ECU Tuning for Reliability

[regal]

With all the DI failures, the head issues, etc I think its time to take a serious look at the factory tuning. Especially if we want to keep the engine past warranty.

This is an interview from Yimi Sport tuning on the EJ, and based on monitoring knock with the torque plugin I belive the factory tune of the FA20 has the same issue:

Unfortunately, Subaru chose an unfortunate ECU tuning strategy in more recent years (mostly '07 and newer) for Federal emissions compliance. Basically there is a long delay in the switchover from closed to open loop operation that forces the car to run very lean fueling even as load/boost rises. At that point, knocking/pinging/detonation is almost unavoidable and the resulting concussive pressure wave will crack the piston ring lands. In large part, this is what has caused some people to believe that the EJ motors are weak, but the underlying cause is actually in the ECU tuning.

ref: http://www.importtuner.com/tech/impp...h/viewall.html

I don't know if the Ecutek tuners are looking at knock logs and correcting these issues or just chasing HP, but I think a collected BRZ-edit effort may be worth while.

[ft_sjo]

Subaru haven't chosen anything, they are FORCED to do it as is every other car manufacturer.

Stop with these threads, please. You're on a frickin' witch-hunt. Get a life.

[Fast_Freddy]

There's one on every forum... I'm sure this FA20 will still be running great at 190k miles just like the "weak" EJ in my other car.

[regal]

Originally Posted by ft_sjo

Subaru haven't chosen anything, they are FORCED to do it as is every other car manufacturer.

Stop with these threads, please. You're on a frickin' witch-hunt. Get a life.

That doesn't mean we can't fix it, get your head out of the sand.

[arghx7]

Ok guys, time for a little lesson about stock tunes, especially those complying with US regulations.

There is a chassis dyno test cycle called the US06:



It's used for emissions and fuel economy certifications. The US06 is supposed to simulate more aggressive driving compared to the old city and highway cycles from decades ago.

The emissions problem is that big acceleration event at around 120 seconds. There are limits to how slow you can accelerate on it so you can't just weasel your way out of it. Small displacement engines and underpowered applications (think big trucks with V6 engines) have to work pretty hard to accelerate through this test. When you work the engine hard, you build up heat and the cat can experience excessive temperatures. The only way to lower temperatures is to advance the timing and/or enrichen the mixture.

BUT if you enrich the mixture too much, you will fail the standard for CO emissions. CO emissions are directly related to how rich the mixture is; if you are in a closed loop targeting lambda 1, and the cat is working, CO emissions are basically 0. So the industry trend is to delay enrichment as much as possible to reduce enrichment, which helps fuel economy and CO emissions. It's not some "irresponsible Subaru bad tune," it's also there for the real-world fuel economy that everyone complains about.

In 2008 Subaru basically doubled the enrichment delay time on a lot of cars. I haven't looked through all the certification documents but I suspect that was to get the CO emission down. What probably happened was a bunch of guys sat around in a room in like 2006 and said that they wanted to improve fuel economy and CO for 2008 model year on a lot of engine programs. Some guy gave some Powerpoint presentation about changing the closed loop delay as a main strategy to be "horizontally deployed" into multiple programs. The higher-ups agreed based on some Excel spreadsheets and some simulations and such. Then the rank-and-file guys had to figure out how to make it work. Subaru was not the only one doing this enrichment delay though--basically all modern cars do this. To meet fuel economy and enrichment requirements, expect even less enrichment in the future as combustion systems become more sophisticated.

The engine still has to meet durability targets with the enrichment delays. Clearly Subaru didn't spend a lot of time on that for the '08 WRX and STi especially. The FA20 engine was a new engine program so there's a good chance they did a lot more study on how make the engine durable, rather than shoehorning in a longer closed loop delay. Keep in mind that minor model refreshes like 2008 typically have a small budget, small engineering team, and often less experienced guys compared to brand new projects. So somebody pushed those enrichment delays internally and let it out the door with less validation and testing than on a new engine program.

[regal]

Originally Posted by arghx7

The engine still has to meet durability targets with the enrichment delays. Clearly Subaru didn't spend a lot of time on that for the '08 WRX and STi especially. The FA20 engine was a new engine program so there's a good chance they did a lot more study on how make the engine durable, rather than shoehorning in a longer closed loop delay. Keep in mind that minor model refreshes like 2008 typically have a small budget, small engineering team, and often less experienced guys compared to brand new projects. So somebody pushed those enrichment delays internally and let it out the door with less validation and testing than on a new engine program.

See this is an excellent post. Exactly what I have been trying to communicate. But there is a lot of evidence of high knock logs with the FA20. Anyone interested just needs to download torque and the knock app.

Having worked in the industry for 20 years I can tell you that R&D generally means un-foreseen issues, no matter how big the budget. I'm not saying the FA20 is a bad design, I'm saying with a tool like Brz-edit we can make it better, less knock prone, less likely to blow DI seals, less likely to wear rod bearings. Isn't that a good thing?


We aren't constrained by US06,LEV,ULEV, or even the tail sniffer. With the high compression and DI we will pass any emissions tests for as long as our rings last.

And our rings will last longer if we tune out this knock rather than sitting on out hands waiting on the OEM who is constrained by government bureaucracy.

[ft_sjo]

@regal - the problem is you don't really have a clue what you're talking about (based on this and other FUD topics you've created)..

You're assuming that all knock is bad/evil. You're assuming that 'other cars' don't knock in similar ways to this. You're spreading paranoia that this 'problem' is going to kill our cars and we 'must' do something about it. As much as what @arghx7 says may sound negative (I think we're probably in similar circles), it's pretty typical. Don't think that Subaru is 'special' or poorer-performing that other companies.

A bit of tip-in det is not going to kill your low power shopping car engine. It might expose a problem with the director injector seal material/spec, but that's a different thing.

The biggest 'downside' to this behaviour is it takes the ECU a little while to get you back up to 'full power' after a light knock event, not that'd you'd notice in the car.


Let me try and guide you so your missions can be focused and useful. Identify and understand exactly what the problem is first before you start throwing blame and trying to come up with solutions. That is what will actually help people with a measurable effect, not FUD and paranoia.

[arghx7]

Basically every modern vehicle runs on the ragged edge of knock all the time, especially if you dont have good gas in there. I'm not saying that to be negative. I'm trying to give some context. Your mom's V6 Camry doesn't blow up does it? You think an Ecoboost engine doesn't regularly knock with 87 octane in there, despite the advertised fuel requirement?

If the car is basically stock, don't worry about it. Generally speaking, if it breaks and it's stock somebody at Subaru dropped the ball on the hardware durability, not on the tune. Although the tip-out knock control thing was probably caused by somebody making a copy and paste mistake. He probably copied that map over from some other engine without thinking.