06 STI VACUUM Hesitation

[mavstang73]

Originally Posted by mavstang73

Unfortunately I think this thread has become a catchall for hesitation across all late models. I think it was the OP's intention to look at just the GD series.

To the best of my knowledge the 08+ issues are something else entirely and have pre-existing solutions that have been proven.
It's the 04-07 people with dbw ej257(ej255) that are a mystery when the obvious is ruled out(air leaks, fueling, mechanical).

I'm quoting myself which is lame but I take back what I said.

After reading more about the 08STI+ stumbling issues I now believe that intake tract resonance is the root cause of the 04-07 dbw ej257, ej255 hesitation.
The MP table Phatron mentioned is the cure.

Once you hit a certain power level or mod combination were experiencing intake resonance and harmonics the factory probably worked hard to suppress at stock power levels.

* Registered users of the site do not see these ads.

[ESMBunny]

There seem to be two main "hesitation" issues;
- The 08+ "intake resonance" problem, where you have to change fuel compensation or something like that. (Mine had many hours of logging and custom tweaking to set perfect - within 1.6%)
- A more mysterious, and wider encompassing and sudden cyclical hesitation/surge problem that shows at idle and low load high vacuum situations. Daluv has that, and a few others in here.

I have experienced both, and they are quite different, though can be hard to explain in words. The second one, when it gets bad, effects the idle, the intake resonance doesn't seem to do that for anyone, though if it was a really early bad map, it might idle worse, but not like an old muscle car with a 3/4 race cam, which is how mine idles now.

(Some posters seem to have some sort of "tip in" problem or weak spot. Their idle is ok, but they get a power loss a bit different form the "intake resonance" one.)

My tuner actually was one of the first to fix the resonance problem, and I beleive he helped Airboy with the romraider patches.

I was lucky enough to get the guy who has created cars that have won 7 years of manufacturers title's in the national rally series, and 6 national wins in a row, for Subaru Canada's rally team. to tune my car. He did about 6 hrs on the dyno, and a couple on the road, and the car was running better than stock for about 8 months before my problem appeared. I think he did adjust cam timing a bit, for the car made 80 more ft/lbs, and about 60 more HP after the tune.

He says from looking at what is going on, that something has to be broken in some way. My surge/hesitation cycle does have points where it runs smoothly for a minute or so, but the ECU isn't the reason for it - no learning or compensation goes on. It's quite fascinating. I guess it means the ECU is insane, or a sensor is feeding false info off and on, that makes the ECU mess up how the engine runs, but the sensor isn't broken enough to pop a code.

I've replaced, modified, or checked all the intake system, and all the exhaust system and gaskets. I've even isolated my fuel system to just a bucket beside the car with a 320 l/hr pump on battery voltage. It's passed compression/leakdown tests, boost pressure tests, intake pressure tests, etc.

There isn't much left to replace, unless it is a really sneaky piston ringland problem. I hope, that when we find what is wrong with my car, it will solve the others who have the exact same fluctuating idle and power/hesitation cycle while driving at constant throttle, high vaccuum, low load.

If it is just the cam sensors or AVCS sensors screwing up, that would be easy and pretty cheap to fix. An ECU is a bit more expensive, I'm sure... ;-)

[PhatBotti Tuning TKR]

Originally Posted by ESMBunny

but not like an old muscle car with a 3/4 race cam, which is how mine idles now.

that sounds like a bad injector or fuel pressure problem

[hatched91]

Mavstang73, I believe u r correct. The 08+ is something else entirely. Hopefully by this weekend, I will have tried out a different ecu and cam/crank sensors

[Daluv]

keep us posted =)

Originally Posted by hatched91

Mavstang73, I believe u r correct. The 08+ is something else entirely. Hopefully by this weekend, I will have tried out a different ecu and cam/crank sensors

[mavstang73]

Originally Posted by hatched91

Mavstang73, I believe u r correct. The 08+ is something else entirely. Hopefully by this weekend, I will have tried out a different ecu and cam/crank sensors

Well... I kinda changed my mind about that last night.
After reading about the resonance issues that 08+ gr series experience I think were suffering from a similar issue.
However if your experiencing a hesitation issue at idle under no load I think there is something else at play.

Phatbotti is probably right about Esmbunny having a bad injector or fueling issue. At idle injector latency plays a big role in maintaining a smooth idle. If the coil in an injector goes bad you could experience the same type of issues.

For those with 04-07 and no issues mechanical or otherwise I believe intake resonance is the culprit. We've altered the airflow, load ranges, and components bolted to the motor.

[ESMBunny]

I had all my injectors (deathschwerks 750) do a 4 hour ultrasonic cleaning, and testing. One has a slightly less than perfect spray pattern, but flow of all is within 2%. The technician said the spray pattern wasn't something that would have any effect.
After cleaning and new fuel rails, there is no difference in the idle.

I have 2 fuel pressure gauges, and a fuel labs FPR. A new AVO fuel pump. Fuel pressure stays solid, and rises properly with boost. I've taken apart the fuel assembly three times, and used walbro 255 HP and the AVO 245 HP pumps.

Despite the new pump and fuel pressure regulator, I added in a perrin fuel rail system, ran fuel lines to a bucket beside my car, in that bucket I had a brand new Deatschwerks 320 l/hr fuel pump running form the battery at full power (not off the fuel module), and the idle was exactly the same. I didn't try driving it like that. Not enough duct tape to tape the bucket to the side of the car. ;-) I also then cut out all fuel pressure regulators and dampers, and the idle was a tiny bit worse without anything controlling fuel pulses.

It still might be the injector, because it wasn't tested in all possible ways - pulse variation at different rpms or anything, but it flowed about 780 CCs. So it isn't obviously messed up.

[Phatron]

try turning your FP up....

[mavstang73]

Originally Posted by Phatron

try turning your FP up....

And when the injectors were tested did they check them to see if the latencies are within tolerance?


Besides you have an 08 has anyone checked if your Maf compensation is out of whack recently?
Did you read through this thread? http://www.romraider.com/forum/viewt...hp?f=15&t=4911

I'm not saying that will resolve your idle issues. That's another animal.

[PhatBotti Tuning TKR]

^ That is the MRP vs RPM load trim table i have been telling everyone to try the spreadsheet to tune the whole freaking thread.....

The spreadsheet to tune that table is in this thread

http://www.romraider.com/forum/viewt...hp?f=32&t=5483

That is the same table Mike offered to send out to people 5-10 posts ago...

[ESMBunny]

Does anyone know exactly what my fuel pressure should be?

I'm trying to do this from memory - and my car is in the shop - so I think I set it to idle (with vacuum hose attached) at about 36/37 psi. With no vacuum hose connected it is almost 44. (I tried for 43.5 psi and erred towards 44).

I did have the pressure turned up a few psi, and it smoothed things a tiny bit, but it didn't change the cycle. I think the extra richness just kinda filled in the gaps a bit, for a part of the gap. The underlying mechanism seemed unaffected. Normal driving, the ECU may see the richness and just cut back, so not sure if changing pressure would do anything, and if it did, it might just be a band-aid.

Having no FPR or dampers in the circuit, may have resulted in as much as 60 PSI (I had no gauges in the circuit without the FPR in place - but I would suppose that the fuel pump with low flow would be running a lot of pressure), and even then, the car idled the same or a tiny bit worse.

[ESMBunny]

It was a simple injector test, so I doubt they did latencies.

Later I saw that they could do stuff like flow at different RPMs and such, but I hadn't asked them to do anything fancy. Didn't know I had the option, lol.

They did spray pattern and before/after flow test, and maybe something else. I would think if the coil had problems, it would be unlikely to flow as much.

I'd need to spend $500 on 4 deatchswerks injectors just to see if there is any difference... I've spent about $3000 so far, and don't want to be throwing money around as much anymore. :-) Even if I had any money left, lol.

[ESMBunny]

Hmm, I could throw in my stock injectors and see if it idles - that shouldn't damage anything. Or would it?
One nice thing about putting perrin fuel rails on, changing injectors should be a lot easier. :-)

[ESMBunny]

And yes, phatron has been telling us about the cam overlap thing since the early-ish days of this thread. :-)

I don't have the software or connections to look at my tune. I'm assuming that the guy who builds the most successful rally cars in the country might be aware of that problem. But ya never know.

[Phatron]

if you're not gonna tune the car for it, then putting those injectors in is gonna lean your car out by 25-30%

[mavstang73]

Originally Posted by PhatBotti Tuning TKR

^ That is the MRP vs RPM load trim table i have been telling everyone to try the spreadsheet to tune the whole freaking thread.....

The spreadsheet to tune that table is in this thread

http://www.romraider.com/forum/viewt...hp?f=32&t=5483

That is the same table Mike offered to send out to people 5-10 posts ago...

I know I know...lol Sorry for being a sheeple.
I had it stuck in my head that there had to be something wrong mechanical not realizing the problem was right there in front of me.

Originally Posted by ESMBunny

And yes, phatron has been telling us about the cam overlap thing since the early-ish days of this thread. :-)

I don't have the software or connections to look at my tune. I'm assuming that the guy who builds the most successful rally cars in the country might be aware of that problem. But ya never know.

Umm to the best of my knowledge that is an entirely different subject.
AVCS and the MRP vs RPM table while related (in the same way every table is related too each other) they are not part of the problem unless they are really off.

Originally Posted by ESMBunny

It was a simple injector test, so I doubt they did latencies.

Later I saw that they could do stuff like flow at different RPMs and such, but I hadn't asked them to do anything fancy. Didn't know I had the option, lol.

They did spray pattern and before/after flow test, and maybe something else. I would think if the coil had problems, it would be unlikely to flow as much.

I'd need to spend $500 on 4 deatchswerks injectors just to see if there is any difference... I've spent about $3000 so far, and don't want to be throwing money around as much anymore. :-) Even if I had any money left, lol.

If someone happens to know the ohm rating of the injector and you have a multimeter you could perform a rough check. It wont tell you whats going on under real world conditions and load but if the resting value was way off from what it should be you would know you had a problem.

[ESMBunny]

Yes - but could it idle for ten minutes like that and be ok? Just to see if it changes the hesitation cycle? Would the computer compensate richness at idle?

[ESMBunny]

Ya, I thought mavstang was talking about cam overlap. I don't do the tuning thing, so have to read links to be sure what they are about.

[mavstang73]

Originally Posted by ESMBunny

Yes - but could it idle for ten minutes like that and be ok? Just to see if it changes the hesitation cycle? Would the computer compensate richness at idle?

I wouldnt do it without changing the injector scaler and latency values.

[Phatron]

Originally Posted by mavstang73

Umm to the best of my knowledge that is an entirely different subject.
AVCS and the MRP vs RPM table while related (in the same way every table is related too each other) they are not part of the problem unless they are really off.

well....the hesitation problem was not prevalent until exhaust avcs was introduced.

And when the exhaust avcs was introduced is when the MRP load trim table came into major play. In pre-08 roms that table only has values of +-2% and some of the roms have it zero'd.....then in 08+ it has values of +20%.

AVCS directly effects the MRP. If you zero the AVCS you will lose 600-700rpm of spool.

[mavstang73]

Originally Posted by PhatBotti Tuning TKR

^ That is the MRP vs RPM load trim table i have been telling everyone to try the spreadsheet to tune the whole freaking thread.....

The spreadsheet to tune that table is in this thread

http://www.romraider.com/forum/viewt...hp?f=32&t=5483

That is the same table Mike offered to send out to people 5-10 posts ago...

Originally Posted by ESMBunny

And yes, phatron has been telling us about the cam overlap thing since the early-ish days of this thread. :-)

I don't have the software or connections to look at my tune. I'm assuming that the guy who builds the most successful rally cars in the country might be aware of that problem. But ya never know.

Originally Posted by Phatron

well....the hesitation problem was not prevalent until exhaust avcs was introduced.

And when the exhaust avcs was introduced is when the MRP load trim table came into major play. In pre-08 roms that table only has values of +-2% and some of the roms have it zero'd.....then in 08+ it has values of +20%.

AVCS directly effects the MRP. If you zero the AVCS you will lose 600-700rpm of spool.

And this is way I wish someone else no would tune my car. But alas no one returns my calls/emails.

I think the Mrp table adjustments would resolve issues with 04-07 2400-3000rpm stumbles too. For a different reason (resonance).

[spec_bg]

but why would it have the 2400-3000 stutter bone stock? stock air box and snorkel. if its in the stock tune than why isnt it more common? this is crazy, when i give it gas to go in boost in that range it wants to stutter before going. i wish we can figure this out.

[Daluv]

for the GD department, let's hope that hatched91 comes up with something

Mart

Originally Posted by spec_bg

but why would it have the 2400-3000 stutter bone stock? stock air box and snorkel. if its in the stock tune than why isnt it more common? this is crazy, when i give it gas to go in boost in that range it wants to stutter before going. i wish we can figure this out.

[hatched91]

I replaced my crank sensor, and no change. It seems to boost better, but I think it's just the ecu relearning. My cam sensor was the wrong one, I get the new ones saturday.

[mavstang73]

Originally Posted by spec_bg

but why would it have the 2400-3000 stutter bone stock? stock air box and snorkel. if its in the stock tune than why isnt it more common? this is crazy, when i give it gas to go in boost in that range it wants to stutter before going. i wish we can figure this out.

Not sure. I didnt have the issue in stock form.
I'll have to think about that one more.