[Cold start] - Exhaust leak or something else?

[DoubleV]

Hi guys,

I've noticed a strange noise recently (not sure how much time is there) that in indoor spaces like a garage or mall underground parking it's harder to be missed.

I can't find any explanation why this is possible to hear only when the cold start program is on. While the cold start program is running, if I press the gas pedal to match the same RPM, the noise goes away, but it I take my foot off the pedal, then it resumes the cold start program and the noise comes back.

I don't have a video of my car doing it, but found a video with the exact same noise:

Sounds like an exhaust leak, but can't hear it while pressing the gas pedal. Like the car in the video, when the cold start is over, the noise is gone...

Any ideas?

[tomm.brz]

maybe you have a small leak near catback or front pipe that exacerbates this
but it's kinda normal
when cold start is on, car has retarded timing, and stratified injection with lean burning
it just sounds different and some leaky gasket make it more evident because of the very retarded timing

[humfrz]

Originally Posted by DoubleV

Hi guys,

when the cold start is over, the noise is gone...

Any ideas?

Yes, I suggest you quit worrying about it, since it only happens at start up (which is quite normal).

[BrahmaBull1990]

Sounds fine to me. I would be more concerned if you heard it AFTER the cold start

Any mods to the car? Mileage?

[DoubleV]

Originally Posted by BrahmaBull1990

Sounds fine to me. I would be more concerned if you heard it AFTER the cold start

Any mods to the car? Mileage?

No mods, around 35k miles.
Apart from the noise, the car drives fine.

[DoubleV]

Originally Posted by tomm.brz

maybe you have a small leak near catback or front pipe that exacerbates this
but it's kinda normal
when cold start is on, car has retarded timing, and stratified injection with lean burning
it just sounds different and some leaky gasket make it more evident because of the very retarded timing

I've replaced the headers gaskets recently (the original ones did not look good) and it made no difference in the noise.

The noise is louder on the right side of the car (left side of the engine).
It just sounds like pressurized air escaping from somewhere it's not suppose to...

[tomm.brz]

still normal for cold start with aftermarket headers

[RZNT4R]

Originally Posted by tomm.brz

maybe you have a small leak near catback or front pipe that exacerbates this
but it's kinda normal
when cold start is on, car has retarded timing, and stratified injection with lean burning
it just sounds different and some leaky gasket make it more evident because of the very retarded timing

AFAIK the spill valve isn't actuating during cold start, so it's running port injection and open throttle (pump your brakes and notice the lack of vacuum assist if you car is manual, as automatics have a vacuum pump).

--

It DOES sound like an exhaust leak, the most likely place would be the tiny flex pipes between the two banks. You would hear it during the cold start because it's running open throttle and retarded as fuuuc, so the exhaust pulses are leaving the engine with quite a lot of energy left in them, quite a bit more energy than simply revving the engine in neutral, and you probably don't hear it when loaded because you're driving down the road and drowning it out with every other noise the car makes.

[tomm.brz]

Originally Posted by RZNT4R

AFAIK the spill valve isn't actuating during cold start, so it's running port injection and open throttle (pump your brakes and notice the lack of vacuum assist if you car is manual, as automatics have a vacuum pump).

--

It DOES sound like an exhaust leak, the most likely place would be the tiny flex pipes between the two banks. You would hear it during the cold start because it's running open throttle and retarded as fuuuc, so the exhaust pulses are leaving the engine with quite a lot of energy left in them, quite a bit more energy than simply revving the engine in neutral, and you probably don't hear it when loaded because you're driving down the road and drowning it out with every other noise the car makes.

Car starts 100% port injection, yes
but very soon switches to 50/50 PI/DI with late injection DI.. any log of a cold start would show this

Anyway I think you're right abouth the sound and makes a lot of sense,
also as reducing the throttle opening during cold start makes it quieter

[RZNT4R]

Originally Posted by tomm.brz

Car starts 100% port injection, yes
but very soon switches to 50/50 PI/DI with late injection DI.. any log of a cold start would show this

Anyway I think you're right abouth the sound and makes a lot of sense,
also as reducing the throttle opening during cold start makes it quieter

You don't need a log, the spill valve is quite audible, no spill valve loudly going click click = no DI

Not a lot of effect on sound with the throttle, the bulk of the sound in startup is coming from the high energy exhaust pulses with just a bit of intake gurgle because there is no throttle plate to shut it's hungry hungry pie hole. It's open to create less vacuum to reduce oil consumption on startup, which is a strategy being adopted by more and more manufacturers, I've even seen a chrysler TSB where they had a new PCM calibration that added a low vacuum cold start to reduce oil consumption concerns.


Other benefit of startup is those high energy exhaust pulses being dumped out of the engine warm up the cat a lot faster, which any government official appointed to an environmental ministry will tell you, is super swell.

[tomm.brz]

so the ecu is lying, and the PI duty cycle and pulse width that halves, and DI injecting 50% of fuel at an angle of 45° are just not true? honestly dont think so..

https://datazap.me/u/tommbrz/io-cold...47&zoom=59-348

Engine cranks with PI only, then switches immediately to 50/50% PI/DI as you can see in the log
THEN (you can't see in the log because it's too short) it switches to 100% PI at idle for a while, THEN it switches to 100% DI on idle, and follow the ratio tables
But at this point the cold start is already over with no late ignition and no late injection

the loud part is mostly when port and DI are 50/50%

The change in sound anyone can hear during every switch of phase is pretty noticeable , like the change in idle stability because of the 2 types of injections that aren't nicely scaled to each other

[RZNT4R]

Originally Posted by tomm.brz

so the ecu is lying, and the PI duty cycle and pulse width that halves, and DI injecting 50% of fuel at an angle of 45° are just not true? honestly dont think so..

https://datazap.me/u/tommbrz/io-cold...47&zoom=59-348

Engine cranks with PI only, then switches immediately to 50/50% PI/DI as you can see in the log
THEN (you can't see in the log because it's too short) it switches to 100% PI at idle for a while, THEN it switches to 100% DI on idle, and follow the ratio tables
But at this point the cold start is already over with no late ignition and no late injection

the loud part is mostly when port and DI are 50/50%

The change in sound anyone can hear during every switch of phase is pretty noticeable , like the change in idle stability because of the 2 types of injections that aren't nicely scaled to each other

You've linked to something that is in your account on a website, for everyone else it's just a page with a youtube video of what the site does. Rather than a datalog, I put a digital oscilloscope directly on the spill valve driver circuit and was looking at it live.

[tomm.brz]

Log wasn't public, sorry, now it is

maybe you measured DI was off, when DI was actually off and PI 100%? shortly after cold start

Anyway not that i don't want to believe you, but logs says so, i can hear the injection switch back and forth, and Subaru wrote on one of those PDF years ago explaining the D4S system what I basically wrote in the last post

[tomm.brz]

here you see the switches even better
https://datazap.me/u/tommbrz/maicol-...-42-43&solo=37

100% PI for cranking
50/50 during cold start (+ late DI injection)
100% PI again during end of cold start

it will then switches to 0% PI during idle


all of this is hard coded and you see it in every brz or gt86