PDA

View Full Version : A New Technique for Tuning AEM Fuel Maps



DG
01-10-2013, 10:20 AM
So I invented (re-invented?) a new way to data-mine logs for tuning data. This is AEM-specific, but the core concept will apply to any tuning package.

I just installed fresh 450cc injectors in my car, replacing the 20 year old 360cc injectors. The AEM Injector Resize wizard worked very well, but it relies on the old injectors flowing exactly 360cc/min and they probably flowed somewhat less. So the fuel map needs a confirmation and retune.

I have a 3 hour log from my drive from Windsor to Toronto; almost all of it 4th and 5th gear highway cruise. That's a lot of log to sift through - but it is also a lot of data reflecting actual engine performance.

This method assumes O2 Feedback is on, and wideband O2 sensors.

Load logfile. Go to "Analysis", "X-Y Plot". Set "Engine RPM" to X axis. Set "O2 Feedback #1" to Y axis. Select point colouring by point density. And then - here's the nifty bit - make it conditional on an Engine Load that lines up with one of the lines on the fuel map.

So if there is a load cell at -5 PSI, the conditional is "Engine Load < -4.9 AND Engine Load > -5.1"

This then plots every point of O2 sensor feeback along that load line, and areas where there are more points they get coloured "hotter".

Drag the cursor to the centre of the hot spot(s), read Engine RPM and correction value off the graph, change the fuel map at that cell by that much.

Repeat for each load line on the fuel map.

DG

Greg E
01-10-2013, 11:44 AM
Does AEM have a MAF scaling and Comp tables like the stock ECU? Or does it just calculate g/s and use that as the fuel map axis?

DG
01-10-2013, 12:40 PM
Far North Racing - Stealth - Programming the Mitsubishi MAF with the AEM EMS (http://farnorthracing.com/stealth/AEM_Mitsu_MAF.html)

It never actually calculates g/sec. It's a "load" number. If you have a MAP sensor (used for boost crontrol), it will assume that the range of the MAP is the range of the MAF and use MAP units along the Y axis, but those numbers don't line up very well (on my current calibration, MAF reports about 2 PSI higher than the actual MAP)

On a trace that overlays MAP and MAF, they will track each other exactly in terms of shape, but offset.

DG

Greg E
01-10-2013, 12:53 PM
:chin:

You're giving me an idea Dennis...

iceorphen
01-10-2013, 05:16 PM
oo I like ideas proceed to the lab >_>

CoopKill
01-10-2013, 05:25 PM
Will this idea be expanded upon here, or do I need to keep an eye/ear out for said ideas?

KeithMac
10-11-2013, 01:38 PM
Good idea, I did similar with the Emanage Ultimate but had to do it manually..