely serve to prevent lean hot re-starts when you use the car in your day-to-day life.

MS-II and MS-III both have ASE taper curves based on CLT temp.

http://msextra.com/doc/ms2extra/Gene...2_ASETaper.GIF



You can also reduce the amount of the software's built in air density correction values (MAT Correction Value) so that it gives less aggressive corrections when the sensor might be heat soaking. This way you don't need to fool with the 'MAT corrections table' to combat the ECU pulling fuel after a heat soaked start.

Right. I'm still having problems with my VE table holding true cell values and it's driving me nuts! I need to tune every day in order for the car drive smoothly. I think it's still to do with the differences in ambient temps and possibly because Gair might not be compensating correctly.

Below is an image of my fuel map that drove great at 15degC this morning. Tonight on the way home in warmer weather (ambient temps 23degC) and my AFRs run slightly leaner and autotune is run and pulls the VE table back into line and the car progressively drives wonderfully again...

Attachment 196001

What the hell do I do next? I've turned off all forms of IAT correction, the IAT sensor isn't heatsoaked as it's in the intercooler endtank so it's got to be my Airden factor or something similarly related? If I don't get it sorted within a week, there will be an MX5 on fire in my driveway, I've had enough of this shit.

I can post my MSQ, logs and ini files if required.

turn on EGO, those changes to your table are pretty minute.

^^^ Word. My car runs pretty nicely, though if I ran autotune every single time I drove the car it would probably want to adjust a point or two up or down occasionally. That shouldn't cause the car to run weird. What I would do is keep the richer table values and let EGO lean it out a touch where necessary.

BTW if your table is pretty well dialed in you should set auotune's cell change resistance set to "very hard."

Yeah Im set to "hard" at the moment with max cell percentage change and value set to "15". EGO is also on with a max value of 4, step size of 1 and set to a quick to react value of 20 cycles. The trouble is my VE table values are dialling in a value of +/- 5 in some of the cells when temps change, that is hardly well-tuned.

Ill give "very hard" a go for a few days but surely it'll just tune to the ambient temp at the time?

Should I lower the cell percentage and max value even more than 15, say 5?

When it's on "very hard" that means it takes more "hits" (data points) to convince autotune to change the value. The more data it has to sample before making a change, the more consistent the numbers will be. My experience with the old MLV and now with the live VEanalyzer is that "hard" is still kind of flaccid, that's why you're seeing it want to make +/- 5 point changes.

You shouldn't need to lower the max values or percentages. Doesn't matter because on "very hard" autotune never wants to change more than that anyway, except for maybe the first one or two times you use it. I think I just leave mine on something like 10 and 10.

For your EGO settings, 4% authority is probably okay. I leave mine on 10% but EGO never uses all that authority because my VE table is pretty good. Also if you're cruising at 4000 rpm and EGO is allowed to make changes every 20 cycles with a step size of 1 (that's 1%) it could be adding or pulling a percent every 0.3 seconds. It might smooth out fueling if you increased the cycles to 60-80-ish. The more dialed-in your fuel table gets, the more slowly you want/need EGO to be adjusting things on the fly.

i have ego at 15% :)

I dont really have issues with driving around in different temp levels, it idle that gets interesting, where you really need to tune the AIT corrections table well.

@Brain - what do your actual EGO correction settings look like with MS2/3?

Thanks!