Miatapower List Archive

This read-only message was archived from a public mail list.
Mail From: "Ralph" <(email redacted)>


doehrmann.net/critter.jpg

This screen shot shows how erratic the O2 sensor reading is after it gets
hot. The 200 row and part of 300 row are all over the place. This is from
a dyno run. The car had several other problems that contributed but this
was the log that really pushed the outer bounds of what could go wrong with
your data. The Critter is intelligent enough to not suggest a timing
adjustment until after the fuel for that zone has been sorted out. Code is
in place for standard dev. but not the storage structure so I cheated and
did an average deviation... call me lazy. Didn't take time to clean up
screen cosmetics either.
Right now the animal does a whole bunch of data checks and looks back x
number of datapoints to see if this datapoint is useable. The number of
historical points varies with boost. With a cool O2 sensor the logs would
have parsed out real pretty.

Where I'm going with this all is that Joe Average can tune without a dyno
and get logs that can tell them exactly which zones to tweak. Maybe even
how much to tweak them (but that varies with injector size). Detailed
instructions will be posted later.
When wideband datalogging gets added I'll include it.

Thank you. Drive through.

Ralph Doehrmann
Sunburst Yellow
BB Garrett, Link, FM, BEGI

This read-only message was archived from a public mail list.
Mail From: Mark Peugeot <(email redacted)>



Wow, pretty cool.



On Wed, 26 Sep 2001, Ralph wrote:

>
> doehrmann.net/critter.jpg
>
> This screen shot shows how erratic the O2 sensor reading is after it gets
> hot. The 200 row and part of 300 row are all over the place. This is from
> a dyno run. The car had several other problems that contributed but this
> was the log that really pushed the outer bounds of what could go wrong with
> your data. The Critter is intelligent enough to not suggest a timing
> adjustment until after the fuel for that zone has been sorted out. Code is
> in place for standard dev. but not the storage structure so I cheated and
> did an average deviation... call me lazy. Didn't take time to clean up
> screen cosmetics either.
> Right now the animal does a whole bunch of data checks and looks back x
> number of datapoints to see if this datapoint is useable. The number of
> historical points varies with boost. With a cool O2 sensor the logs would
> have parsed out real pretty.
>
> Where I'm going with this all is that Joe Average can tune without a dyno
> and get logs that can tell them exactly which zones to tweak. Maybe even
> how much to tweak them (but that varies with injector size). Detailed
> instructions will be posted later.
> When wideband datalogging gets added I'll include it.
>
> Thank you. Drive through.
>
> Ralph Doehrmann
> Sunburst Yellow
> BB Garrett, Link, FM, BEGI
>
>
>
>

This read-only message was archived from a public mail list.
Mail From: "Jerry Malsam" <(email redacted)>


Tell me about it...: highspeed.malsam.com/critter1.gif

Seriously, tell me about it!

--Joe Average, eagerly awaiting "detailed instructions" =D

----- Original Message -----
From: "Ralph" <(email redacted)>
To: <(email redacted)>
Sent: Wednesday, September 26, 2001 11:44 AM
Subject: Here's an example of why dynos and O2 sensors don't like each other


>
> doehrmann.net/critter.jpg
>
> This screen shot shows how erratic the O2 sensor reading is after it gets
> hot. The 200 row and part of 300 row are all over the place. This is
from
> a dyno run. The car had several other problems that contributed but this
> was the log that really pushed the outer bounds of what could go wrong
with
> your data. The Critter is intelligent enough to not suggest a timing
> adjustment until after the fuel for that zone has been sorted out. Code
is
> in place for standard dev. but not the storage structure so I cheated and
> did an average deviation... call me lazy. Didn't take time to clean up
> screen cosmetics either.
> Right now the animal does a whole bunch of data checks and looks back x
> number of datapoints to see if this datapoint is useable. The number of
> historical points varies with boost. With a cool O2 sensor the logs would
> have parsed out real pretty.
>
> Where I'm going with this all is that Joe Average can tune without a dyno
> and get logs that can tell them exactly which zones to tweak. Maybe even
> how much to tweak them (but that varies with injector size). Detailed
> instructions will be posted later.
> When wideband datalogging gets added I'll include it.
>
> Thank you. Drive through.
>
> Ralph Doehrmann
> Sunburst Yellow
> BB Garrett, Link, FM, BEGI
>
>
>