Exhaust issue - long read, but this issue is driving me INSANE

Maintenance/Repairs
, , ,
1

Australian here. I have a car problem that has plagued me for nearly two years and I’m still no closer to a solution.

The car;
Mitsubishi
Diamante (Magna)
All Wheel Drive (not front wheel drive)
2003
6G74 V6
Automatic

The problem;
Aftermarket headers make a hissing sound when installed. Not a constant hissing sound though.

If I’m in neutral the hissing sound is heard after you rev the engine - bring revs up to say 3000, then close the throttle and you hear a hissing/rattle sound as the revs return to idle.

No leaks heard at idle.

When you’re driving the sound is heard on acceleration - VERY LOUDLY. Same sort of sound - loud hissing. Not a ticking sound like you might expect from a manifold gasket though. This sounds like a dead flange gasket when accelerating under load.

If I accelerate very, very, very slowly there is no unusual exhaust sounds.

This might sound simple enough, but let me walk you through why I am confused.

On first install of Brand X headers I get this noise. So, I assume I have installed correctly and attempt to eliminate possible leaks before reinstalling them… the list of things I have tried is as follows;

  • completely remove the EGR system. Still leaks.
  • Reinstall EGR system. Still leaks.
  • REPLACE the entire EGR system. Still leaks.
  • Replace the cat converter. Still leaks.
  • replace the ENTIRE exhaust system, cat back. Still leaks.
  • Replace ALL engine mounts. Still leaks.
  • try three different brands of gaskets. Still leaks.
  • try three different brands of gaskets with a sealant. Still leaks.
  • try Remflex gaskets. Still leaks.
  • Pressure test each individual section of the Brand X headers. No leaks found.
  • straight edge the manifolds and cylinder heads. All within tolerance.
  • manifold studs checked and within tolerance (not stretched).
  • Have the Brand X headers ceramic coated. Still leaks.
  • give up and have an exhaust shop fit the headers. Still leaks
  • give up again and have a second exhaust shop fit the headers. Still leaks.
  • give up on Brand X headers and try Brand Z headers. Still leaks.
  • install stock headers. No leaks.

Extra info;
Brand Z headers make the EXACT same hissing sound at idle (revs up, close throttle / hissing sound) as the Brand X headers. But, the acceleration noise is slightly quieter , or lower pitched, on acceleration with the Brand Z headers.

  • both exhaust shops acknowledge the existence of the hissing sound. Both claim to be able to eliminate the sound. After installation the first exhaust shop denies the existence of the hissing sound, despite is still being audible.

Second exhaust shop, after installation, tells me the hissing sound is “normal”.

Third exhaust shop - yes, there was a third - says he cannot find the leak, and to bring the car back when it gets worse.

Both Brand X and Brand Z headers have two flex joints in the Y-piece. On both Brand Z and Brand X headers I had both these flex joints replaced. With Brand X I had a whole new Y-piece sent. None of this fixed the hissing sound or loud acceleration noise.

There are no heat shields on either Brand X or Z headers.

The engines compression test results are 180psi over all six cylinders with a deviation of plus/minus 5psi.

There are no other modifications to the engine of drivetrain.

I have also tried two different brands of airfilter, as well as replacing the two accessory drive belts. And, I’ve inspected the timing belt - replaced the hydraulic tensioner - and there are no abnormalities. No wear marks of the inside of the timing covers.

No wear marks on the engine bay from clashing headers.

No soot marks around any flanges or surrounding body work.

The primary and secondary inlet manifold gaskets have been replaced. The inlet manifold vacuum test is within tolerance.

So. What the heck am I missing here?! IS it an exhaust leak, if so where? Could this be an inlet leak?

If needed I will upload an audio file of the leak.

2

Has the exhaust manifold been inspected?

3

Exhaust manifolds on both Brand X and Z have been straight edged - both within tolerance.

With the Brand X headers I had a spare set of the primary section of the headers - switching between the two did not change the sound.

The Brand Z headers have a very thick manifold flange - 1/2" thick stainless steel.

I’ve installed the Brand X headers three other times on FWD Diamantes without this issue.

4

I also had some steel plate laser cut to blank off each section of the headers - primary sections and the y-piece - then filled each section with water and pressured them with compressed air. No leaks found.

I only did this with the Brand X headers. I have done this with the Brand Z headers as the hissing sound is unchanged between the two headers.

5

I see the dilemma. Leaking on acceleration and also during engine coast-down from 3000 rpm makes no sense. During coast-down, there’s no pressure in the header. So, let’s assume it’s not a leak but a high-frequency vibration from the header. Try pushing against the header pipes during coast-down to see if there’s any change.

6

@insightful

“…During coast-down, there’s no pressure in the header …”

Don’t say that!
With open intake valve(s) downward moving pistons want some air to follow them. And – trust me – it does.
With open outlet valve(s) upward moving pistons want that air to get out of their way. And – well, you know….

But during coast-down from higher revs there is a constant change of the vacuum level in the intake manifold, due to much higher friction, which fresh intake air has to deal with while the throttle is closed.
That results in constant changes of air “mass” (not “volume”) entering and leaving the combustion chamber. Which results in constant pressure changes in the combustion chamber – and when the air is pushed out through the outlet valve – also in the exhaust system. During coast-down the pressure peaks in the exhaust system can even be significantly higher than the constant pressure during full engine load.

That is where the “rattle sound” in skapper’s

“… then close the throttle and you hear a hissing/rattle sound as the revs return to idle. …”

comes from.

I can not think of anything else but a problem with the inner pressure inflicted on the exhaust system by the new headers somehow.

7

Sorry, I should rather say:

That COULD CAUSE the “rattle sound” in skapper’s

“… then close the throttle and you hear a hissing/rattle sound as the revs return to idle. …”

(not: is where the “rattle sound” … comes from)

8

Thanks HD and Insightful

I am going to try as Insightful suggests - apply some external pressure to the outside of the headers to see how it affects the hiss-rattle.

I could probably live with the sound if it came down to resonance and so on.

The ffffffffffffttt sound it makes under acceleration though, I can’t live with that.

Is there anything I’ve missed, any tricks to locate exhaust leaks, any secret series of curse words I should be doing? I’m at the stage where I’m considering immersing the entire car in water so I can watch for bubbles.

9

Can you get the sound by revving the engine? If so, then try using a garden hose to find out where it is coming from. If not, this is going to extremely hard to diagnose.

Another thought: Could this be a sound from the induction side of things?

10

+1 to CapriRacer (I was so disappointed when I got to the bottom hoping to be the first to toss that idea out)

11

With no pun intended, it sounds like the headers are “ringing” with the exhaust pulses hitting the tubing. Headers are always nosier than a cast iron manifold. You’ve done everything I would suggest short of stuffing a potato in each exhaust outlet to create high back pressure. The engine will still run if there’s a leaks and it will make a ton of noise. From your post, you’ve done the equivalent several times.

I hate to say it but you might have to either live with it or put factory manifolds back on.

+2 to CapriRacer, could be the intake… I’d check that, too

12

+3 to Capriracer. I too wondered if it was the induction system making the noise.

Is it possible to put something over each side one at a time and see if the sound changes? Perhaps a piece of leather or underhood heat insulator? From the description, I’m thinking you can do this with your head under the hood and revving the engine manually or having a friend rev the engine.

Another possibility that occurred to me is the joint(s) between the header and the cat converters. I know you said you replaced the cat converter and exhaust system. Did you replace all the metal “doughnut” gaskets?

Do you have access to a smoke machine? It might be possible to see any leak in the smoke pattern if you smoke the outside of things while the engine is running/revving.

If all attempts are fruitless, have you considered putting heat wrap on the header pipes? If the sound is just a normal byproduct of the headers, this should insulate them sound-wise.

13

I’ve installed the Brand X headers three other times on FWD Diamantes without this issue.

Did they have the same engine and exhaust components as the current car?

The car;
Mitsubishi
Diamante (Magna)
All Wheel Drive (not front wheel drive)
2003
6G74 V6
Automatic

Trying to understand if we’re comparing apples to apples. If so, then the historical experience has more weight.

Personally, I think you’ve eliminated the issue being a leak. What is the construction for the OEM header? Is it a cast iron manifold/header?

Stainless steel is much harder and will have a distinct resonant frequency that is much higher than the cast iron manifold. Just hang each by a wire and tap it with the plastic end of a screwdriver. The SS will ring like a bell compared to the dead sound of the cast part. That new header will magnify any sound going in. It may also produce new sound based on the construction (tubing size, length, thickness etc) and its interaction with the pressure pulses going in.

Audible noise is a hard issue to solve sometimes. We have systems that have very restrictive audible noise specifications. But broadband “noise” level is only one part of the equation. Not all sound is created equal. Think fingernails on a chalkboard for example. Not loud but certainly obnoxious to humans. A noise that the OEM manifold is masking can be amplified and sent out at a different frequency that is now unacceptable to humans…

If the historical perspective is valid, another possibility is that the new header is enhancing some condition unique to this motor or exhaust system. A loud valve, cracked ring, loose bearing…those sounds could be amplified by the new header. An overly restricted exhaust could also cause weird noises when the highest volume of gases are being “processed”. Have you checked back pressure with a gauge between the OEM and new header?

Anyway, perhaps some things to think about.

14

“During coast-down the pressure peaks in the exhaust system can even be significantly higher than the constant pressure during full engine load.”

Please supply a link that corroborates this statement. I’ve never heard of this. The key appears to be “pressure peaks” vs. “constant pressure.” It’s constant pressure that would cause the constant hissing of a leak, not pressure peaks. I can’t believe a pressure gauge in the exhaust manifold would read higher under coast-down than at full throttle. Here’s a question to OP: Is the sound the same under throttle and during coast-down?

15

@insightful

“…It’s constant pressure that would cause the constant hissing of a leak, not pressure peaks. …”

That’s correct!

And no one said, that the pressure peaks would cause the constant HISSING of a leak.

I said, that the pressure peaks could cause the RATTLE sound that skapper talks about in the second sentence of his initial posting!

There skapper said:
“… then close the throttle and you hear a hissing/rattle sound as the revs return to idle. …””

16

Let me start by saying that I seriously appreciate all of the input. To help with diagnosis I’ve attached an audio file - I only have one arm, so I left my phone on the strut tower while I revved the engine while recording this sound.

Now let me try and cover all of the points raised.

FWD vs AWD Diamante, and the difference in header construction with the Brand X headers; Only the Y-pipe changed, the primary manifolds are identical FWD to AWD. The AWD Y-pipe is shaped to clear the transfer case hanging off the side of the transmission. Pain in the A to fit AWD headers, on both Brand X and Z headers.

Standard headers; Cast iron primary manifold to a Y-pipe of similar construction to the aftermarket headers - 2 flex joints - but with a very restrictive collector between the two banks. The stock Y-pipe is as you would imagine - shrouded in heat shields.

The Brand X headers are are constructed of 440 stainless (I think) and use much smaller primaries and a thinner walled tube than the Brand Z headers. The Brand Z headers are constructed of heavy walled 304 or 316 stainless steel. The Brand Z headers are of HEAVY construction, with a much more simple collector than the Brand X headers.

Heat wrap - tried that as well on the Brand X headers. No change in the hissrattle.

The location or source of the hissrattle is somewhere below the engine… on first hearing it you would immediately think it was the flex joint leaking. No dice. Worse still, when you’re under the car the hissrattle appears to “move” (might be a placebo) and it could be from inside the flex joint, or inside the flanged joints where the headers bolt together. Changing all three flange gaskets on the y-pipe to Remflex gaskets made no change to the hissrattle. Replacing both flex joints (twice) made no change to the noise on both Brand X and Brand Z headers.

As an example of how the hissrattle appears to “move” - yesterday I could swear it was coming from the front bank. Put my header under the car - yep, its the front bank. Popped the hood today and, nope - its the rear bank making the noise. Stuck my head under there again and the noise sounds like its coming from behind the engine. Confusing.

I’ve been over the car with one of those stethoscopes and I cannot pin point the source of the hissrattle. I’ve wrapped the flex joints tightly with as much rag as I could wrap around them and it does not change the noise. I’ve coated the outside of the flanged joints of the headers with sealant to see of bubbles/breaks would appear - nothing.

The sound at idle might be resonance - I could accept that. But the noise under acceleration is ridiculous, it sounds like there is a HUGE exhaust leak. HUGE.

On the inlet side I’ve replaced all of the inlet manifold gaskets, and throttle body gasket - Felpro. I didnt straight edge the mating surfaces, but they cleaned up fine and there were no obvious marks. I’ve sprayed all around the mating surfaces, with the engine running, and I get no change in engine note or speed. I’ve disconnected the brake booster and run the engine - no change, hissrattle persists.

On the drive belt side (AC and PS) I’ve replaced both belts as well as the two idler pulleys. No change to hissrattle or ffffffftt noises.

The part I’m struggling with is the hissrattle is 100% identical during the stationary “free rev” check with both the Brand X and Brand Z headers. Ceramic coating the Brand X headers made no change. I’m confused how a resonance sound could be identical between two very differently constructed header designs.

The ffffffffftt sound was louder with the Brand X headers - which uses 2 bolt flanges. The Brand Z headers use 3 bolt flanges and I’m pretty sure the fffffffftt sound is somewhat quieter?

One thing I didnt mention - just thinking now - is that the stock headers, when reinstalled, with make the same hissrattle sound for the first few moments of running after installation. After that the stock headers do not hissrattle nor do they make the fffffffttt noise on acceleration.

One thing I havent tried is NEW genuine manifold gaskets. On first install of Brand X headers I cleaned and reused the existing genuine manifold gaskets. Eventually thinking that was the cause I’ve my leak I used a sealant on the gaskets… that didn’t work so I tried ACL gaskets, then ACL gaskets with sealant… then Felpro gaskets… then Felpro gaskets with sealant.

As the car sits it has Brand Z headers installed, with Felpro manifold gaskets and Remflex flange gaskets. The cat convertor has been replaced and the exhaust system from the cat convertor backward is new. The engine mounts were replaced a week ago with solid urethane mounts - no change to hissrattle or ffffffftt noises after installing solid engine mounts.

Other engine internals? Compression test is within specification. Timing belt was replaced around 7000 miles ago, and I updated the tensioner to an NTN unit. Changing the tensioner did not affect the hissrattle or fffffffft. To my knowledge the engine is completely stock, and has around 84000 miles on it. Coolant is fresh and free of oil. Most recent oil change appears free of water/metal/contaminates. No signs of leaks around the head gaskets.

I do have a noisy lifter - rear bank. All 24 will be replaced this weekend though.

One final point - I use a tension wrench to do bolts up, and follow a pattern to prevent warping.

17

“Please supply a link that corroborates this statement.”

I don’t have links, but every gasoline car engine I’ve heard with a loud or leaky exhaust system sounds louder on over-run (throttle closed & engine revved up by transmission + coasting) than under light throttle.
My understanding is that on over-run the combustion chamber is under strong vacuum when the exhaust valve opens; air “pops” back into the chamber from the exhaust manifold, which is near atmospheric pressure.

I recommend a book called “Scientific Design Of Exhaust & Intake Systems” by Philip Smith & John Morrison.

18

man. with all the effort you ve put in im almost inclinedto think you may be missing something simple , and not related to the exhaust. perhaps the induction end of things, idk. try looking outside the box of what you are thinking is the problem

19

“…every gasoline car engine I’ve heard with a loud or leaky exhaust system sounds louder on over-run (throttle closed & engine revved up by transmission + coasting) than under light throttle.”

I believe we’re talking about over-run vs. heavy throttle.

20

Wesw might be right. But, man, have I run out of ideas. I’ve even tried with/without the dipstick, engine covers, different PCV valves, cleaning the injectors and removing the front half of the airbox. Heck, I’ve even run the car without the centre muffler.

In my mind it’s still an exhaust leak. The hissrattle is momentary and caused by the swift change in exhaust manifold pressure either negative or positive. The ffffffftt noise is (in my frazzled mind) is caused by positive manifold pressure - air try to escape.

The catch is there is no obvious sign of where the air is trying to escape from. And for the sound to be constant across two brands of headers? Then refit stock headers and no sound? It’s got to be exhaust.