Traditional to Modelled

[Rob W]

Just a quick one people.

I'm about to take the plunge and set up modelled mode. I'm aiming to get to multifuel eventually.

I'm happy with what I've got in fuel main etc. I'm going to leave the overlay table out to start with and come back to it later if there seems to be something niggly.

I'm really just asking before I go out there and do it the hard way, what's the go with the main fuel table? Can I multiply it out by a percentage as it sits and expect to get in the ballpark up to a point or is it a drive and extrapolate sort of task.

It's an EJ25 build with 272 cams and 165kPa on a 3076 size turbo. 

[Rob W]

Sorry for starting a new thread. Please delete.

I've done a full internet search and actually found the information I need in here where the forum search function didn't pick it up.

Sounds like it's best to start from scratch with the fuel tables.

[Confused]

Yep, an existing Traditional mode fuelling table won't be much help.

You could however look at one of the existing base maps, from memory the Atom base map is in Modelled Mode, this should get you a rough starting point to get you up and going.

[Rob W]

Well, I got it idling pretty easy. There was a huge jump in the cell numbers though. Went from 24 to 52.

Had to add about 70% in the 2500rpm low load area.

I added 50% to the whole map before I started and another 20% to get it close once it was running.

I reckon I'm going to need a driver for this one

I'll double check my fuel pressure first. Then I'll go and set the Fuel Charge Cooling Coefficient correctly before I get too far ahead of myself.

 

[Adamw]

You can import the fuel table from our G4X WRX11 base map which is an EJ205 using modelled fuel.  It will probably be a decent starting point.  

[Rob W]

Excellent. Thanks Adamw. I just pulled that Fuel map down. If that produces good fuel trim the spikey looking old thing I came up with in 15 minutes on the road yesterday might be ok.

 

[Rob W]

Righto. Now the really hard part of this project to work out is going to be the Charge Temp Approximation Table.

Now, how do I know if it's the charge table or the VE table. I could chase a spikey old ve table around when I should of been trimming the Charge Temp Table.

What do I have to do to know which one to interfere with?

I'm using the supplied table and figures at the moment.

Should my IAT control be left on?

My IAT sensor is about 150 mm before the throttle body inside a custom fmic pipe.

 

[Adamw]

The charge temp table from our WRX11 base map will be pretty close.  The general way to tune the charge temp table is to hold one variable as constant as possible while varying the other one.  On our subaru it was easy with the top mount intercooler, with my intercooler fan off I could just hold it at constant load/speed on the dyno, the water temp would stay reasonably static while the IAT would quickly run away.  If charge temp approximation number is correct the lambda will stay on target.  If it goes lean as IAT increases then you need more bias towards coolant.  I do that at say 6 different loads/RPM's and that gives you a pretty good feel for the trend.

IAT fuel trim should be off.  

[Rob W]

Thanks Adam. Awesome info. I'm off to have a go in a minute.

[Rob W]

So I ended up putting a value in the top left hand cell that gave me the numbers I knew worked for idle and I think 130 in the bottom right, then filled the map. That gave me a flat map to start. Then When I started driving I multiplied the whole map by the same percentage to richen it up enough to get me in the ballpark.

So I'm doing the whole things off the logs in mixture map on my own. It's going really well but.... and it seems to be in the same places as in traditional mode for a cell with plenty of samples in it.  I've got a few filters in the mixture map to stop the transients noise,  but, there will either be a spike or a hole in the map to get it to return the desired lambda.  As much as I try and compensate in the surrounding cells and blend around it the odd lump and hole is still there. It's got non avcs 272 cams in it and I always wondered if the hole was something to do with vacuum at rpm or load. I did install them with a bit of overlap initially but have since taken them back to "straight up"

The Charge temp table doesnt really seem to be giving me any trouble. I haven't set the fuel temp coeficient yet, as I'm still working through getting it all stable at lambda 1.

It sounds nice though and it drives nice. And I do understand that I won't necessarily be cruising in the cells of concern but Jeez, it would be interesting to see what it would spit out sitting on a load bearing dyno cell for cell.