RomRaider

I have changed the location of my front O2 to the DP. I get AFR fluctuations that indicates the STFT are adjusting too fast for the amount of time it takes for the exhaust gasses to reach the sensor.

I imagine there is some sort of PID logic that controls the STFT.

Any chance of them getting defined?

I assume they are different in the OEM TS cars.

This would be very helpful for anyone that wants to run OEM O2 in the PD, or anyone with a TS set-up.

I've see the fluctuations as well, like low-14s to low-15s every second or two. I'm pretty sure I saw them back when I had the sensor in the headers too though. I think it's by design. Someone posted something a while back the described deliberate AFR changes to improve catalyst efficiency.

It would be great to have someone log several seconds of idle, with a 32-bit ECU, warm engine and stock O2 sensor setup, just to confirm / dispute my "not related to the O2 relocation" theory. Interesting parameters would be just AFR (ideally wideband, but stock may be OK) and time. Nothing else. Faster logging of AFR would be better than slower logging of AFR and anything else.

Maybe, but with the front O2 in the DP there is a definite drive-ability problem. The swings are too big, and affect consistent vehicle performance. Something that was not there when the O2 sensor was in the header.

I do not have logs showing AFR swing when it was in the header, but I have lots with it in the DP. There is no mistaking the feel of the car now in CL.

That's odd, I didn't notice any difference when I moved the sensor to the bellmouth. I've asked a local guy to log his stock-location O2 sensor, so hopefully we'll have something to compare against in the next few days or so.

STFT logic is very complex - basically taking current target afr and comparing it to current sensor reading and adding/subtracting fuel (jist of it)

there are a bunch of bits that its based off of and a few other compensations additions that complete it. Its not simple like long term fuel trims, there is no store maximum (persay) - I mean there is, but its not stored in the same fashion.

Given we are talking 16bit, its not something that is extremely easy to manipulate.

Do you have those AFR swings at idle or steady state low RPM\TPS\LOAD?

Or do you have drivability problems under changing TPS position (at low\middle RPM)?

In the last case you probably need to run open loop for 300-500 ms just after throttle movement. And adjust tip-in enrichment more closely. I believe it is possible to control.

I see the swings at idle.

Fortunately this could not limit the maximum velocity..

My car has a 32 bit ECU.

Idle is good, 50% of steady state driving sees the swings, the 50% doesn't.

There is the odd driveability problem at low throttle positions. I skewed the O2 sensor's calibrations, so the corrections wouldn't be soo bad, it worked, but not 100%.

Obviously tip-in has no affect on this issue, but I am sure I could adjust it a bit better.

Not sure exactly what you mean by, "In the last case you probably need to run open loop for 300-500 ms just after throttle movement."

Thanks

After throttle position is changed it is useful to stop AFR corrections to be done. For about a part of a second. Run openloop this period.

has any one done testing to see just how accurate the front o2 sensor is when moved from pre turbo to post turbo? I know pre turbo can cause issues do to back pressure. I'm wondering if the stock front o2 becomes more accurate/reliable for higher load situations when placed post turbo?

Thanks for the link, it would appear as though my searching skills could use some work

We already did that thread.
http://www.google.com/search?q=relocate ... raider.com

This is from a log of the "AF Sensor #1 (2-byte)" parameter from an 05 Legacy GT with the OEM O2 sensor in the stock location. The engine was warmed up with some driving, then this log was taken while it just idled. The Time column in this log was in a weird format (35:41.4 ???) so the X-axis is just the samples number. There were about 10 per second.



The owner is going to move the sensor to the downpipe in the next few days, so we'll get a comparison.

Might have missed it somewhere else, but was there ever an update to this? How'd this owner fair with the sensor in the dp?