The MAF has to be scaled in such a way that it will read around 30hZ at idle.

I found this by swaping out my HKS suction pipe with a stock rubber pipe and instantly the LTFT Low made changes.

I plan on running the stock rubber tube for a while so I can develop proper scaling for the HKS since the HKS idles at about 53hZ.

 

MAF scaling doesn't change the frequency that the car runs at though? The MAF frequency is just a product of air velocity through the sensor.

If this is true, that it requires the MAF frequency to be within a certain range before it allows LTFT updates, then finding the values in the ECU that control the range would be the way to fix it.

 

That makes perfect sense to me.

I guess all we need is for someone to code the LTFT values and make it adjustable.

This would be wonderful to allow the use of hard intake pipes, cams, and the MAF. Not all of us are so eager to go full SD.

 

I messed around all weekend with scaling the MAF several different ways.

My car idles at 38 Hz and the LTFT are pretty stable at "not moving any where." Even the STFTs don't really do much.

I did notice some interesting behavior though that wasn't explained by any maps I have seen. In open loop with the EGR disabled, the ECU wants to jump around between 14.7:1 AFRs and 16.5:1 AFRs when the main map shows nothing of the sorts.

 

Maybe the EGR is used to stabilize idle in some way?

 

It has been noted in another post that the trims correspond to MAF hz in the way you mentioned. If I recall correctly LTFT low shows for hz at or around 30hz. I can't recall what mid and high refer to though.

 

My mid was adjusting at idle with the HKS suction intake pipe at 50hz.

I seriously believe that if someone can find the definitions of where each trim starts adjusting we can use these to change the values to make hard pipes work with the MAF and big cams. This means those of us that want the few HP a hard pipe can offer, but still want daily driving capability would have the best of both worlds.

 

If the evo remained anything like the DSM, ltftlow is probably near 0-88hz, while mid is 88-180hz

See here: http://www.fullthrottletech.com/showthread.php?t=105 (image in "Base Knob Tuning" section)

It would be great if someone could find the range for the evo ECU!

Calibrating the MAF on my 2g with this knowledge and simply cruising with a 0 trim at that airflow while watching STFT made my drivability incredibly better and also got me up to 27mpg!

 

Also, the DSM has a "current LTFT" value that is loggable. This would help us in determining the value above. This would also be great for not having to worry about hitting 0 LTFT in the range while adjusting based on STFT. If we had "current LTFT" we could easily make a "combined FT" loggable item that would be an easy way to reference while tuning outside of a 0 current LTFT.

 

Oh man, after looking at the protocols.xml file for log works, I now realize why what I was seeing wasn't making any sense.

Logworks has three fuel trims:
(MUT 0C) Short Term Fuel Trim
(MUT 0D) Mid Term Fuel Trim
(MUT 0E) Long Term Fuel Trim

This lead me to believe I was logging a short term, mid term, and long term fuel trim values.

Compared to EVOscan, it's logging the wrong MUT channels for what the discription is. They should be:
(MUT 0C) Low Long Term Fuel Trim
(MUT 0D) Mid Long Term Fuel Trim
(MUT 0E) High Long Term Fuel Trim
(MUT 0F) Short Term Fuel Trim (O2 feedback)

I'll update the logworks entries and see what it does.

 

I believe our Evos MAFs low will be much lower than the DSMs. I say this because with my HKS suction it would idle stead at about 50Hz and the LTFT Mid would adjust. I installed the stock rubber intake and idle is now around 30Hz and the STFT Low has adjusted. Knowing this really does nothing for us because we do not have the ability to adjust the base Low and Mid crossover points.

 

Zee germans, they will do their best to confuse you

I guess I don't understand why this is a problem. It seems likely that a more free flowing intake would allow for a higher airflow at idle (since it is easier to draw the air through, but the car doesn't really know that) and that the car is just compensating properly.

The DSMLink relic of g/rev of airflow would be nice too... I should work out that equation. That is much more likely to tell you if your idle airflow is in the ballpark than MAF Hz.

LTFT low is just a range of long term adjustment from STFT. You want it to encompass all near idle airflow conditions, and 0-88hz seems like a very viable range for that.

 

The problem is that as you approach 50Hz the Mid trim starts to take over. It causes idle stability issues since the car is designed to run at idle around 38Hz.

I fully agree that a free flowing intake is better for HP. I am trying to solve idle issues with CAMS and hard pipes.

 

After changing logworks, everything is making sense and I saw some interesting behavior.

I am using a Tial BOV, so I have so lift off 'issues.'

I noticed that the problem isn't directly the rich condition induced by the BOV. The real problem looks to be the STFTs. On lift off, AFRs go into the 10s, but the motor runs fine. STFTs drop to about -80% though. This then forces the AFR into the 18:1-19:1 AFR range, and this is when the car wants to stall/run rough.

Would it be possible to limit the max/min STFT values? Limiting it to say +/-25% may greatly reduce the lean conditions that cause the rough running conditions while still allowing plenty of range to allow proper adjustments under normal conditions.

 

Cams can also change your idle g/rev pretty easily. Take a look at the ecmtuning knowledgebase for a reference on why that is a better measurement.

I'll have to see what frequency I'm idling at and see if I can reproduce your 50hz LTFT Mid changes with some temporary BISS adjustment