Non performance, driveability style tuning, cl/ol delay etc.

I'm new to ECU tuning with RR and I'm just curious if there are any other easy things to do before messing with boost or fuel? I haven't really poked around in the ROM that much, just on here and RomRaiders forum reading and trying to soak it all in.

 

I zeroed the cl/ol delay table already. Are there any other "must do" little things I can adjust? 06 XT is bone stock. PLX Wideband is going in soon, just have to take it out of the old turbo legacy.

If you zero out the delay, you might want to change what AFR it goes into OL depending on where/when you want it to go into OL on the fuel map.

If you decide you want to go the etune route from the bests let us know! We do offer etuning via open source (rom raider and tactrix cable).

 

-Brian

If you zero out the delay, you might want to change what AFR it goes into OL depending on where/when you want it to go into OL on the fuel map.

 

What might be the benefit of this? Is it so it's not switching over to open loop all the time if I'm not being smooth with the throttle?

Its independent of throttle input/smoothness. If it is set to switch to OL at 14.41, you could be in OL a lot more than you need to be. In most cases, you want to set it so that it switches to OL when you want to actually be in OL. Is there a reason why you zeroed it in the first place?

I am experiencing a hiccup when I get into boost, so I decided to zero the delay to see if it would help. After this weekend of driving though it's seems like it's back.

See post #3, there's a reason some of us leave this to the experts...

Hehe well, yea, some people do leave things to experts. But I can fix it when I break it, and I want to learn.

I suggest you ask TA for guidance and buy your parts from them when you need them. Customer service is priceless.

Guys, this isn't a classifieds "WTB tuning solutions" thread.

 

If you decide you want to go the etune route from the bests let us know! We do offer etuning via open source (rom raider and tactrix cable).

 

-Brian

 

Max, you're being tactless.

I don't think I'll use Tuning Alliance, I'm a fickle consumer and this little bit of prodding has swayed me go elsewhere. Sorry TA.

 

See post #3, there's a reason some of us leave this to the experts...

 

I suggest you ask TA for guidance and buy your parts from them when you need them. Customer service is priceless.

You're gonna wanna post some logs if you need extra help... a hiccup going into boost sounds to me like you might be getting some timing pulled on tip-in, but it's hard to say without logs.

Good Luck, if you get your tuning wrong, the cost can be pretty high.

 

Just trying to help. I know I don't have 20,000 post on here so maybe I've never seen what can happen...you may want to ead covertrussian's threads, he's got a good handle on what he's doing.

I know I don't have 20,000 post on here so maybe I've never seen what can happen..

 

:lol::lol:

biggreen96, you sound as stubborn and as bone headed as I am, They are giving suggestions for those that are not willing to learn and are not careful when doing changes (some believe if you have to ask, you should change, but I believe we all start somewhere). Since you seem to be willing to learn I think you'll be fine as long as you don't do something stupid

 

Here are some links to the tune changes that I did to my stock car, when it was stockish. This covers a lot of Subaru's emissions stupidity removal .

 

Closed Loop to Open Loop Delay Disable

Timing Compensation Zeroing

Disabling AF Learning for Open Loop

Max Airflow Learning Restricting

 

Its independent of throttle input/smoothness. If it is set to switch to OL at 14.41, you could be in OL a lot more than you need to be. In most cases, you want to set it so that it switches to OL when you want to actually be in OL. Is there a reason why you zeroed it in the first place?

 

Good call on lowering the OL switch. I just checked and I'm still running 14.41, 2007's run 14.00. I probably was busy testing other variables before lowering this value myself (hard to do accurate gas mileage testing if you change more then one thing).

 

As for why zeroing out the CL to OL Delay, I'm sure I know why this graph will tell anyone that knows anything about AFR's on why....

http://i160.photobucket.com/albums/t188/covertrussian/Cars/05%20LGT/ECU/CL%20to%20OL%20Delay/CLtoOLDelayDynoGraphpng.jpg~original

That's crazy how long it takes AFRs to settle with the delay!

 

Here's a good tuning guide. It's not everything there is to know, but it has a ton of good info

 

https://sites.google.com/site/asubienewbietuningguide/ver-1-0

 

Dave from Cryotune recommended some books by Greg Banish and Ben Strader that help lay the groundwork for tuning yourself. Not everyone wants to leave it to the experts, because you don't gain any practical knowledge that way.

On the other hand, I paid for 2 tunes from Cryotune and my car runs great. Now I have a number of revisions I can look over along with all my datalogs. I'm able to go over the revisions to see how Dave got to that point.

Best of luck and keep us updated.

Thanks guys!

I was logging 2 nights ago just to see just how often the car would switch to open loop during my errands at the stock 14.41 AFR threshold, and the damn hard-drive started grinding away and the thing crashed! Soooo a new SSD is on the way. Better to fail while it was logging than a ROM write I guess.

Thanks guys!

I was logging 2 nights ago just to see just how often the car would switch to open loop during my errands at the stock 14.41 AFR threshold, and the damn hard-drive started grinding away and the thing crashed! Soooo a new SSD is on the way. Better to fail while it was logging than a ROM write I guess.

 

Keep in mind it's not in open loop as soon as you hit 14.41. For example, my map is set with the minimum switch for OL at richer than 14.00. However, monitoring my ECU in real time will show AFRs of 13.6-13.9 or so while still in closed-loop as long as my calculated load is below 1.0 g/rev.

So as soon as I get to see those logs after I fix my comp, would it be reasonable to expect to see afrs of around 14.00-14.30 while still in closed loop(when looking at load below 1 g/rev) since I haven't adjusted the switch yet?

So as soon as I get to see those logs after I fix my comp, would it be reasonable to expect to see afrs of around 14.00-14.30 while still in closed loop(when looking at load below 1 g/rev) since I haven't adjusted the switch yet?

 

Your AFRs should be near your target (plus-or-minus any other compensations) while you're in closed-loop, unless your learning + correction can't adjust accordingly (which generally points to mechanical problems if you haven't modified the tune, but could also point to calibration issues). Generally if you have correction or learning pegged at their maximum values (I believe default maximum is ±15% on correction, and ±25% on learning), you've got serious issues, and large learning values (>5-10%) may also be indicative of minor leaks or similar.

I used to zero out CL/OL delays by default but I find even a small value helps with fuel economy while feeling in use exactly like zero delay. Start with a third or a quarter of the stock value and experiment until you get what you like.

 

IMO the foundation of a solid tune is a good MAF scaling. Yes there are ways to change fuel and timing without changing the MAF scaling, but the right way to go is to get this perfect first and then start making other changes. Not doing this is a bit like starting on a building without first defining what a meter or centimeter is.

There are great tools available to help you and great tutorials as well, just look in the relevant sections on RR.com. Just note that Subaru MAF scalings as stock are set up a bit high, meaning the load numbers are higher than they should be. More load means more fuel and less timing, and lower load the opposite. So, as you adjust load down to a more accurate number, you will spray less fuel and run more timing advance for the same MAF voltage signal, both of which can be dangerous.

Rescaling the MAF will also give you good experience rescaling the load axes in all the tables that use load (which is a lot) which you will use when raising boost further.

 

You could also try tuning the load compensation tables, and I've written a tutorial here on that process. Just bear in mind these only need large changes when running different intakes, deleted TGVs, different injectors or a combination of.

 

The advice given to get an etune or pro tune is solid. There is a lot wrong with the stock tunes like too much timing and too lean fueling in some areas, all done to pass emissions standards. Subaru are well aware of these shortcomings but they have to sell cars, and they have to pass testing. I think you'll have to do a lot of research to find out all the details and what needs to be done about it. There's no shame in enlisting a pro to get you started on the right road.

 

If you're really determined to go it alone then start off with MickeyD's ROM and tweak from there. They are not perfect maps but they are carefully thought out and well tested. Get your head in there and figure out why everything that is no longer stock was modified and why it was done the way it was. Once you understand that, start tweaking to suit your own hardware perfectly.

I used to zero out CL/OL delays by default but I find even a small value helps with fuel economy while feeling in use exactly like zero delay. Start with a third or a quarter of the stock value and experiment until you get what you like.

 

Interestingly enough, I actually slightly gained gas mileage when I disabled the CL/OL delay. Tested on the same highway stretch, which does have pretty hefty hills (600ft-2200ft). I think think the CL delay was causing the car not have enough power going over the mountains thus more throttle for longer periods of time was required causing MPG loss.

 

Play around with it and see helps in your case basically.

For a minute there, I thought the purpose of this forum was to tell people to stop asking questions on the forum.

 

This might be contentious, but here goes anyway... I never liked the throttle response of my stock LGT. It was too twitchy at low throttle, and too vague at high throttle. So I redid the main DBW tables to make it feel like I want it to feel.

 

There's a table that maps the throttle pedal angle to a "requested torque" value. I just made it linear. The table is designed to have different curves at different RPM, but didn't see much point of that. It's the same curve (actually a straight line) from 0 to 7000 RPM.

 

I did however change the values at 7000+ RPM to be half the values used in the rest of the table, so that if I approach the rev limiter, the car automatically cuts throttle. It's much less abrupt than hitting fuel cut. (Of course, if you don't lift right away, you will still hit fuel cut.)

 

I also raised the rev limit fuel cut to 7200, same as the STI, since the engines are virtually identical.

 

There's a table that maps the requested torque value to a throttle plate (butterfly valve) opening angle. I changed this so that the cross-section of the valve opening increases linearly with the requested torque value. It increases very slowly near the bottom, then steeply near the top.

 

The stock throttle curve is actually the reverse - steep at the bottom, flattening at the top. So you spend most of your time with a twitchy throttle, and the top 50% of the throttle pedal range feels like it adds almost nothing.

 

If you're used to the stock setup, this will feel really vague at low throttle openings. At first, anyway. Personally I think it's much better. If you don't like it, you can always flash back to the stock setup.

The factory tune will add a crapload of fuel when load is rising rapidly. This makes for poor throttle response when boost is building - which is when you really want the throttle to respond.

 

As far as I and the others at RomRaider can tell, there's code in the ECU that is trying to model "wall wetting" (aka "'Tau") which is the tendency of some fuel to stick to the intake manifold walls rather than getting blown into the combustion chamber. And while the model works fairly well for lower-power driveability, it sucks (horribly) at high power / high airflow.

 

With enough airflow, wall-wetting is not much of an issue anyhow, because the intake runners just get blown dry (or very nearly so) with every intake valve opening.

 

These tables are in an Alpha section of the EcuFlash definitions, but IMO they're well proven now. There are actually several undocumented tables in the ECU for the Tau model, but most of them are very hard to make sense of, and not part of the problem anyway. The three that are part of the EcuFlash definition are fairly straightforward. All three act as multipliers for the Tau value computed with the other tables. The key thing about the changes I've made is that they disable Tau at high load when the engine is warmed up.

 

If you disable Tau at low load, the car stumbles when you try to raise RPM up from idle. Stoplight turned green, I gave it some throttle, bleh... nurtured it up to speed, drove home, un-did that change right away. I didn't experiment with the not-yet-warmed-up parts of the tables, because I always drive gently while warming up anyway. It never uses those low-temperature + high load cells anyway, because I never let it.

 

You can log the Tau compensation with RomRaider, I think it's "AF Correction #6." Cruise at 2500 RPM, stomp on the throttle, watch it blip while boost is rising. Or log "Primary Open Loop Enrichment" and "Final Fueling Base" and watch the latter dip 0.7 - 0.8 lower than the former. For example I had 11.2 in my open loop table, and the ECU was asking for 10.5 during spool. To make things worse, there's something else going on that pulls the actual AFR down into the high 9s, which costs a lot of power. Zeroing out the Tau at high load kept my AFR from going below about 10.5, so power didn't suffer. (I think that "something else" might be air accumulating in the intake tract as it pressurizes, but you'd need a second MAF sensor at the throttle plate to actually confirm or refute that theory.)

Also, if you're going to be flashing much, go into the "Ignition Timing - Knock Control" and change the "Advance Multiplier (Initial)" to 1.0. That way you won't have to wait for the ECU to re-learn your fuel quality after every reflash.

 

But only do this if you are consistently buying fuel that has enough octane to keep the IAM at 1.0 all the time. You can verify that with a logger or with the Learning View utility.

 

And change Fine Correction Retard Value to -1.01

And change Fine Correction Advance Value to 0.20

 

Then you can use the hundredths digit of the values in Learning View to see how many times a particular cell saw knock, even if the ECU learned the knock away. E.g. if you see -0.42 you know that it has knocked twice even though it has since put most of the timing back. This gives you more information to work with.

 

Also increase the min/max load and RPM ranges for Fine Learning Knock Correction to cover everything. For some reason Subaru only wanted FLKC active in certain load and RPM ranges. Perhaps they saw too much false knock outside of that range, but I didn't have a problem with it and, again, it gives more information to work with.

 

These aren't really about driveability, but they're handy tweaks if you're about to start tuning.

He lives! Afters years of reading your awesome posts over at RomRaider, nice to see you actively posting here again!

 

@Tau, thank you for explaining that, that's a very interesting concept and I wonder if going to plastic manifolds that are smoother and having vertical runners have something to do with this? I haven't heard of such tables for my inline 4 motor with runners are are much more horizontal, but it is an older engine.

 

@DBW tables, I really played with them a few years ago on my 04 FXT, which has the same exact stock table as the 05 LGT. Main reason for playing with the DBW tables was, while offroading FXT had way too much power causing it to lunge forward a lot more. This linear table made it a bit easier to drive in those conditions, but daily driving it didn't feel right so went back to stock DBW tables.

 

I tried a linear table on the Legacy too, but since my AVCS table was not stock, it actually required too much foot throttling. Thus I've been back to stock DBW table since then.

 

As for AVCS bit, I noticed that changing the AVCS table from 0* to 10-15* reduced the power enough in those areas that lunging was reduced greatly, especially while on cruise control. Making the car be much more pleasant.

 

Stock AVCS MAP:

http://i160.photobucket.com/albums/t188/covertrussian/Cars/05%20LGT/ECU/AVCS/05LGTStockAVCS.png

 

Modified map:

http://i160.photobucket.com/albums/t188/covertrussian/Cars/05%20LGT/ECU/AVCS/AVCS10.png