lambse, kams... rich?

Stanger007 said:

Looks like you are mixing tuning the cruise and the WOT.

Click to expand...

My thoughts as well

You want to tune for each SEPARATELY and DIFFERENTLY :Word:

Making a change to the ENTIRE curve

It might can help one
but
It will almost certainly ... H O S E - U P ... the other

I would respectfully say
Get back to the basic info on ... How The PCM Works

It is difficult for one to improve on a thing
if
One might possibly not have a grasp on

How the thing works in the first place

Stick with it

You can do it :Teh-Win:

Grady

Mustangless said:

Got the new tune loaded and had my lambse pegged at 18.30 most of the time. Tried raising my high slope from 40 to 42.5 and was still like that.

Put my old maf curve back on and it is much better. Now my lambse is 14.4-16.0 and my kams are .91-1.013. EA said to raise my high slope to 43.5287.

Now with the questions. Most of the time my kams are below 1 and lambse is 14.4-16.0 during CL until I come to a stop. That is when my kams go above 1 and lambse at 13.3ish. It is also wanting to die when this happens.

Should I lower my low slope some? Doesn't the low slope rich/lean the idle?

Also, now with my new tune I have a hard start. I am pretty sure it is flooding because I see smoke when it does start. I can't say for certain, but I think this is only a problem after it is warm. It did not give me any problems this morning starting (32*) but when the ect was 115 is when I had problems)

Would crank_fuel_modifr_for_start be the problem for this? Or is there something else that could cause this?

Click to expand...

Some general info

Low Slope is used at low maf volt conditions/lowest part of the t curve

High Slope is used when you get higher up in the t curve

At idle ... you now see ..............

K's >1.00 so it is adding fuel
thus
The lambse is commanding fat ratios (<14.7 to 1.00) of around low 13's

See how those two things go hand in hand

Sounds like you need to lower the low slope a bit

I know what you need to do to help with the hot starts

but

It seems that some of those functions and table names were changed in the
newer software

In the old software I use a table called fuel_table_startup_OL

never the less ... whatever they call it these days

Basically ... the startup table is based upon ect and time

The values in the startup table are subtracted from the normal table
thus making things fatter. As time expires ... the mixture becomes
leaner until you slip into Closed Loop.

Grady

Mustangless said:

I was thinking I need to lower the high slope, but EA was not saying anything about it. Wes told me to try messing with the battery offset some, so I am going to try that. If that does not work then I will lower the low slope alittle.

Also, that table you are talking about is called "Fuel - Open Loop Startup" aka FN1361. I changed it back to the one I had on my previous tune. I will try to get more logging tomorrow.

I will probably have more problems by then too.

Click to expand...

Looks as if you are making some progress

Yes ... the offsets could make a difference

They mainly give you some fine tuning ability for different inj behavior when
a heavy load draw in current is encountered like on a rainy night when you
would have the ac (defrost), headlights, wipers, etc working.

You also might need to up the flow on the t curve points for enrichment
at and around idle ... but ... before you do that

A good rule of thumb for stable CL tuning is to ......

1) use as close to known t curve you can find for a start
2) use low/high slope, min inj pulsewidth, breakpoint, and offset values to
get K's all up and down the CL range of operation close to the 1.00
benchmark value.
3) finally ... a little tweec here and there on the t curve might be needed

I found a middle ground approach on the offsets was best for me

By that I mean ... I picked a selection of accessories that was kinda average
and made sure I tuned each time using the same selection.

As for the hot starts
I never did any kind of datalogging

I just did some common sense kinda tweecing

I used the thinking my 30's were a certain % size larger than the default
24# size used in the j4j1 and reduced the values by that amount.

I also hosed around with the temps and amount of time a bit.

I might have some of that stuff on my site but I just can't remember

If the screen shots of that table are not displayed along with the text of
tuning tips ... I do know you could download the shots which displays
comparisons of an unmodified j4j1 and one of the last files I made.

Grady

EEC Analyzer is sure to explain dialing in your inj's better than I can

I will share some random stuff
so
You might wanna break out the opcorn:
as
I might get a bit wordy here

I kinda look at working with the slope values as

Low for idle and just off idle driving conditions
High for everything above that

BUT .......... never forget
you really can't go all that high before you slip out of CL

Also ... if you change just the low or high slope ... you will see things
are effected on the other side of the breakpoint. More change closer
and less change farther away you could say

I see the breakpoint as a way to kinda influence the transition between
Low Slope and High Slope operation

The offsets can be effected by M A N Y variables :Word:

Always ... Always ... Always ...................... Remember This

You have datalogging at your disposal
The smart cookie kinda guy will use it to his advantage
Sooooooo ........... USE IT

I can use the offsets to give you a good example of how datalogging can
be such an EYE OPENER

Here is a little story showing some things I discovered .........
the hard way ........

I had been working on a CL tune for a while and was getting within
3% all up and down the line on my K's.

I had been doing my test runs in the late afternoon before dark

I made a run one evening a bit after supper and on the last half of my
15 minute run, it started to sprinkle.

I was shocked to see my K's were way off from the tight 3% results
I had been seeing

I threw the dlog in Excel for some really close scrutiny

I was able to see the headlights changed things a bit
but
It was the windshield wipers that made the K's really take a dump

This made me put dialing in my injectors on hold for a while and pursue
what I might see during different conditions as far as battery voltage
was concerned.

By datalogging, I could see the amount of battery voltage drop various
accessories caused.

By datalogging, I saw the battery voltage did not stabilize from a cold
start until between 20 to 30 minutes of operation.

See how the ability of datalogging allowed me to use hard data to tell
me why and how things could change

A lot of peeps might not get all that involved and I'm not saying they
should ... but ... I just like to see what make things tick

Story over and back to your Q's and concerns !!!

The deal about changing both low and high slope together

I did it all the time when I was far away from getting things close
but
I usually did only one at a time when I was fine tuning

The deal with CL tuning is so many variables can hose up your efforts
thus
You wanna be focused and keep the variables to a minimum

I'd use the t curve you got now
I'd use offsets for what ever size inj's you now got

That right there gets two variables out of the picture

Then ... I'd strive to get the K's to a range of 5 to 7% total variation for a
rough tune. You will use the slopes and bp for this.

I found I could get to this point fairly quickly by doing a run of 15 minutes
after a pcm reset to get some reference points for the adaptive strategy.

During this first 15 minutes ... I did not gather data but I did keep an
eye on the real time values.

After that first 15 minutes ... I duplicated the run again gathering data

This would allow me to see trends with the adaptive strategy :Word:

I repeated this procedure until I got the K's to that rough range I spoke
about above.

As you get closer ... or ... the range of where K variation diminishes ...........
You will have to do two things to be more accurate
1) Allow the pcm more time to get reference points
2) Gather data over a longer period of time

Of Course ... if you just datalog longer after a pcm reset ... 1 & 2 will
be captured all together on the same dlog.

Now say you got the higher end of the range looking kinda good with
the K's being close to 1.00 but the lower end still looks bad and the
idle is a bit unstable.

You can do several things in a situation like that

Slide the breakpoint to the left
Change the low slope
Change the offsets that are at and around idle voltage
Change the maf t curve points that are at and around idle

Other idle considerations

Have you done anything with your tb airflow scalar
Have you done anything with your isc neutral idle airflow function

Also ... Your mechanical adjustments like .............

TB blade stop screw
Bleed screw under that little rubber plug

Those can hose up the idle if they are kinda far out of adjustment

Have you got any stuff like breathers in place allowing unmetered air

Have you looked at the seal around the pcv valve cause they can get
brittle and leak around them causing an unstable idle as well

Hope some of that helps
and
If I was not all that clear on some of it ... just let me know

Grady

I used this idea to tune the Battery Offsets:

First before you start to mess with the slopes and MAF use both slopes equally , this will allow you to tune directly to the injectors without any other factor moving the data. Do a couple of runs using different accessories ( AC , Windshield Wipers, Headlights, etc) make EA work till you found your best Battery Offset, then start with the slopes, and finally do the MAF transfer work.

But also first of all you should really dial your idle data (TB Flow) otherwise if you tune that in the end or in the middle you will be chasing your tail.

It’s a really long work and it could be painful sometimes but in the end its worth to have a near stocker reliability.

m0l0 said:

I used this idea to tune the Battery Offsets:

First before you start to mess with the slopes and MAF use both slopes equally , this will allow you to tune directly to the injectors without any other factor moving the data. Do a couple of runs using different accessories ( AC , Windshield Wipers, Headlights, etc) make EA work till you found your best Battery Offset, then start with the slopes, and finally do the MAF transfer work.

But also first of all you should really dial your idle data (TB Flow) otherwise
if you tune that in the end or in the middle you will be chasing your tail.

It’s a really long work and it could be painful sometimes but in the end its worth to have a near stocker reliability.

Click to expand...

I certainly would not disagree with the above info

I would like to point out one more thing

But also first of all you should really dial your idle data (TB Flow) otherwise
if you tune that in the end or in the middle you will be chasing your tail.

If you do this first ... and I think it is wise to do so
and
You make notes of K readings after these changes

After working with your slopes ... You WILL see the K's change :Word:

By that ... I mean the K values at and right around idle conditions

also

Don't just focus on the word ... idle
cause
A low load cruise driving condition is the same as an idle condition

I like to focus on the maf voltage values a good bit of the time

Pulsewidth values can sometimes mislead you as well if the adaptive is
hard at work and close to it's maximum limit.

anyway

After getting the tb scalar and air flow function points all squared away ...
I'd not change them

I'd get as close as possible with the low slope
and then
See what can be done with slight tweecs to the idle t curve points

Yes ... you can spend a good bit of time with the CL tune
but
These days ... So many have given good, reliable, info

That info can R E A L L Y cut down the trial and error part of it

Grady

another thing whose importance is often overlooked is the exhause pulse delay.

if you have long tubes, or if for some other reason your O2s are farther away from the exhaust ports than stock, you should try to get those right early on, because otherwise, the feedback from the O2s back to the computer is out of sync with what the computer thinks, so the decisions it makes are not very accurate

also, if your cam is very different from stock, the injector timing is important, because if the injectors start firing too late and the valve closes before the injectors finish firing, then the O2s will report lean, and the computer thinks it needs more fuel, so it fires the injectors more, but that doesn't help because the valves close before the fuel makes it in. then every so often enough fuel has pooled somewhere and it alot of it gets sucked down into the cylinder, causing a misfire and the car stumbles. and the gas mileage is terrible. i believe that is the situation i am in right now

Don't know if you are talking low slope or high slope
but
Many have used this rule of thumb to start out with slope values

High slope value = inj size
Low slope value = inj size X 1.2

or in your case ........

High at 42
Low at 50.4

Grady

Mustangless said:

I know that high slope is idle, but can also be light throttle. What is considered light throttle though?

Sorry for all the questions, but I am

Click to expand...

Believe me
I understand ... It took me a while to catch on myself :Word:

Your low slope effects idle and just off idle
and
Very light load cruise

High Slope will effect .........

higher loads
greater pulsewidth
higher maf volt

kinds of conditions

Have you ever taken the time to read through all those EA help files

Here is a thought that might help

Take a cruise with the tps position included in the payload

Then you can see how the amount of open throttle relates to stuff like .....
CL or OL
load
injector pulsewidth
maf volts
yada yada yada

It still seems like to me

You're trying to tune without a basic understanding of how
the pcm sets up fuel and spark values
based upon the input from the various sensors

Maybe some of the other peeps can lay it out in a more clear way

Grady

Going lean on coast is a-ok.

Datalog your wideband in with the twEECer and you can see if you have any lean tip-in issues. Use scalar PID "AIR_MAN_VOL" to nix a tip-in lean/rich condition.

Wes

Mustangless said:

The only thing I dislike about it is the car starts jerking, kinda like driving over bumps. This could be another problem though.

Click to expand...

Log the load and spark when you experience this. Try dropping the spark by 10 degrees and try it again.

Mustangless said:

When I calculate my offset EA tells me to move my 14.00 from 0.7500 to 0.8932. Does that mean it gets more fuel?

Click to expand...

Correct, that will increase the injector pulsewidth and add more fuel.

Mustangless said:

Also, what would cause the car to get out of CL? I am driving around without giving it much throttle at all. Looking over my log, it shows go from CL to OL about every 3mins for 30 seconds or so.

Click to expand...

You got me on that one man.

Mustangless said:

BTW, when I did the offset for my wideband it did not work perfect. Like when I put 2volts for high/low in lmprogrammer, calcon was showing 19.79-19.84 (not 19.79) For the low it should different fluctuating numbers. I just added the two together and avg them out If I compare it to logworks it looks very close.

Click to expand...

Sounds like you got some jitter. I have mine scaled as 10:1 = 1v 16:1 = 5v. Gives you a bit more resolution and will negate any sort of electrical noise you might be getting.

Wes