Cold Weather Heatsoak Issue

[sdtib]

2008 Hyundai Tiburon, Supercharged, MS2X

When I first installed the MS2 I had the IAT in the cold air intake tube a few inches upstream from the throttle body, pretty much in the stock location. In that configuration, it was always seeing pretty close to ambient temperatures which created a problem when boosted; it would not detect the increase in the temperature of the air actually entering the cylinders after being heated by the supercharger. In fact, the temp being detected by the IAT would decrease with the increase in speed. This was causing it to run very rich on the top end. To combat this, I relocated the IAT to one of the intake runners after the supercharger and intercooler. This worked great during the summer months (95* to 100* here in FL), detected temps rose with boost as they should and the tuning became more predictable. I did notice more of a heat soak issue on hot starts but was able to work around that with some increased afterstart enrichment.

Now that winter is here (lows in the 30*-40* range) I am having a huge heatsoak issue at idle….within about 5 or 10 minutes after start up, the IAT reads 110*-120* due to under hood heat, when the air entering the cold air intake is only 40*.EGO control is correcting as much as 125% trying to pull the AFR down. Even at that, I’m seeing AFR’s in the 15-16 range at idle. It gets a little better when cruising, the increased flow of air through the intake tract brings the IAT reading down to around 100* so my summertime tune works ok, EGO correction is 5% or less.

I was thinking of ditching the cold air intake and switching to a short ram intake. I figure if the intake tube is in the engine bay, it will be sucking in air at approximately the same temps as what the IAT sees. Not the best solution for overall performance but I think it may make the engine easier to tune at all temperatures.

Any thoughts, do you think it will help?
Is there a way to tune around this?
Has anyone else experienced this same scenario?
What did you do to overcome it?

[Matt Cramer]

I would try tuning the MAT correction curve and the MAT/CLT Correction setting to see if you can dial this out.

[sdtib]

I don't think MAT correction is the right way to go, I could fool it into working when it's cold, but it would throw my fueling off when the air temperature actually is what the IAT is reporting. I tired playing around with the MAT/CLT correction once without much success but I don't think I fully understood it at the time and I wound up making things worse (ran even leaner than without it).

So, help me understand this...If actual air temp is 50 degrees, but the IAT is reporting 100 degrees due to heatsoak, I have an error of 50 degrees. So, that means that I need to adjust the percentage so that is is equal to the 50 degree error. In this case, if CLT temp is 200 degrees, 50 degrees is 25% of 200 degrees so I should set the correction factor at idle airflow to 25% then decrease it as airflow increases. Since the ECU now sees the correct temperature, I need to fatten the VE table at idle to compensate and also taper that down as RPM and MAP increase.

Does this sound correct??? I think where I went wrong before is not understanding how I had to adjust the VE table; I let it alone and the results was that it ran even leaner than with no correction.

I also like the idea of using 2 sensors as discussed in this thread http://www.msextra.com/forums/viewtopic ... 34&t=56840

[Matt Cramer]

Can you post a data log from both summer and winter running and a copy of your current MSQ?

[sdtib]

I have been modifying the tune a lot lately trying to get it run right. The temperature should be in the 40's when i get off work tonight around 11 pm so I will load an older tune and run a data log on my way home then post the log and msq. In the meantime, I have attached 2 log files, one with the IAT sensor in the intake tube, and one after I moved it to the manifold. you can easily see the difference in behavior of the sensor from one location to the other.

[sdtib]

I loaded up a tune that I was using back in August when the temperatures were still in the 90's. I don't remember this specific tune so I can't guarantee the engine was running perfectly with it but it had to be somewhere close. The temperature on the way home tonight was 45 degrees, the 1st log file is soon after start up but after warm up enrichment was over and EGO control kicked in. this is slow cruise toward interstate, about 45 mph and then up the on ramp to the highway. After getting off the highway, I shut the engine down for about 10 or 15 minutes to allow the intake to become heat soaked. The 2nd log file is right after restarting the hot engine and part of the rest of my way home.

[Matt Cramer]

Are you using a closed element sensor for the IAT by any chance?

[sdtib]

No, open element. It's the oem sensor.

[Matt Cramer]

This sensor seems to have a considerable heat capacity that is affecting its output curve. You may be able to tune around this with the MAT/CLT correction number, but it may need a different sensor or some change to the way it's mounted.

How exactly is the supercharger set up?

[sdtib]

It looks like this.....



The IAT sensor is in the intake runner on the left. The sensor is plastic so shouldn't conduct too much heat and I don't know where else I could mount it and still get a reading post-supercharger and intercooler.

I tried messing around with the MAT/CLT correction and remembered why I gave up on it the first time. It does the opposite of what I think it should do. When I put in 25% coolant at idle airflow, It increases the MAT correction factor putting further down the curve where it wants to decrease the PW and causes it to run even leaner. You can't see it on the gauge cluster, but when looking at the MAT air density table while changing the coolant value you can see the cursor jump up. In my case, with a zero coolant value, and the coolant temperature at 185 degrees, the IAT reading was 120 degrees, as soon as I put a value of 25% in the table, the MAT correction jumped up to almost 140 degrees. I was expecting that when I blended coolant temp into the equation, that the MAT correction would move in the other direction so that the PW would increase and help enrich the mixture.

Am I totally missing the concept here or is the MAT/CLT correction not working properly?

[BigLou240sx]

When using the MAT/CLT correction, you need to add fuel to the affected areas in your VE table. Since CLT will almost certainly be higher than MAT in this situation, it pulls gamma air correction lower as you saw.

I needed to add about 10-20 percent to the lower left portion of my VE table when i started using the MAT/CLT correction.

[sdtib]

I did adjust my VE table as you mentioned and that worked when the engine was hot but on cold starts it ran extremely rich.

I guess the way it works makes sense but only on an NA application where the sensor is in the intake and you need to increase the MAT reading. In my case, what I need to have happen is for the perceived temperature to be lower since the sensor is in a hotter location than the ambient air. Except when I have high MAP readings, then I need it to follow the actual temperature and go up. Dual sensors may be the only way to go.

[BigLou240sx]

I'm unable to comment on the rich condition during warm up... I did not experience that. Regarding reading the real MAT temperature when MAP is high:
The MAT/CLT correction curve does allow the MAT to read the real value when MAP is high depending on how you setup the curve. My MAT/CLT correction curve is 0% at 48000 flow (load*rpm) (where my curve ends). So by my curve, if I see 50kpa at 1000 rpm, the real MAT reading is used. Likewise for 25 kpa at 2000rpm and so on.

Dual sensors is an interesting idea. Myself, I will be pulling my MAT sensor outside of the engine bay but downstream of my intercooler (my engine bay gets very hot with the turbo/manifold setup and steel intercooler piping). Unfortunately that doesn't look like an option for you. Maybe more playing with the MAT/CLT curve (and VE table in the affected area) will make it driveable... it took me awhile to find the happy medium. Admittedly, it doesn't solve all of my heat soak issues currently, but it gets my car to not stall from running lean and to get me driving (and passing more air over the sensor/engine bay to cool back down)

[sdtib]

Yeah, temps at high MAP are not a problem, I drop the MAT/CLT correction to zero just above idle so it works in that respect. It's at idle that I have the problem and the correction does the opposite of what I need it to do in my case.

I've been playing with the VE table the last couple of days and here is what I came up with....
I ran a log file on a cold morning, after the WUE was done, and then ran VE analyze on that log file. Did the same thing later in the afternoon when ambient temp was in the 70's and MAT readings were between 110-120.I then looked at both tables and set my idle and low MAP areas to values in the middle of the two. Then I increased my WUE a little bit and made the MAT air density curve a little steeper on the cold side and I'm gonna let ego control take care of it. This way EGO control runs at around (+ -) 10% at idle depending on the what the MAT sensor is reporting. Seems to working OK......