glapsson wrote:

redeye wrote:Set 10 degree @ 3000rpm and 60 lq and do a data log should give the answer

10 degree -- it is ori

Set it for check. This will see the degrees

Im viewing from my phone at moment if its 13 degree ori try bigger change for test set 5 degree and do log just example you can then check if axis = same as log

Ori is 10 in 3000rpm above 55mg

my is 13 in 3000rpm above 55mg

I think pilot injection blocked SOI map

may need to change as

but factor in pilot injection is 0,01 ...... is it ok?

In SOI factor is 0,023438

????

Different factors for the same function which is relative to crank angle? Sounds fishy to me, but I might be wrong.

I guess if the DAMOS are reliable then pilot SOI factor will be obvious from there (pretty sure I checked all SOI and they had the same factor 0.023...)



Bit confused what is going on here but I'd be more inclined to say that the three or so maps we conventionally call 'SOI' maps are actually the limiters, ie, the max allowed values (the ones with 20 or so deg @ 5000rpm and 55mm3 IQ)


I've been told that there are a lot of 3d/2d maps and some logic which operates to determine actual SOI and SOI quantity of the three injection phases (pilot, main and post)

It's really not simple and from the sounds of things they are somewhat inter-twined when it comes to calibrations, so making main SOI earlier might bump into pilot injection and so on... could get messy quickly unless you have the full understanding of the process of determining IQ's and SOI's to make sure what you want is actually given.


I suppose there comes a point where change and observe works, but when the system gets so complex it really isn't ever going to be fruitful to use that approach.

This is where a technical approach might work better. Ie, learn how to disassemble code, or invest the time on a dyno with an OLS300 owner who can let you sit for a day and watch the SOI values be calculated off the appropriate maps :)


Hmmm, still all very interesting. Five years ago there was nothing like this on forums, so in five more years we might finally have flow-charts for the entire ECU processor logic and do things perfectly hehe! Here is to hoping!


Dave

ross2482 wrote:Dave I think we had some emails bounce between us back in 2006 or so over on one of the 306 forums, comparing HDi files - I've still got a few of them on my PC!

Yep, September 2006, haha... wow. Have I really been tinkering with HDi's that long! Eeek!


But back to SOI/logging. I think what is probably always an important thing to do is plot what you have, decide why it's not what you want, and then decide what you want, and then work out how to get it hehe.

So going out and logging and then tweaking overall for more power, then logging again and trying to figure out what SOI has changed. Then adjust one SOI related map and see how that impacts things. You might start to get a picture building up.


It's impossibly hard though, most SOI maps in the damos are IQ vs rpm but no other header data so you have no idea what they do or any thing. I've not seen one SOI related map (ie, outputs are in SOI range with SOI factor applied, or are in the 0...1 range for a factoring map) that has engine coolant temp on one axis like one of those posted above for example.


SOI is a nut well worth cracking though!

Dave

glapsson wrote:but factor in pilot injection is 0,01 ...... is it ok?

In SOI factor is 0,023438

????

Factor is 0.0234 for pilot injection, same as main injection. Confirmed by logging and comparing to maps.
Thanks for the file in PM by the way.... Are you sure some of the maps are correct though? For example I'm sure things like SOI and rail pressure should be set 2D inverse... ;)
Also which engine is your SOI maps from? Looking at those you've posted above, they seem way retarded compared to stock. At 55mm3 and 4000rpm mine is about 19 degrees (From memory) and 4500rpm about 23 degrees.

glapsson wrote:Ori is 10 in 3000rpm above 55mg

my is 13 in 3000rpm above 55mg

I think pilot injection blocked SOI map

may need to change as

have a study in this alfa damos a lot of the maps follow the same pattern as edc15c2 you will see also the that the soi limiter is
rpm vs iq not rpm vs temp like in the vag edc15p it looks like you are using mm3 axis for for both x and y you pic above ?

I know it's my fault mm3 factor is 0.0084 mg
but it was about SOI

So what is 2D inverse setting in WinOLS?

3D maps are basically rows and columns, defined by linear strips of data (the spikes in 2d for instance) for each row or column, and the 2D or 2D inverse setting is basically saying if those strips are building up in column or in row?

Thus an 8x16 map could be either 8 rows of IQ vs 16 of rail pressure for example, but in 2D inverse it would be 8 rows of rail pressure vs 16 rows of IQ?


Just trying to get my head totally clear before thinking more.

Is there some binary flag to say if a map is inverse or not? Maybe something in the values like signed data?


I guess it's pretty hard to work out but why even bother having 2D inverse maps, why not just write the data with less rows and more columns or vice versa? Why not always say a linear strip is ALWAYS a column and each recursive linear strip is ALWAYS a row?!

Just another peculiarity of stored data, made even harder when axis are shared across many maps and so hard to reference!

Dave

ross2482 wrote:

glapsson wrote:but factor in pilot injection is 0,01 ...... is it ok?

In SOI factor is 0,023438

????

Factor is 0.0234 for pilot injection, same as main injection. Confirmed by logging and comparing to maps.
Thanks for the file in PM by the way.... Are you sure some of the maps are correct though? For example I'm sure things like SOI and rail pressure should be set 2D inverse... ;)
Also which engine is your SOI maps from? Looking at those you've posted above, they seem way retarded compared to stock. At 55mm3 and 4000rpm mine is about 19 degrees (From memory) and 4500rpm about 23 degrees.

PS, are you generally finding that across that characteristic SOI map and pilot SOI map that PP readout values trace those maps values quite accurately?

I suppose it's hard to tell due to the sampling rate being quite low...

Dave