Why you shouldn't remove the "silencer" box

[cava23]

I just recently got my 05 STI so naturally I've been hitting up the forums. One of the first mods I read about was the silencer delete and I went ahead and bought one of the elbows. I installed it and, yes, I could hear the turbo and recirculating air valve better. The next hot day (yesterday) I also heard some other new sounds, knocking. Since the only thing that changed from this hot day to the others since I've had the car is the MIA silencer, I blamed it.

Some research and a few conversations with people at work later I think I have an explanation. The silencer is not really a silencer at all, the fact that it makes your engine quieter is merely a side-effect of its intended purpose which is to damp standing waves in the intake piping. Our cars measure air flow using the MAF sensor (and probably the MAP sensor for some redundancy/error checking). MAF sensors have a few drawbacks: 1) they can't tell which direction the air is flowing and 2) they really only know the air flow at one point in the intake piping. Engines which use valves have significant pulsations in the intake due to the rapid open and closing of the valves. Granted, the compressor probably does much to damp this but not entirely. So any flow in the reverse direction is counted by the MAF sensor as positive so the ECU injects fuel accordingly => rich mixture. Since the O2 sensor is too slow to tell the ECU it made a mistake in its estimate until several combustion events later (at mid to high engine speed) you end up with a significant number of combustion events which are richer than intended. But fuel inhibits knock, so clearly this didn't cause the condition I noticed.

The issue of the MAF sensor only reading flow in one place must be the culprit. There are two possibilities: 1) the elbow significantly changed the velocity profile of the air moving through the MAF sensor so that the flow model was incorrect or 2) a standing wave occurred and the location of the MAF sensor was an anti-node (high pressure, low velocity). The likelihood of the elbow changing the velocity profile enough to cause significant error in the flow model is low considering that the air first travels through the air filter before reaching the MAF sensor. The air filter probably does a good job of reorganizing the flow pattern. The likelihood of a standing wave, however, is actually very probably since that "silencer" was basically a spring/damper element tuned for some bandwidth of frequencies.... and I removed it.

If any of you are into home audio you may be aware of this phenomenon. It occurs at low frequencies in a room since the distance between walls is large. Ever notice that the bass is very loud in some places and almost not existent in others? That's a standing wave. It can be fixed by changing the impedance of the boundaries, the walls. So people install bass traps. The silencer box is essentially a bass trap, although probably not tuned for frequencies we would consider bass. The silencer is there to damp out pulsations inherent in IC engines and help ensure an accurate MAF sensor output.

Its very unlikely that the ECU could learn that this standing wave is present and its MAF sensor readings are inaccurate at certain speeds. This is because you would have to run the engine at steady state at each of these speeds long enough for the O2 sensor to be considered valid by the ECU, then long enough for the A/F mismatch to persist, then long enough again for the learn fudge factor to fully account for the sensor error. And that is assuming the learned parameters have that capability in the opcode.

So, my recommendation: leave your silencer box installed and don't call it a silencer box.

You might find a cooler name for it here: http://en.wikipedia.org/wiki/Helmholtz_resonance

Also, I have a silencer delete elbow for sale if anyone wants it. Only $50 and your car sounds so sweet!

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

[Unabomber]

Oh Lord.....you kids and your internet and your ideas. We've been doing this since you were in grade school, spare us. If your theory held water, we would have seen it LONG ago.

[kpluiten]

Wirelessly posted (LG Voyager: Mozilla/4.1 (compatible; MSIE 6.0; ) 400x240 LGE VX10000)

This is the first I've heard of knocking due to removing the silencer. You do have a very long write up though, I'll give you that.

I don't believe you are going to sway anyone here with this no matter how well thought out or researched your data is though. Also, the vast majority of people do not stay on a stock intake or tune for very long on here.

I sort of agree with Unabomber too; some very smart and talented guys have been working with these cars for years and years and I feel some one would have stumbled onto this by now.

Did you log your car to confirm knock present?

[the real suby dude]

I took that stupid box out and everything else ... Can't be a coward all your life

[theotherguy]

My head hurts now. I'm trying to install a home audio speaker under the hood of my car to try to counteract the standing wave anti node that I created by deleting my snorkus.

[Uncle Scotty]

well.....there may be something to all that....and most of that has been posted before......but

I think the real issue is just plain heat and poor factory tune and possibly a bit of lower quality fuel and if the wrong oil is being used(there are several 'popular' ones) and is vaporizing and being sucked through the pcv system and thus into the intake...THAT can dramatically lower the octane that the engine 'sees' and cause(more) knock as engine/ambient heat increases.

[fivestringslinger]

Um...If any of this were true, and these designers at Subaru went through all this trouble to design a box tuned to cancel out problems such as you describe, then why would they offer a short-ram intake branded by SPT that maintains the factory warranty?

The box is simply a silencer to minimize the NVH (Noise/Vibration/Harshness) factor to suit your average consumer buying the car as a daily driver.

I second the notion that people have been modding these cars for a very long time. Someone would have uncovered this by now.

[fastnoypi]

Originally Posted by cava23

Our cars measure air flow using the MAF sensor (and probably the MAP sensor for some redundancy/error checking). MAF sensors have a few drawbacks: 1) they can't tell which direction the air is flowing and 2) they really only know the air flow at one point in the intake piping.

if you don't have logs presenting the knock, you're only guessing.
You must take fuel quality, stock tune and intake temps into account too to make your deduction

FYI, try reversing your MAF and see how crappy the MAF meters the air, MAF direction does make a difference.

[Back Road Runner]

You can data log and test to see if your MAF numbers are funky and giving you a ragged air fuel ratio.

[Dirty25RS]

so in this pictures the guy would be the MAF?

[Gigabelova]

I also think you are overthinking the problem. The compressor wheel will significantly dampen the pulses and beyond that, I believe, removing the air silencer box hardly will change the pulse dampening properties of the rest of the abs plastic.

Any combination of a bad tank of gas, a bad tune, pcv failure, oil vaporization, slipped timing belt(An STi probably uses other means than this to measure crank position...), broken crank sprocket teeth, etc can lead to leaner afrs which can lead to predetonation. BTW, how do you hear knock? It must be really bad for you to actually hear it.

[iheartscooby]

1 WRX + 1 STI = -2 silencer deletes with a combined 120,000 miles with no problems whatsoever.

[lowrider89]

I have nothing useful to say

[NSFW]

Correlation is not always causation. If you weren't monitoring knock before and after the delete, you really don't have much evidence to go on.

If you really want to know what's up, plot AFR Error over MAFv, before and after the delete. If you have a logger and Excel, it's easy to create scatter plots that will show where your MAF is accurate and where it isn't. That's the kind of thing you can draw conclusions from.

[Rxsti7]

wall of text

[Jay89]

Originally Posted by lowrider89

to the OP.... your an idiot... you have a crappy tune and are trying to find something to blame... coincidently you just installed the elbow delete so you automatically blame it... you once again are an idiot

hey hey hey be nice. hes just trying to provide us with information based on what he found out. its sort of like what unabomber does with his manifesto except i think he actually does quite a bit more research. i wouldnt call him an idiot, but rather just mislead by his friends.

to OP
your theory sounds good but as someone somewhere on this forum once said, if you have an idea, chances are someone has already thought of it before you. till this day i find quite true. almost all the ideas ive come up with on my own for my car ive found somone who has already done it, or a kit to do exactly what i was going to do myself. dispite what everyone is saying, we all appreciate the fact that you are trying to help the community by providing us with what you beleive is fact, but i strongly suggest you do a little bit more research, and/or ask around to see if anyone else has that problem. there are a boatload more people here that drive and mod subarus here than your co-workers, friends, family, or even the techs at your local dealership or garage. if it indeed is that part that is causing the "knock" then im almost positive someone has already either posted a solution, or posted warnings about it.

[nhmtns]

Nice to see some folks with actual knowledge chiming in to a thread in General for a change.

[JohnQTard]

Originally Posted by Dirty25RS

I never thought I'd see a pic of a surfer at Munich's Englischer Garten in a Nasioc thread

[MConte05]

The theory holds water for the NA guys, just not so much the Turbo guys. Go check out the NA Forums for willaty's "Hybrid Intake" where there is lots of data to show that removing the silencer has some negative effects in low RPMs compared to leaving it in.

[Evilchargerfan]

ibtl

so this place is a real surf destination?

Originally Posted by JohnQTard

I never thought I'd see a pic of a surfer at Munich's Englischer Garten in a Nasioc thread

[el_guapo]

Originally Posted by cava23

I just recently got my 05 STI so naturally I've been hitting up the forums. One of the first mods I read about was the silencer delete and I went ahead and bought one of the elbows. I installed it and, yes, I could hear the turbo and recirculating air valve better. The next hot day (yesterday) I also heard some other new sounds, knocking. Since the only thing that changed from this hot day to the others since I've had the car is the MIA silencer, I blamed it.

Some research and a few conversations with people at work later I think I have an explanation. The silencer is not really a silencer at all, the fact that it makes your engine quieter is merely a side-effect of its intended purpose which is to damp standing waves in the intake piping. Our cars measure air flow using the MAF sensor (and probably the MAP sensor for some redundancy/error checking). MAF sensors have a few drawbacks: 1) they can't tell which direction the air is flowing and 2) they really only know the air flow at one point in the intake piping. Engines which use valves have significant pulsations in the intake due to the rapid open and closing of the valves. Granted, the compressor probably does much to damp this but not entirely. So any flow in the reverse direction is counted by the MAF sensor as positive so the ECU injects fuel accordingly => rich mixture. Since the O2 sensor is too slow to tell the ECU it made a mistake in its estimate until several combustion events later (at mid to high engine speed) you end up with a significant number of combustion events which are richer than intended. But fuel inhibits knock, so clearly this didn't cause the condition I noticed.

The issue of the MAF sensor only reading flow in one place must be the culprit. There are two possibilities: 1) the elbow significantly changed the velocity profile of the air moving through the MAF sensor so that the flow model was incorrect or 2) a standing wave occurred and the location of the MAF sensor was an anti-node (high pressure, low velocity). The likelihood of the elbow changing the velocity profile enough to cause significant error in the flow model is low considering that the air first travels through the air filter before reaching the MAF sensor. The air filter probably does a good job of reorganizing the flow pattern. The likelihood of a standing wave, however, is actually very probably since that "silencer" was basically a spring/damper element tuned for some bandwidth of frequencies.... and I removed it.

If any of you are into home audio you may be aware of this phenomenon. It occurs at low frequencies in a room since the distance between walls is large. Ever notice that the bass is very loud in some places and almost not existent in others? That's a standing wave. It can be fixed by changing the impedance of the boundaries, the walls. So people install bass traps. The silencer box is essentially a bass trap, although probably not tuned for frequencies we would consider bass. The silencer is there to damp out pulsations inherent in IC engines and help ensure an accurate MAF sensor output.

Its very unlikely that the ECU could learn that this standing wave is present and its MAF sensor readings are inaccurate at certain speeds. This is because you would have to run the engine at steady state at each of these speeds long enough for the O2 sensor to be considered valid by the ECU, then long enough for the A/F mismatch to persist, then long enough again for the learn fudge factor to fully account for the sensor error. And that is assuming the learned parameters have that capability in the opcode.

So, my recommendation: leave your silencer box installed and don't call it a silencer box.

You might find a cooler name for it here: http://en.wikipedia.org/wiki/Helmholtz_resonance

Also, I have a silencer delete elbow for sale if anyone wants it. Only $50 and your car sounds so sweet!

Just because.

[munkis]

Put your MAF sensor in backwards and see if the car starts...it wont, because the MAF sensor only reads airflow in one direction. Anomolies with air flow due to a valved engine in a turbo subaru will not make it back to the maf sensor, because there is a big intercooler, bypass valve and turbo in the way. The bypass valve also dumps after the MAF

2nd Corrections to the AF mixture are done alot quicker than you think. Short Term Fuel Trim is active all the time, if you pull a vacuum hose off your car and you were to monitor short fuel trim it will spike immediately. Over time if the vacuum leak stays there, this is transfered to Long Tern Fuel Trim. (af correction and af learning respectively on a Subaru)

Further more your car use a Front Air Fuel Ratio Sensor, this sensor reacts sharply and quickly to any changes in exhaust gas. That is why it is used to make adjustments to the AF Ratio, it is much better than an Oxygen sensor.

the end.

[cava23]

Lol. The response to this is actually less negative than I anticipated.

[AWDshenanigans]

why the elbow? just take out the silencer, reset your ecu if you want, and go on your merry way.

[kpluiten]

But, the responses are actually making some very good points.

If you still believe you are correct in your assertion, you must log the car before and after the modification, keeping as many variables (gas, temps, driving style, etc) as consistent as possible.