[Diy_efi] tuning question

Tim Marsteiner tmarstei
Thu May 12 20:16:01 UTC 2005


I have been collecting wideband O2 data for the last
few months on my large displacement air cooled V-twin
motorcyle. 

http://www.superxtune.com/Tunes.htm

Turns out for my setup the best O2 reading occurs at
about 300 ms of delay from capturing TPS and RPM. I
arrived at this by putting some big spikes in a few
bins and adjusting the delay until they were centered
where they belonged in my results.

I have written a perl script to filter out the
transient data with limits on rate of change of TPS,
RPM and target AFR. I also put a primitive standard
deviation check. If the sample is beyond for example
+/- 4 standard deviations from the average I toss it
out. I have not determined what kind of distribution
the data is so I don't know if this is valid.

I have also sorted the results into finer granularity
than the 16x24 bins in the map to better reveal the
surface contour. (I can change the bins with my
software if needed)

I still notice there is quit a bit of variation from
sample to sample. I sometimes see as much as +/-5%
variation in data collected for a given bin.

However, I am getting decent results. The bike seems
to be running a bit smoother with more power. I am
making adjustments slowly making sure I am not
thrashing around and giving more "weight" to those
bins with more samples. My plan is to finish up with
the wideband before going to the dyno.

It is just that the sample to sample variation
concerns me a bit. 

Is +/- 5% variation within the realm of acceptable for
a engine of this type? 

I have heard that the wideband O2(NTK) reading can be
made more accurrate by correcting for EGT. Does anyone
have an idea of what AFR correction I might apply
based on EGT?

 
Thanks,
Tim


		
__________________________________ 
Yahoo! Mail Mobile 
Take Yahoo! Mail with you! Check email on your mobile phone. 
http://mobile.yahoo.com/learn/mail 




More information about the Diy_efi mailing list