bagodoosh

i spent a good portion of today reading this stuff. i'll make a hypothetical situation. tell me if i'm correct in understanding the theory:

let's say we log block 003, and get the following timing data (in degrees) per 1000 RPM @ WOT:

4, 5, 6, 7, 8, 9

A) this is the "ideal" timing minus CF (correction factor)

then we log block 020 & 021 and see about even CF for all 6cyl ranging from 5-9 @ WOT, 0 elsewhere.

B) i read people try to tune for 3-5 CF
C) OEM spec finds up to 12 CF acceptable
D) most people don't prefer over 10 CF

now we use lemmiwinks ch09. we set timing to -3.

E) this retard the "ideal" line by 3 degrees
F) block 003 numbers should now appear like (assuming we could reproduce the exact same scenario):

1, 2, 3, 4, 5, 6

G) CF in blocks 020 & 021 should now range between 2-6 @ WOT

H) assuming the ignition retard is caused by knock (over 20v in blocks 026 & 027), we can also use lemmiwinks ch08 to increase (secondary) fuel for WOT.
I) doing H should reduce CF & knock.

did i grasp A-I correctly?

Ochlockonee

I have seen conflicting posts about the effects of secondary fuel. Someone even used a wideband and added secondary fuel and the wideband read even leaner!
Also I believe we take it under assumption that a CF of 1=1 degree of retard. It probably is. In some vehicles however,a CF (or whatever that particular manufacturer calls it) can equal 5 degrees of retard.
There are some other things about knock that others will correct you about too, but I have to go.
Good post!

-M3=Lunch-

I thought 15 was the Max amount of retard?

S4saken

I've got bits and pieces of it, ie:

1. O2 voltage should not exceed .85 @ WOT
2. CFs should not exceed single digits
3. Injector duty cycle should not exceed ???

what other parameters need to be optimized?

Ed

PSmalls::DoggieStyle

02 voltages can go above .85 @ WOT. Seriously, different cars perform better at different AFRs.

OCA::DIY Motorsports

2nd!

bagodoosh

this is where i get confused because i think people confuse it with lambda of .85.

i'm going by this graph:

<img src="http://bagodoosh.home.comcast.net/s4/lambda-afr.gif">

a lambda of .85 = 12.5 AFR ~ .915v

i thought this was desired, where as..

.85v ~ lambda of .925 = 13.6 AFR
.80v ~ lambda of .95 = 14.0 AFR

not exceeding .85v means you're running leaner than 13.6, which seems way too lean @ WOT.

PSmalls::DoggieStyle

A.)I don't understand this statement at all. What do you mean by "ideal" timing? The timing you see in block 003 is the *actual* timing as I undersatnd it.

And what do you mean by minus CF? Are you thinking that the 003 timing is the desired and then when you subtract the CF, you get the actual timing? That's not the case at all AFAIK or else everyone on this forum is doing it wrong.

B,C,D) Try to stay in single digits IMO. However, everyone has a different philosophy with regards to tweaking. I guess it just depends how you weight performance vs. reliability!?

E.) I still have no idea what you are talking about with this "ideal" timing line.

F.) You are oversimplifying everything. Things don't quite work out like this in the real world from what I've seen. Reducing timing by 1 degree with lemmi won't reduce CFs by 1 across the board, or reduce actual timing by 1 degree across the board either. It would be nice if it were that straightforward, but unfortuantely it really isn't...from what I've seen at least.

G.) See F above.

H,I.) I've heard of mixed results when attempting this. I will be sure to do quite a bit of testing once I get ECUx working (should be soon, hopefully). You can try, but take this "fact" with a grain of salt. I know I'm one of the people who have perpetuated this "fact", but I'd like to do more testing.

If I screwed something up or missed something or whatever, please feel free to correct me...hope this helps.

bagodoosh

"desired" / whatever the ECU was programmed to achieve -- but can not because of such and such which need to be corrected via CF.

how about this:
J) if CF=0 then block 003 is pretty much what the ECU was programmed to do under the conditions
K) so then i concluded (block 003 timing) + CF = "desired" / "ideal" / originally programmed timing table

i simplified everything to understand the basics. then i'll work my way up. if there are any bookmarked threads relating, i'd like to read.

Ochlockonee

I think the graph speaks for itself. You can't use a narrowband o2 sensor like a wideband.