So my buddy and I were tuning my car last night. Everything was going about as perfect as could be. Part throttle was as crisp as I have ever seen. VE table was so so so so close. So we changed the values ever so slightly, like by 2 digits or maybe 3. That's it.

What happens next? Car starts running extremely lean, as if we made huge changes to the VE table. It was so lean it was almost undrivable. My LC1 wideband was reading as high as 20.0 at times and averaged around 17.
WTF?!?

We double checked our changes (which weren't many) and nothing was drastic or out of line. It was almost identical to the previous file that worked so good. Next we uploaded the previous file which was the one that was so so so so close. So it should be back to near perfect, right? Nope. Still way too lean!
WTF?!?

So we basically had to start all over. Currently we are back to being really close again. So we decided to leave "close enough" alone.

Has anyone ever encountered something like this before?

 

One more thing...
Fuel pressure was fine, about 62 at hot idle, 63 at cold idle.

 

If the IAT get hot I have seen it change things alot. Make sure that it does not heat soak

 

Do you mean the IAT sensor can get heat soaked or just that the intake air temp can become high?

Interesting that you say that. At the end of our log (the one where we were seeing really good data) we sat in a parking lot letting the car idle for about 5 minutes. We also kept adjusting the commanded idle by +/- 200 rpm increments so we could hit some more cells. Then we shut the car off, made very slight changes, uploaded the new program, restarted the car and presto, very lean.

Could that have caused the IAT sensor to become heat soaked?

Only thing is, we monitored the intake air temp and it was nothing extreme. Also, we even tried driving around for a few minutes to see if that helped and it did but very little.

 

When the car was sitting, I bet the IAT and ECT got a bit warm. That alters the charge temperature (in degrees Kelvin), which is a factor in the fueling calculation. The PCM determines grams/cylinder with the following equations:

g/cyl=(VE*MAP)/K

charge temp: K=273.15+IAT+((ECT-IAT)*factor)

IAT and ECT need to be in degrees celcius for the 2nd equation to work. The factor is pulled from the PCM and usually is closer to 1 for low airflow scenarios.

A good example is to take an idle situation. Keeping everything constant (IAT, ECT, VE, and MAP), changing the factor from 0 to 1 is the difference between .180grams/cyl and .142grams/cyl. That's using a nearly stock VE, IAT of 56*F, ECT of 195*F and a MAP of 35kPA. Because idle is a low airflow situation, the factor should be close to .55 (.156grams/cyl).

If you increase the IAT and ECT in the example above to 110*F and 218*F, changing the factor from 0 to 1 is the difference between .162grams/cyl and .137grams/cyl. So, now your calculated idle airflow has dropped to .147grams/cyl using the same .55 factor.

The point is as IAT's and ECT's rise, the charge temp should naturally increase (passing through the hot heads into the cylinder) and the grams/cyl of air should decrease naturally (hot air expands). The PCM tries to compensate for this. But, I'm thinking it overcompensates if it leans out as temps rise. I'm currently experimenting with changing the Charge Temp Factor closer to 0 to see if the higher calculated grams/cyl actually richen up my fueling some at lower airflows. However, with more focus on the IAT, the swings become greater (.18~.162) where as ECT focus shows swings of only (.142~.137). I've also heard that pre-'01 cars might not respond to this.

 

In actuality I did nothing to return AFR back to normal, we pulled over for awhile, started the car back up and the problem vanashed. I'm wondering if there was a wiring issue, injector problem ect.

At the time the issue started ECT was only at 185, intake air temp not above 85 to 90* The issue did not resurface the rest of the night, even idling the car for awhile.

The only mechanical issue I noted was inconsistant fuel pressure with key on/key off operation.

 

With the car running, though, fuel pressure was always around 62.

Josh, look at your logs last night. First look at what the IAT and ECT were when we drove from the gas station to Hooters. I am willing to bet they were a lot lower than the log from Hooters to White Castle. Probably by the time we were getting close to White Castle, IAT and ECT had dropped again. That could be why the problem corrected itself.

Now, if this is the case then this is kind of a bum deal because what are we going to do about hot summer days sitting in the staging lanes at a stand still or creeping forward. Won't the IAT and ECT rise substantially in those situations? If this is the case, and since my car is mostly a race car, wouldn't you want to tune for those situations more than normal street driving conditions?

SSpdDmon, thanks for the informative reply.

 

One more thing I wanted to add...

This was discussed over at the EFI Live forum months ago. I never got around to it, but the proposed solution was to relocate the IAT sensor further forward (away from the MAF) in the hopes of avoiding the heat soak that is prone to that spot in the lid. Because the swings are more pronounced when IATs change, this would theoretically stabilize those swings somewhat and maybe kill this whole lean@hot idle issue. Not 100% on this though as there is no real world proof.

 

My stomach was churning this morning from Whities

IAT was no different at Whities than Hooters. I'm going with it was a fluke. As far as heat soak, we took it on the highway really quick to see if that would cure the problem, it didn't. IMOP it had nothing to do with heat, IAT ect, it was some fluke, either in the software, wiring, stuck injector, the list could go on.

 

My car randomly started doing this not too long ago. If I get up higher in the RPMs though, it seems to go away and if I play with the throttle enough, it will correct itself. I too am trying to get my VE dialed in.

 

I dropped my charge temp blending table significantly. Afterwards, I saw a max of 15.1:1 AFR when my IAT's hit nearly 110*F. I think I'm on the right path to a fix here. Will keep you posted.

 

http://www.hptuners.com/forum/showth...t=lean+startup

That's the only help I can offer...after relocating my IAT sensor completely out of the airflow stream, and setting my IAT/ECT bias to 1.0, it finally doesn't do it enough to worry about. Just one of those weird quirks about doing an SD tune on these cars...after all, they weren't designed to run full time on them.

 

It'll be open loop +maf once the tune is finished

If the problem occours again i'll look into relocating the IAT.

 

See, I set my blending to 1.0 also across the board. I noticed there were extremely low values in it to start with so it almost solely depended upon the IAT temps.