AIT Correction Table

Hi there,

I already asked some quesitons about AIT correction on the last forum because I had problems with getting my AFRs stable in different weather conditions.

We got to the point of the gas law and all the details how to do the map (+2,5% and -2,5% for each 10C of temperature change below and above 20C). I hoped that my issues will get solved with a table set like this but sadly I still cant get the AFRs to be stable.

My problem is, I am driving the car at many different conditions, ambient temperatures and altitudes (from -15C up to +40C, from 0meters up to 2000 meters altitude above sea level). So I really need a AIT correction that will work perfect. My car is a Cooper S Supercharged with the intercooler on top of the engine so it gets heat soked very easy and also the AITs are moving a lot and the deltas are big, that does not make it easier.

I am in the tuning process right now and found strange things, I just cant get my AFR stable when the AIT changes, the -2,5% for each 10C seem to be too high.

Today we had around +18C outside temp, giving it WOT the AIT climbed from around +25C to around +50C, in that area the AFRs are stable. But once I let the car sit for some minutes the AIT goes up to around 50C and when I give it WOT then, the AITs climb up to around +70C and then It seems that I am getting way too much AIT correction (-9%) and AFRs go lean (around 0,2 to 0,5 AFR out of target).

What should I do to solve this? Should I make the AIT correction numbers lower in the zones where it seem to bee too much? or will I get problems this way in the other direction (that I wont get enough AIT correction) when in summer the outside temps will go up to 40C and I will reach 70C of AIT in normal WOT conditions withouth heat soak, so even higher with heat soak?

another thing, in idle with heatsoak, I need to add fuel above 40C AIT correction in the idle zones because othervise the AFRs go way too lean- at 0% correction above +40C AIT I am getting around 15,5 AFR, at +20C AIT my idle AFR-s are at the normal 14,2 that I am targeting. So I need to add around +2,5% correction at the +40C AIT and even more at +50C to have an idle AFR of around 14,7. But the other problem then comes when driving away, as soon as I press the throttle i get to the 20%TP zones and above where the correction is again -5% or even -7,5% and my afr are very lean (16afr) and car nearly dies and has no power on drive away from the standstill. It is really annoying to drive like that because after each pause the car sits parked for some minutes and I drive away it looks like I cant drive the car, but my problems are just in the MAP. :) should I make the AIT correction lower in the 20% TP and 30% TP zones or what should I do?

would appreciate a good discussion about this problems, this is really driving me crazy that I cant get it to a stable AFR, so I would be very thankful if we could together find the right solution.

Hi,

were are you located in Austria? If you're near Vienna, you can come by our shop and we could try replicate the problem on our dyno.

What ECU are you running?

cheers,

Andy

Hi Andy,

I am from Klagenfurt, but my MINI I have in Slovenia. So I am not often in Vienna with it. Where in Vienna are you?

ECU is a Vipec plug n play for the MINI.

I can replicate this on the road withouth a problem, just need to let sit the car for 5 min and then give it WOT right after. So for this problem the AIT table is the cause I think because it goes too lean when it corrects to much fuel. This AIT correction now really drives me crazy, I need to know if I should adjust the AIT table and make the corrections smaller at higher AITs or will this be work in the wrong direction and the problems will be then on another side?

would be nice to hear some ideas what else to do that I could try.

Our shop is located in Unter-Oberndorf, which is about 25km outside west of Vienna.

Anyways, I assume you had your IAT correction table done before you tuned the fuel map? In anycase, I'd set this up as a 3d table against load, and set it accordingly.

Also, your setup would really benefit from the charge cooling coefficient option... if you've tuned in VE mode.

hope that brings you on the right path,

Andy

That's right KG ... the AIT compensation map should be configured before any tuning is done or you will be just chasing you tail with AFRs.

Where is the AIT sensor located on the engine? If its located in the intake manifold then heat soak will occur at a rapid rate and this is something you just cant avoid.

As KG suggested its best to setup a 3D AIT compensation map and I would use throttle position as the load axis. At 0% TP across most of the AIT temp ranges (heat soak temps) I would do no compensations for those cells.

Unfortunately this means retuning your fuel table but you shouldnt be far off.

Hope that helps

Here in NZ we don't have the wide range of ambient temperatures that you are seeing, although our new location is currently experiencing snow and sub zero temps so that may change.

IAT correction 'should' be fairly simple and the air temperature (or more correctly how the temperature effects the air density) is taken care of inside the ideal gas law. That being said it's important to understand the implications of the ideal gas law. What we are worried about is the charge temp as it enters the cylinder, not the air temperature in the inlet pipe or inlet manifold. Depending on air flow and engine rpm, the intake charge can draw a lot of temperature out of the intake manifold and cylinder head/valve and hence this can skew the actual air temperature from that measured by the sensor.

As others have suggested, it's worth using TPS as a load input for the IAT correction table and reducing the amount of influence the correction has at low throttle openings. If your ECU is the newer i-Series Vipec you could use the charge temperature estimate function instead of the normal IAT correction table which bases charge temp off a 3D table of IAT, ECT and rpm.

One more question that is important - Are you 100% sure that the IAT sensor calibration you are using is correct? If this is wrong it can understandably have a large impact on the accuracy of your tune and may explain some of the problems you are seeing.

I'll add that I had similar problems with the IAT correction on the Motec M1 in our 86. I found that the control seemed too aggressive. At idle and normal IAT the mixture would hold target accurately however with the bonnet down or after a hot restart, the mixture could be up to 10% lean. I fixed that by fudging the IAT calibration table to reduce the control the IAT had over the mixture. Obviously that's a fairly ugly workaround though so more recently I've added a charge temp approximation function in M1 Build. I've found that this charge temp approximation table coupled works much better and offers very stable control. I ended up weighting the result of the table quite heavily towards ECT in the idle area and the upshot of this is that the charge temp approximation table also plays a part in the cold start and warm up enrichment areas of operation (when ECT is significantly colder than during normal running).

Thanks for your replies.

Yes, the IAT correction Table was done before the tuning and I am using the 3D table with TP on one and RPM on the other axis so this is done correctly.

The IAT sensor is in the intake manifold just before the charge air enters the cylinder head, it has to be there because the TP is before supercharger, intercooler after the Supercharger and after the intercooler I have the methanol injection injector (water/meth injection was off when getting those problems) and then there is the intake manifold with the IAT sensor. The whole system is from aluminum.

On cruise I am using either no correction and on higher intake temps I have to add fuel on 0 and 15% TP where heat soak occures. That way its working ok on cruise and then as well the Closed loop controls the left over difference of AFR very well. So on cruise I dont have many problems for now.

My problems are happening on Wide open Throttle (100% TP).

Yesterday we had 20C ambient temp, my AFRs were spot on up to 60C IAT where the ECU has trimmed -10% of fuel. It was working great. But the day before when we had just around 10C ambient temp, the problem with lean IAT happened when i started the run with a slightly heatsoaked engine (at around 40C), and then it seems that the ecu is getting a too high value of Airtemp or that the air that enters the engine is colder than the air temp measured and the ECU IAT trim seems way to high.

The IAT sensor should work fine, it is the stock sensor that works on stock ECU cars perfect.

Andre,

How can I check that the IAT sensor calibration is correct? where can I find this? I just find where I can select the sensor in vipec, this is for Inlet Air Temp stated as Std Bosch NTC and it should be preconfigured by vipec as part of the base map.

I dont have the i series, so i need to find a solution with the functions of the standard vipec v series.

But what is making me headache here is that the correction seem to work fine when the outside temps are 20C where I tuned the engine, but once the ambient temps get colder and the engine gets some heatsoke everything gets out of borders.

But on one 100% TP pull through 2nd and 3rd gear yesterday happened another strange thing. i got a AFR rich down to 10.5 (target 11,8), the IAT trim there should have been -9,5% but it seems that the ecu there just did not make any fuel trim there, in the logs it says -9,5% but the AFR is rich like there was no IAT trim done.

This IAT fluctuation is really driving me crazy. I am slowly starting to believe that there could be a fault in my ECU and its doing problems? Would something like that be possible or did someone ever had heard about something like that?

I am adding the download link to my log file where you can see normal runs and than once of a sudden on the last wot run like there was no IAT correction done:

https://drive.google.com/file/d/0B_bF6LsjEAivWUJQTDY4WW9rSkE/view?usp=sharing

What is actually sounds like to me is that the AIT sensor isn't reacting as quickly to conditions as it could do. Do you know if the sensor has an open element or shielded?

this is the sensor:

http://www.1320.co.uk/content/images/thumbs/0001224_inlet-map-t-map-sensor-r53.jpeg

http://www.1320.co.uk/content/images/thumbs/0001221_inlet-map-t-map-sensor-r53_300.jpeg

Just a thought but how old is the supercharger? What size pulley is on it and do you have a larger intercooler or stock?

supercharger is giving the same amount of boost all the time, the pulley is smaller but not too small, the intercooler is bigger and it is doing the same cooling work (same deltas) all the time. Fueling is tuned for higher boost and the hardware on the engine. for this cars it is normal seeing high IATs, therefore it is very important that the ecu corrections are working perfect.

I know this has little to do with the IAT correction tables and how they're working, BUT from your posts it looks like the supercharger is pumping hot air.

so either it is out of its efficiency range or the intercooler isn't cooling the charge properly (WOT and IATs climbing).

just throwing out ideas, if there's maybe a way to improve your setup.

Andy

The IAT sensor pictured isn't one I'm familiar with. Link/Vipec should be able to advise which calibration to use.

In theory it shouldn't matter if the ambient air temp is -20 degrees or 40 degrees C - If the air temp entering the engine is say 60 degrees, then that is what the ECU needs to compensate for. However it seems that this is where your problem is - The variable IAT coupled with heat soak. I do agree with Chris, it sounds like perhaps the response of the sensor is quite slow. Even with a heat soaked manifold, a burst of full throttle should introduce sufficient airflow to almost instantly eliminate this. If the IAT reading doesn't keep up then you would see a lean condition.

What you could try doing which may be an effective fix would be to add an ambient temp sensor to a spare AN T input and use this as an axis for a 4D fuel comp table. It's a workaround that shouldn't be necessary but it may be a quick solution to a frustrating situation.

The OEM ECU will retard ignition when AIT is over 50c. Whilst not directly effecting the heat from the charger it does encourage the driver to back off the throttle thus slowing the revs to the charger and introducing colder air to the intake.

Calibration for the sensor is as Standard Bosch.

Thanks for your replys. I am as well retarding the ignition as soon as the IAT gets over 50C. This usually does not happen once I use the Water Meth injection. :) But I want the map to be safe both, with and withouth Water Meth just like on the stock ecu.

Like Andre mentioned, hot air from the supercharger should have nothing to do with my problem, the ECU needs to adapt to hot air and adjust the fueling correctly. My hardware setup is great and the IAT-s are much better than on the stock car, this engines run with very high IATs as well in stock form.

To get back to topic, do I need to wire an aditional sensor to the vipec for Ambient temps or would the original ambient temperature sensor be enough and possible to connect it with vipec? on the stock ecu this sensor is connected to the ecu and readable through the obd port, so maybe it is as well connected with vipec already because it is a plug n play ecu?

This with the ambient sensor seems a good idea, because like that I could let another table add or take away some fuel just when needed. I also thought using a 4D table with Engine Water Temps, I am running a cold thermostat so my normal when driving water temps are at 82C, in heatsoak the temps go up to 98C where the fans kick in and cool back down to around 92C. Will try to play with such a map if needed as soon as it gets warmer again outside. :)

I retuned the car last week at 20C ambient temp and the AFRs were spot on in all gears back then. I got some idea there and i introduced in the fuel table additional rows for 190kpa and 210kpa (before I had it like 180, 200, 220). That Way I hope to get it much more stable in the different conditions because in 2nd gear boost sits between 190 and 200kpa, in the 3rd gear it is at 200kpa, in higher gears it goes higher, up to between 210 and 220kpa cells in 6th gear. The boost varies very much on this supercharged engine depending on altitude, ambient teperature and as well a lot dependant on heatsoak.

In the last 2 days the ambient temps had fallen to just around 5C and the AFRs were still spot on in third gear and got out of target on the rich side just in the higher gears and high rpm. So I really hope that this is just a fuel map issue and can be solved now with the fuel map with much more resolution- leaning the 210 and 220kpa cells a little bit. Will retune it and lets hold thumbs that this is the cause for my problem.

Will keep you updated how it works out.

Thanks for all the help

You should be able to use a factory ambient temp sensor with no trouble if you wire it to a spare AT input. It will depend if the calibration matches something generic like a Bosch sensor (which it probably will).

I'm not familiar with the Mini ECU or what functionality the Vipec does and doesn't support. The manual however will tell you where the factory sensors are connected. Have you fitted an additional IAT sensor post intercooler or are there two sensors from factory?

Usually tighter load zones of 10 kPa aren't essential however if you are struggling to reliably tune in the zones either side, this can put out the interpolation. For example you could have the engine running correct AFR at 210 kPa by nature of the interpolation but have the 200 kPa row too lean and the 220 kPa row too rich. This would look fine until the boost either increases or decreases from 210 kPa and then you will see the AFR swing rich or lean.

Hi guys, I am back and still have some strange issue.

The ECU seems to make some dropout on its own and it makes the AFR go rich a lot (more than 10%), some minutes after the afr-s are back to normal. I am speaking here from going of around 11,8 target to around 10,4. this happened today and I have made a nice log where this happend on first 2 runs where I have given it WOT and some miles later the AFR was back to normal.

Does anyone have a clue what could cause the ECU to add around 10% of fuel on its own? This happens every so, 4 days ago I got one such a event where I was accelerating through 4 and 5 gears and afr go down to around 10,4 and car has not accelerated as it should, minutes later it was normal again. then last 3 days the car run great, no problems, stable afr, and today again this events with around 10% rich afr on WOT once of a sudden, some minutes later everything good again.

I have attached the log file, please take a look if you can see something strange?

Is there any possibility that the VIPEC ECU could be faulty and its making strange things or what could cause this occuring events?

hi

I just spent littel time on your log file. I'm quite sure that I found the problem. As you can see on the picture (where the mouse is) you have setted up dual fuel table. The ECU is using fuel table 1 on first WOT run. On second WOT run, fuel table 2 is active which has alot higher values in it. INJ PW is about 10% higher in second run.

hope that helps.

Apart from this, think I found a bug in the PC Link. As you can see on second pictures, the RPM for the overlayed run doesn't match.

Hi Adrian,

you are right for the map switching, but the afr goes rich on both maps. The map switching is set by purpose, map1 is with methanol injection, therefore the numbers are that lower, map 2 is without meth injection. first run is with meth and second run withouth meth- on both runs afr goes down to 10,4. on normal operation i am pretty spot on to target 11,8 on both maps. the richening problem was present on both maps.

any other idea?

what do you mean by the overlayed run and not matching rpm? I am also contacting vipec right now about the problem and could tell them about this problem as well. give me some details please.

Thanks for your help

Hi

Okay couldn't know about the meth dual fuel. Are you confident that always the right table are used at the right time? Maybe the switch doesn't work right.

The rirst thing what you should do is to check if the ECU really injecet more fuel when the rich AFR happens. Compare the

Inj actual PW

on a clean run and a rich run.

If temperatures are similary the PW must be very close from run to run. Should it really be the case that the Inj Actual PW is higher during a rich period, you can analyse the log to find the table which adds the fuel.

Otherwise it's a mechanical issue. Anyway I would check the boost system for leaks. You can make an adapter with a tire valve an give 1.5bar pressure on the system. I do that on every car bevore I start with tuning. 90% of cars have leaks!

Hi,

Map switching does work right, you can see it in the log, the first wot is with meth on, you can see it runs on table 1 (ign and fuel). then you can see the point where i turned the meth injection off and the ecu switched to map 2 (ignition and fuel). Then I gave it wot for the 2nd time and you can see the AFR go rich again once of a sudden on the no meth map, just like it has happened on the meth map. Then i drove some miles normally and gave it wot later again and afr was good again. In this test i have proven that it is happening on both maps.

as you can see in the logs the boost is constant, so i am shure its not a boost related issue. also i dont have any leaks, have tested the engine for leaks just as you mentioned, have made such an adapter for my car and this really is the best way. :)