[Dean]

So looking at maps, I see a whole lot of spikes, pot holes, jagged/steep walls and plateaus in the stock, Cobb and custom ones I've seen so far. At least in the 3D maps. Most 2D maps are pretty smooth.

I can understand the plateaus on certain maps, but the other stuff is slightly mind boggling. Are people smoothing their maps out. I would think these jaggies and such would make any interpolation the ECU has to do potentially inaccurate or at least imprecise.

For discussion, take the Stage 2 v110 rescaled map which is available for multiple years.:
Dyn Adv. A or B: 200RPM, 1.3-1.9CL Spikes
Dyn Adv. B: 4000 RPM, 2.5-3.4CL Wall
Heck all the advance tables look like some form of jagged mountain range.
WGDC low: 3200-3600RPM Great Wall of steps, etc.

Even though I have a lightened flywheel I appear to have fixed my misfile CELs by smoothing my dyn. adv. tables out a bit. (Though it could have been the intake tuning I did at the same time. I should know better than making 2 changes at once.)

So how smooth are your tables, and any thoughts on smoothing?

The primary ignition table is very nice except for the couple ski jumps in the extended CL section which are excusable and a dimple here and there. Boost targets are also very smooth. Why wouldn't advance, wastegate and fuel tables be equally smooth? (fuel isn't bad, just abrupt cliffs) For those of you who have been doing this a long time, care to share pictures of your maps, comment?

[STLSTI]

Good thread I need to work on it. smoothing that is

[06naSTI]

Ya I noticed that my graph is very spikey. I just got tuned this weekend.

[flatthump]

Haha have you seen the stage 2 AVCS map 3d graph? It looks like a crumpled piece of paper

[WolfPlayer]

FIRST OF ALL ...

Don't call that my map. It's Cobb's map - just rescaled. No way would I have a map like that. If I ever release a map that is really my map, it would have smooth 3D graphs.

t

[Dean]

Originally Posted by WolfPlayer

FIRST OF ALL ...

Don't call that my map. It's Cobb's map - just rescaled. No way would I have a map like that. If I ever release a map that is really my map, it would have smooth 3D graphs.

t

Sorry, fixed. So what do your tables look like? JPG without labels & scales won't give to much away.

[crystal_Imprezav]

Originally Posted by WolfPlayer

FIRST OF ALL ...

Don't call that my map. It's Cobb's map - just rescaled. No way would I have a map like that. If I ever release a map that is really my map, it would have smooth 3D graphs.

t

Can we say defensive

[WolfPlayer]

Originally Posted by crystal_Imprezav

Can we say defensive

Defensive.

Now, with all that said, I also need to say that those maps aren't really Cobb's maps either. Don't get me wrong, they are Cobb's maps, but the Dynamic Advance maps are stock - that means that they are Subaru's maps

t

[daemon]

Originally Posted by WolfPlayer

that means that they are Subaru's maps

Yeah. I always looked at some of those stock maps and thought WTF. But then Subaru has the resources to hire some great tuners, maybe they know something we don't. Hahahahaha... Or maybe the tuners forgot those maps were there and they just ended up being the random garbage that was in the memory chips at the time. Or, sun spots flaring at the time of the ROM burn? (that is a joke)

[jaxscuby]

in a ideal world the maps should all have relative smooth transitions.
but the odd thing i observed when tuning some cars afrs was that
the afr was not a consistent number, say 11.2, but the whp & tq
graphs where so smooth i did not want to touch the fuel map.

yes it is odd that tuner maps have some peaks & valleys where
the factory map does not.

[crystal_Imprezav]

Not with an ideal and properly mapped MAF calibration

[Dean]

Originally Posted by crystal_Imprezav

Not with an ideal and properly mapped MAF calibration

I'm beginning to think this may possibly the most important part of tuning if you want everything else to line up, not that my noob opinion means much.

I think I'm up to 5+ hours of logging and probably 4X that in spreadsheet land evaluating the data. Closed loop isn't bad, but open loop is tougher. Lots of data, but so few line up and I have as yet to come up with a mean / interpolation / target lookup function to do the work for me. Any thoughts?

Today's task was WGDC table clean up to resolve an slight boost overshoot issue. More is not always better.

[R4ND0M_AX3]

Originally Posted by Dean

I think I'm up to 5+ hours of logging and probably 4X that in spreadsheet land evaluating the data.

Hey Dean, Haven't seen you on this side of the web before.

Your 4X factor is pretty close to what I spend after a one hour datalogged commute.

Most people do shoot for a super smooth map. I think the OEM maps are the way they are becuase of the time they spend tweaking these things. I'm sure it's by committee and if there is a batch of cars that have a problem in a specific area they just fix that one area so as to minimize the loss on the cars that don't have an issue. They have thousands of cars with the exact same setup. We have one car that is almost allways unique.

[Dean]

Originally Posted by R4ND0M_AX3

Your 4X factor is pretty close to what I spend after a one hour datalogged commute.

Most people do shoot for a super smooth map....

Glad to hear I'm not alone staring at the pile of numbers all night. The most frustrating part so far is is when the closed loop data conflicts with the open loop while trying to smooth out the intake map. Closed loop trims are adding fuel but open loop is missing A/F targets on the rich side...

I've almost decided that closed loop wins below 2.5v, open loop wins above 3v mostly and they battle it out for a smooth curve in between.

So, how smooth are your maps?

[R4ND0M_AX3]

Originally Posted by Dean

Glad to hear I'm not alone staring at the pile of numbers all night. The most frustrating part so far is is when the closed loop data conflicts with the open loop while trying to smooth out the intake map. Closed loop trims are adding fuel but open loop is missing A/F targets on the rich side...

I've almost decided that closed loop wins below 2.5v, open loop wins above 3v mostly and they battle it out for a smooth curve in between.

So, how smooth are your maps?

I haven't touched the fuel or ignition yet. Using cobb stage 2 stuff with streettuner. I've been working on getting my PE850 injectors to agree with the maps via the fuel scaler + injector latecies + tip-in. I'm pretty close. You may need to look at the latency values also to get the high & low load areas to jive at the same time.


I have touched the right side of the boost target & duty cycle tables. They're not super smooth but not spikey either. They're just what is needed to get a smooth boost output.