RomRaider

One more thing just for completeness. I know the link I posted for avcs numbers is for is for a single avcs motor and that on our motor the results will also be dependant on exhaust cam timing. For the particular range that I am discussing my exhaust cam restard is set between 17 and 20 degree so exhaust should close 8-5 degree ATDC so that's really close to where it would be on a single avcs engine

Anyhow not trying to poke holes in your logic, I am just taking advantage of the fact that you seem to be one of the person that has put a lot of thoughts into this and is most willing to discuss it

Thanks !

I believe those curves come from that large AVCS thread on NASIOC. I think I've seen these graphs before as they look familiar. So, as far as power at spoolup, I believe those numbers are calculated from load which comes from MAF. The thing about that is, it is a calculated number based on the sensor reading. On the surface, one could say hey, the load numbers are higher, therefore car is generating more torque therefore we're getting better horsepower. Not so fast! The ECU is calculating air mass based on the flow. That is going someplace, but is it all going into the cylinder? Given the high overlap, how much of it goes out the exhaust side? High overlap time, lots of time for the air to escape out the tail pipe. We do know that if you advance the intake thus creating large overlap, at low to mid RPMs, the pressure differential favors the intake side. You can bet some of that air will escape out the exhaust. Air wants to flow from high pressure to low pressure. That's just simple physics. So, problem is, we don't know how much of it is trapped. I believe most folks forget that MAF may show the flow but it doesn't mean it's all going in for combustion. To truly confirm this, you'd need a real dyno to show if the car is actually making more/less power in these scenarios.



I have noticed a delay in AFRs but I am seeing the opposite. I get a slight rich dip while accelerating. My fuel targets eventually stabilize and more or less hit the desired targets. I believe my behavior is due to tip-in enrichment which I have modified to a point where the car has very good throttle response but I still get that slight dip. I believe there may be other tables in play that are not exposed to us. If I reduce my tip-ins further, the car becomes sluggish and overall driveability suffers so I am going based on feel and my general perception rather than the number off the gauge. If the car feels like a turd, what's the point? I am pretty satisfied with my fueling at this point. I think people tend to get obsessive about these things a bit too much and try to get it perfect. It will never be perfect. It needs to be close but some minor deviation is OK.

One note, I find it odd that you're looking at boost in relation to your fueling. I think you should be looking your load as same boost in August can put you in very different loads than the same number this time of year. Load vs. fueling is the way to go as that is what the ECU is mapping.

Hey,

Yes this make perfect sense, taking another look at those graphs I agree with your idea that it's not really possible to know how much air get trapped in the cylinder from those graphs alone. Thanks for pointing that out ... I guess you could integrate MAF over the non overlap period to get some idea though. Easiest way is with a dyno, I have lowered my Intake advance by 10 degree and I think the car feels better but I couldn't really measure a big difference in VE or hp with just a road dyno.

I hear you on non exposed tables. I started learning how to use IDA and did simple things like exposing my per gear timing tables etc, I might try to dig in and understand the fuelling logic a bit more next time I have a bit of a break.

Thanks for taking the time again, I enjoy reading your post, I end up learning something new every time. I agree with monitoring load instead of boost, there wasn't a lot of temperature swing until recently and yeah boost target get adjusted accordingly.

On a side note, I just recently switched to CAN logging and do I ever like the added resolution. It allowed me to pinpoint some load site where I was experiencing little knock much more precisely... Wish I had done it earlier. I also added the CEL feedback knock feature that Merp posted the code for, what a great feature. It helps me feeling more confident about my tune when I don't have a computer hooked up

Let me know if you are interested in getting it ...

UPDATE: posted rev 65 which has minor changes from rev 63 located in CL load comp tables A and B for addressing an infamous GR stumble as well as improve cruise gas economy.

Hello Guys

I have read your complette tread. Great information! I'm new here, and new in the subaru map scene. I have a general question about the 2.5l STI-Engine.
Our local subaru tuner tell me, the 2.5l STI engine is very difficult to map. The engine/piston melt before the ECU reports "knock"!
I ask him too, if we can work with a external knock control unit (phormula etc). He mean: "thats also not possible, Piston melt some points bevor the knocking". You must know the right "cylinder-pressure".

Can you tell me, as you map these engine's if these story is true?

Thank you and best regards

Marcel

P.S

I hope you are understand my english, is at the moment not the best

Salut Marcel,

Great thread hein ? The map author is actually pretty good at sharing information and discussing about some of his strategies. It really helped me getting started. So to answer your question, and take it for what its worth since I only tuned 1 car that has 10 000 km on that tune, my understanding is that the part that usually get damaged are the cast pistons ringland when they hit the cylinder wall during a knock event. If you look at alot of broken bottom end, you'll see a cracked piece of piston, which I think is the result of impact and not melted material. (I have no expertise to back that statement however).

Now assuming that I am right and that knock breaks our piston. I think that you can trust the engine built in knock detection/correction hardware/software. It has to work because the engine has to effectively be able to adjust to 91 octane gas without grenading itself on stock tune (it drops timing, I have seen it on my stock setup). I have also been able to sucessfully rely on the knock detection to tune my car to the point where it doesn't knock very much (-1.4 FBKC every once in a while, not consistent). I think it works

I recommend that you take the time to read and learn about that stuff a little bit. If not to do it yourself at least to be able to peer review what a tuner will give you if you get your car tuned. There are alot of people that have a very good reputation that actually do some pretty scary things. Which actually convinced me to start tuning my car myself because, by looking at some of the stuff some tuner do locally here I felt like, jeeesh these engine are actually tougher than people say

Anyhow, also for what its worth, after 4-5 month of tuning. I find this map pretty reasonable in terms of how safe it is. I am currently running about a degree less of timing and slightly richer but close, my plan is to work my way closer to Zyggirama numbers slowly.

Hope this help

Just wondering... will it make more power if I run more boost, but lower timing few degrees?

hey guys its been a while. I can see OP is still hard at work and i thank you for all your dedication. Dont know if you remember my situation but Ive pulled out the grimmspeed bcs and put the stock one back and Im boosting normal. Im leaning towards a faulty bcs, however i would like to try running the bcs configured the same way as the stock bcs in bleed mode to see what happens. its been cold so been kinda lazy...

anyway to answer your question magnet, i think the turbo will dictate whether higher boost and less timing is more advantages vs less boost more timing. A compressor map of whatever turbo youre running will give you your answer. Im just going to assume youre on the stock turbo, whatever model that is. I dont know the ej25 quite as well as i do a 4g63 but from what i do know, the stock turbo is relatively inefficient at the 17-20 psi range. Someone correct me if im wrong. way too tired to find the turbo model and corresponding compressor map tonight but maybe in the next few days.

Thanks for the input the turbo is VF48 to be exact.

The big question is where in the rev range you'd like to do that. Mid range or top end? Mid range, yes, you can probably raise the pressure and you'll gain some torque. The map is already set up to have pretty good mid range but if you want me, you can set it up that way. As you approach 5K RPMs, no, you will probably not make more power. The turbo is the limiting factor. As the engine spins faster, you consumes a lot more air and so the turbo has a much harder time maintaining the pressure. If you dial in higher pressure, you will overspin it which will put it well outside its optimal range. You'll end up heating the charge a lot more than before. Hotter air will mean less density, which means less air mass, which means less power.

This is good advice. Take active interest in the health of your car. If anything, it will allow you to review what your tuner did and maybe learn a thing or two about your engine.

Sorry guys, I've been absent lately. Between work, home and taking care of my new born daughter, I've had little time to do much of anything else.

ziggy, simply unacceptable. you owe us your undivided attention and we will accept nothing less

JK thats cool. I think i can say with confidence that we all appreciate your time and dedication and willingness to help fellow members when you can. Keep up the good work with your other obligations and share with us whatever extra time you can

Hey Ziggyrama,

Any new updates?

Not much new to report. The current ROM is pretty much complete. I've been riding on it for a while and it is working very well. The idea is to be done at some point and just enjoy driving the car

That being said, a ported and polished exhaust manifold along with Grimmspeed crosspipe just arrive the other day and it has been begging me to install it. I don't expect any tuning to be required for this but I may have to adjust boost control settings. If that happens, I will post a new rev specifically dedicated to this setup. The P&P manifold along with a cross pipe is supposed to improve low and mid range efficiency which could affect how wastegate settings. I think I will attempt the install tomorrow.

Hey Ziggyrama,

Hope all is well. I am curious about the change you made to address the stumble. I was diff'ing rev 63/65 version and as you mention, you seem to only have modified the CL load comp table. If you follow the big thread on stumble, I seem to recall that people only talk about modifying Engine load comp cruise/non-cruise and come people even suggest the fuel pump DTC table. I assume the lean spot is there all the time, not just for CL. So I am curious

I get a bit of feedback knock in the 3-4 region and I have to run sensibly less timing in that region to keep it at bay. My next strategy was to try to work the lean spot, to see if it improves. I was planning on working the load comp table first using airboy spreadsheet. Then maybe look at the fuel pump if I can't get there ...

As always thanks for your time