ecu fuel trim

[totta Crolla]

Can anyone help me with interpolated fueling on a programmable ecu ?

I understand the concept of interpolation but I'm a bit stuck on where
to add or remove fuel.
Say I want a particular area of the map to be 10% richer do I add
2.5% to each of the four adjoining cells in that area or would I add
more fuel to fewer cells ?
Is there another way ?

Thanks

[jondee86]

Each cell is a separate entity i.e. a finite point on the map. Interpolation
only comes into play if the operating point falls in between cells. For
example, if the two adjacent cells have a value of five, any point between
will have a value of five. If the adjacent cells have values of five and
three, the midpoint between them will be interpolated as four.

So to make a single cell or cells 10% richer, add 10% to the existing cell
value. To avoid creating a "skyscraper" effect, you need to taper down to
surrounding values by proportionately adjusting adjacent cells. Some map
software allows you to select a range, and interpolate between the first and
last cells in the range. Very useful for smoothing

Check what options come up when you highlight a cell and right click on it.

Cheers... jondee86

[Uzelac]

What ECU?
Map via VE or direct fuel ?
Does it use AFR target ?

[totta Crolla]

Ok it's a Freedom ecu
Direct fuel I think and my particular Freedom does not use target AFR
which does make getting a fuel map even close somewhat of a challenge.
It does have a basic learning function that will get it to approx. 14.7 AFR
but that gets a little scary beyond 6000rpm under load.

I'm using TPS vs RPM with an Innovate LM2 AFR meter and it is fairly
easy when data logged to see where fuel needs to be added or removed
on the map.

[jondee86]

My ECU (Adaptronic) has a learn function that adjusts fuel in each cell
according to a target AFR table. Once it has "tuned" a reasonable number
of cells, I manually smooth any widely varying cells, and also fill in some
of the "un-tuned" cells = interpolation by eye Then, I use data logging
to verify that my auto-trim values are not excessive (running closed loop
with an LC-1).

If you only have open loop control, you will need to set up the high load,
high rpm part of your map on the rich side, and manually trim the cells
after data logging as you do now. FWIW with ITB's on my 16V I found that
the top end of the map is fairly planar. The bottom end had most of the dips
and ridges.

Cheers... jondee86

[totta Crolla]

At the moment the ecu is not applying any corrections to the fuel map
but I can set it to add or subtract a percentage of fuel at a given TPS
opening, so theoretically it would be possible to allow learning to say 60%
TPS and then apply the correction to richen the map for the last 40%

How does that sound ?

Good info about the top end of the map, I've set the rev. limit at
6000rpm to keep things relatively safe.

[jondee86]

I fired my engine up on this map...



It was just to have a starting point for a TPS x rpm tune. This is the only
one I have a pic of, as I switched to VE tuning shortly after, and have
used VE ever since. As time went by and I refined the map, and it grew
some troughs and ridges, but the gradient of the top end did not change
much.

If you are using ITB's you will be at near enough to atmospheric pressure
from 40% TPS. So from 60% the amount of fuel will increase in proportion
to the increase in rpm, any improvement in VE, and the richer AFR that
you want at higher loads. Therefore, you could take your 60% line and
(for example) add 15% more fuel to give you a value for the 100% line.
Interpolate for the actual enrichment steps.

Don't know if that makes a lot of sense, so take it with a grain of salt

Cheers... jondee86

[totta Crolla]

I have studied what you have said there and yes it does make sense
and I have also changed a few cells as you suggested and it is of course
very subtle and 10% over four cells it seems is a massive change, to
remove a lean spike the fuel I added was probably in total less than 3%

You mention VE tuning there Jondee and whilst I'm aware of this tuning method is it something that needs to be supported by the ecu or can a VE tune be calculated ?

[jondee86]

totta Crolla Wrote:Can a VE tune be calculated ?

Essentially, the ECU has to be able to do the calculations. It uses the basic
parameters of cylinder capacity, injector size and rpm to calculate how
much fuel to inject to obtain your target AFR. It then compares the actual
amount of fuel required to obtain the target AFR with the theoretical
amount of fuel that it calculated. The resulting ratio, expressed as a
percentage, is shown in the fuel map. VE requires that the ECU can run in
closed loop.

Since the ECU knows how much fuel it is injecting, it is possible to simply
switch the map display to ms.

If your ECU does not offer VE tuning, I'm sure it is possible to calculate
the VE external to the ECU with some simple software. Running closed
loop in ms gives exactly the same performance end result... the only
difference is the way the results are displayed. And of course, open loop
is a lot more laborious !!!!

Cheers... jondee86

[totta Crolla]

Thanks for the input and help.
Freedom is interesting in as much as it appears to use interpolation
to find the values to learn, it then changes the cells to get the AFR to
around 14.7
Turning off the learning mode allows the ecu to run in closed loop, going to open loop at a given TPS opening value.