NSX Valvetronic Exhaust System (first ever?)

RYU

Legendary Member
Moderator
Tech Expert
Joined
1 August 2008
Messages
9,622
Location
City of Angels
I've pondered my custom exhaust system for almost a year now. I'm posting this in case others need help building their own custom exhausts. I've followed every single exhaust thread on prime and probably every other exhaust topic available on the internet. I was ready to settle with my trusty HKS muffler but, at times I wanted something louder, more aggressive yet still be completely docile for date night while not hampering horsepower or torque. The units that sound good were too loud for city driving (i.e. GT One V.x's, Gruppe M, Sorcery, etc) and all the ones that were quiet (i.e. HKS, Comptech, etc) weren't aggressive enough for those special times when testosterone gets the best of you. Not to mention, all the loud ones tend to DRONE like no other. I can't tolerate drone because it gives me a migraine.

Inspired by Mac Attack (Thanks Big Mac!) I set forth to design my own. Several versions later it now incorporates a vacuum actuated bypass valve and as well as a quarter-wave resonator. It's really has a Jekyl & Hyde personality. I'm calling it the Harmony Exhaust :biggrin:

Using the NSX's natural vacuum allows for some interesting options.

Option 1 (Throttle Pressure) - By using true vacuum the valve can open and close based on throttle input alone. In other words, when you stomp on it its immediately at full flow, full sound, and 100% gangster regardless of rpm. I've observed the valve tends to open at around 50% throttle. However, to keep the valve closed for when I want it quiet I find myself driving it like I was a Hyper-miler Prius driver. This does lend for an aggressive driving tone for the fwy on-ramps. More effort can be spent fine tuning this by controlling the flow of vacuum.

Option 2 (RPM Dependent) - You can configure the valve to open at 4800 rpms and above (aka VViS). This will open at any level of throttle input but only above 4800. I'm going to drive it more in either option 1 or 2 but I think I prefer Option 2 since it allows for brisk driving around town w/o the need to feather the throttle as much yet still have the option when at WOT.

Option 3 (Wide Open) - Leave it open by incorporating a Hobbs switch to bypass the vacuum. I'll do this when i'm at the track.

Anyway, on to the vids...

I've annotated on this vid to help you understand the valvetronic action - Option 2
<iframe title="YouTube video player" width="480" height="390" src="http://www.youtube.com/embed/tY8mDijgYJk" frameborder="0" allowfullscreen></iframe>

Typical 1st & 2nd gear WOT pull - Option 2
<iframe title="YouTube video player" width="480" height="390" src="http://www.youtube.com/embed/1mkhQrqtV4s" frameborder="0" allowfullscreen></iframe>

The ever popular tunnel video - Option 1 (yes, those people must hate me)
<iframe title="YouTube video player" width="480" height="390" src="http://www.youtube.com/embed/t6EPmZd1mmQ" frameborder="0" allowfullscreen></iframe>

DISCLAIMER:
My camera is terrible. It doesn't remotely capture the volume or fidelity of the sound of this exhaust in real life. This camera sucks!

FAQ's (because i've already been asked :))
1. Why did I use a bypass valve?
- BMWs, Lambos, Ferraris, and many other exhaust manufacturers since the 60s have implemented this as a way to control exhaust noise and flow. Why not us!
2. Why use a vacuum valve vs. an electric valve?
- Mac Attack has had much success using the electric valve. I tried this one first but the valve just did not open fast enough my intended purpose. A vacuum valve opens and closes almost instantaneously and vacuum tends to the best and easiest measure of throttle input. Either works well. Just depends on what you want.
3. F1 sound?
- LOL, I dare not say for fear of getting crucified. It's somewhat high pitched but I think the sorcery GT center exit is even higher pitched. My version is more of an unrefined fierce bark. It's pretty darn awesome (scary?) and when Kuni drove the NSX, it left me standing with the biggest grin on my face.
4. Why center-exit?
- I wanted to avoid all that extra piping to save weight and i'm in the process of making my diffuser and most exhausts sits too far low. I also think diffusers look funny on the NSX with the dual tips getting in the way. However, I could probably have built this with the traditional dual side exits I suppose. I just bought a used rear valance and cut out the center section.
5. What about Drone?
- With help from Mac Attack we were able to isolate the drone frequency. It was occurring between 2000-3000 rpm and at about 130-160hz - every exhaust configuration will yield different drone range. See attached spectrum analysis. By properly placing the valve it doubles as a quarter wave resonator when closed. Genius in its simplicity! It was actually very shocking to me how quiet this is in the cabin when in quiet mode. For the first time in a long time I realized how much wind noise the NSX has on the fwy.. imagine that. It's reminded me of my OEM exhaust from years and years ago.
6. What's next?
- This is just the "Proof of Concept" but i'm so happy with this version that i'll have a local exhaust fabricator remake it to do it up in full mandrel and proper tig welds and much thinner gauge stainless steel piping. I should save about 3lbs(?) from this alone. Right now it weights close to 19lbs because of the crappy 14-16 gauge muffler shop grade piping I used.
 

Attachments

  • RYU Drone 15to25sec.jpg
    RYU Drone 15to25sec.jpg
    45.1 KB · Views: 1,508
Last edited:
Regan, it turned out great. Good call on the vacuum actuated valve, I think I like that better than the electronic version. Next time im down that way I will hit you up so I can check it out in person.
 
Regan, awesome man! Great work!
I'd like to see a video where you slowly rev through the range that the bypass valve opens. I'm thinking one where you're just reving standing still, but reving up slowly. And maybe one where you go WOT in 3rd, start at around 4000rpm and go through the valve opening (so we can hear the tone change - I think you can kind of hear it in the 2nd video at the 19sec mark).
 
Last edited:
what do you have in your car? it redlines pretty fast.

nice sound to
 
Last edited:
Awesome. Exactly what I have been looking for. Just like the Vette NPP exhaust. Quiet at partial throttle. Screaming at WOT.
 
Thanks for the compliments gents! There are so many more tweaks I'd like to do such as try out a different muffler, or increase/decrease piping size, but i'm going to be a bit busy with work in the coming months so I might let it be for now. I need to finish that darn diffuser!

Adrian - I know what you're looking for and its probably Option 2. I have a couple of vids illustrating that but my camera (5yrs old at least) simply will not capture the tone change better. It also won't capture the higher sound spectrum since it was probably calibrated for human voice. I'll try to see if I can pick up an HD cam. The best I have now is 45 secs into the 1st video. You can hear a momentary blip for a micro second in which the valve opens briefly and a quick change in noise is noticeable.
 
RYU- I've just recently started researching exhausts for my stock '91 and we seem to have very similar requirements. This is exactly what I have been looking for...I want one! I'm not really sure of how to go about getting one like this fabbed up. Any tips? This may seem as a stupid question as I'm new to this, but is this something you are planning on bringing to market or are there general guidelines on how to build something like this for myself with the help of a local fab shop? Any feedback is much appreciated. Sounds great on poor vid by the way, bet it sounds wicked in person!:biggrin:
 
RYU- I've just recently started researching exhausts for my stock '91 and we seem to have very similar requirements. This is exactly what I have been looking for...I want one! I'm not really sure of how to go about getting one like this fabbed up. Any tips? This may seem as a stupid question as I'm new to this, but is this something you are planning on bringing to market or are there general guidelines on how to build something like this for myself with the help of a local fab shop? Any feedback is much appreciated. Sounds great on poor vid by the way, bet it sounds wicked in person!:biggrin:
I'm happy to help. I love DIYs because you can make it exactly the way you want it. If you're looking into making something for yourself then i'd start with reading these threads.

Here are some from Mac Attack and some of the calculations and theories (mostly from us) on why things work like they do.
http://www.nsxprime.com/forums/showthread.php?t=134710&highlight=cutout
http://www.nsxprime.com/forums/showthread.php?t=138291&highlight=cutout

Some additional reading on exhaust theory in general:
http://www.walkerexhaust.com/support/exhaust101/productsOfCombustion.asp
http://forums.pelicanparts.com/pors...28-whats-quiest-muffler-3-6-conversion-5.html
^ this one shows a rare cutout of a GT3 exhaust and how Porsche incorporates the bypass valves.
gt3+center+muffler+cutaway31270785264.jpg


Regarding drone specifically, I found this post #32 particularly helpful as well. I copied it here for convenience. All credit goes to "Aussie"
Thrasher, you have done a good job, well done. There are two other ways to cure the dreaded drone by using a helmholtz reflector chamber. The blank ended J pipe works by being of length 1/4 of a wavelength. It travels "there and back", hence travels 1/2 wavelength and on its return to the exhaust pipe, its "sine" wave is now out of phase with the incoming sound wave and when you add the two together, the resultant wave is cancelled out. The relationship being RPM =Frequency *60/4. The 60 is to get the minutes converted to seconds, and the 4 is because there are 4 spark plug firings each rev. Hence RPM =Freq * 15. The drone revs seem to be around 1800 rpm, hence a frequency of 1800/15 = 120 hertz.

Wavelength = Speed of sound/Freq. The sound speed varies with temperature, hence is a bit hard to calculate precisely. The muffler gas is at about 50c, not all that hot really and speed of sound in exhaust gas at 50c is 1180 ft/sec. As a comparison, at 100c, C = 1235 ft/sec. Wavelength is 1180/120 = 9.83 feet. A J pipe needs to be 1/4 of a wavelength long, hence around 2.46 ft = 29 inches. It can vary a bit from this length because it will still work over a small range as the reflected sine wave will reduce the height of the incoming wave. It seems to work over a 300 rpm range.

The other options to build a helmholtz chamber are common to those who play a guitar. The chamber has a hole for the sound to enter, that hole is a certain length, and enters a box of a certain volume. Sound enters the hole and is reflected back such that it can either add to , or decrease a note of a certain frequency. Hence a muffler can also have a box that samples a sound wave using a tube of length L and diameter d, entering a box of volume V at a speed of C . By using the correct sizing, the aim is to get the reflected pulse to be half a wavelength out of phase, the same as the J pipe achieves.

The magic formula to design a helmholtz chamber to fit inside a muffler body is:

Freq = C/2pi * [A/(V*(L+.8d))]^.5. This is for units in ft. If you use c=1180 ft/sec and inches for dimensions for the entrance tube of A, length L and diameter d, and box dimensions to get a volume V in inches3, then multiply the formula by 12.

Try this to cure drone at 1800 rpm as a first trial: Box size: 6" * 4" * 4" V = 96 Tube to pass the sound wave 1" long tube of 3/4" diameter, has A=0.4418. Use C= 1180. Freq = 12 * 1180/2pi * [.4418/(96*(1+.8*.75))]^0.5 = 126 hertz = 126 * 15 = 1897 RPM. Fairly close and would work over a range of 1700 to 2000 rpm.

The other option is to build the same helmholtz chamber but instead of putting it inside the muffler body, it can be part of the exhaust pipe. It looks like a "hotdog" and if you look at a standard G8, you will see one under the gearbox. Same as the chamber, it has a large pipe, about 4" diameter welded over the existing exhaust pipe. The exhaust pipe has a hole drilled in it of diameter d, and length L where the L is really the thickness of the exhaust pipe. Try a "hotdog" of 4" OD fitting over a 2.5" exhaust pipe, with the hotdog being 4" long and the hole size drilled to 1" diameter. The length of the hole is 1/16" inch, the thickness of the ex pipe. Work out the "annulus" volume, meaning the volume of the air space between the two pipes, V = 27.5 in3 Plugging this into the same formula above, gives freq = 124, and RPM is 1857. This quite small "hotdog" would fit easily on any straight pipe length but works best closer to the muffer end. Building several sizes would give a wider rev range to improve effectiveness.

A simple speadsheet allows you to play around with dimensions. If you look inside a Magnaflow or Flowmaster you will see they use the Helmholtz chambers decribed above. Corsa uses the 1/4 wavelength approach by having several sealed tiny tubes of various lengths to give a wider range of revs to fix the horrible drone issues.

I hope this info helps.
Last edited by Aussie; 03-26-2010 at 06:38 AM.
However, if you're just lazy which is completely understandable then I suppose when I get mine fabricated it should be fairly easy to reproduce multiple copies just the same. I'll check with the fabricator when I get a chance.
 
Last edited:
For those interested in the valve operation...Here's a video I took early on in the project when I was testing the behavior of a vacuum valve in Option 2. By design, this is a slower acting valve I borrowed from the Magnusson SC setup on my truck. It's pretty neat and I figure... if VVIS is good enough for Honda it's certainly good enough for my exhaust ;).

Who knows.. maybe at the low end i'm picking up an additional 1 ft/lb of torque because of this. :biggrin:

Again, this is using Option 2 (aka VViS) so it will only ever open past 4800 rpm. Option 1 will have different behavior.
<iframe title="YouTube video player" width="480" height="390" src="http://www.youtube.com/embed/DKZA4wXmbVo" frameborder="0" allowfullscreen></iframe>
 
RYU- Thanks for all of the info., seems I've much reading to do.:smile: I can indeed be lazy when unmotivated, but this thread has definately got the creative juices flowing for me! Regardless, this endeavour will have me going into uncharted territory. I will learn as much as I can here, but please keep us posted when you finish your project as I'm sure there will be a few of us who would just rather purchase the finished product or obtain the schematics to have a local shop fab one up. I like to tinker and what not, but sometimes I'd rather leave it to the pro's. At least now I can begin the "Know enough to be dangerous" part of my exhaust theory education!:biggrin:
 
Wow, has it really been this long since we started this? I'm waiting for your updates too! Thanks much for your help my friend.

Here's where i'm at so far with the RYU 2-Stage NSX Exhaust (lol). Still have some tuning to do with the valve control mechanism which is what prompted this thread. The "build" is done and i'm pretty happy with it. The quiet mode might be enough to get me thru Laguna Seca too. We'll find out soon enough. I need to get an EMS since I notice in WOT, valves open i'm running about 0.2-0.4+ leaner.

Changes from the prototype above:
* I've gone FI since then so I needed a lot more flow. I'm now running CT headers out to two 2.5" secondaries into a 3" straight thru muffer - this is for each bank.
* Now running two vacuum bypass valves before the cats instead of one
* Both cats are ceramic lightweight and advertised as 50-state legal but I have not yet had it smog tested yet. I don't smell that nastiness NOX so i'm probably ok.
* When valves are closed it flows as much as a stock NA1 exhaust and pretty quiet and smog legal but when open it's header > 2.5" dual secondary pipes > 3" straight thru muffer > 3" dual tips (for looks only). It's pretty flippin' loud and sounds insane but not STMPO Messiah, ear piercing loud (from the videos i've heard).
* In total I've saved about 50-55lbs when compared the stock manifolds vs. CT headers, stock cats vs. bullet cats, and stock muffer vs. the single canister i'm using. The whole thing below weighs 39.6lbs. A bit more than i'd have liked but oh well... I'm building a GT car not a race car is what I tell myself :)

So far so good but hoping to build in some cool automatic controlling mechanisms for the valves with the EMS Series 2 like you have Dave. Right now i'm using the OEM ECU VVIS solenoid to actuate the valves for street/quiet mode and pure intake vacuum to actuate track mode. I'd love to make this even more finely tuned with an EMS. I'm all about turn-key automation like you are! Right now I start the car, select street or track mode by flipping a switch and drive away. The valves open/close automatically depending on engine load/VVIS/throttle position. I can definitely see where an electronic valve has it's advantages though. It's too bad the factory ECU is so limited.

I also notice the exhaust packing breaking in. The tone has changed ever so slightly after 500 miles.

311568_10151065727491588_2023691968_n.jpg


EDIT: Btw... The wonderful rainbow welds and cross section curves are courtesy of Efren Built. The man is an artist with metal!!
 
Last edited:
Wow - That turned out very nice :eek:

Very nice design - Efficient, legal, light, and sounds awesome I'm sure! Yes, bypassing my Magnaflow cerametallic cats would increase AFR by ~0.5 too due to the better breathing.

Any drone? Or does your canister do a good job of filtering out the low frequencies?
 
I sense a little bit of drone between 1500-1900 rpm. It's a little subjective because my buddy says he can't hear it. It's definitely not enough to bother me. In terms of cruising speeds, I hear more road noise than anything which unfortunately is rather annoying ironically enough! It's a quite a good date car when in street mode. That was one of my goals checked off the list.

With this cannister and the 3" design (with valves closed) it's low sounding. My 90day trial to the program you recommended expired so I haven't been able to measure the frequencies. My guess is it's roughly a 100-400hz dominant. I have no other exhaust to compare it to since I know of none that sounds quite like it. Hoping to get some vids or have someone else drive it while I follow. I do hear some rasp at WOT throttle probably due to the cats being bypassed but still rather pleasing from inside the car. I have no idea how it sounds from the outside yet.

It sounds really good driving it in the street from inside the cabin. I'm not trying to be bias either. So far i'm pretty darn happy. All the work and expense was worth it.
 
damn, that's looks and sounds awesome!! i myself also find most exhausts to loud for regular driving. any plans for producing these for the community??

The sound clips above are from the old setup. Sorry for the confusion. Blame Dave for bumping an old thread! lol The new setup sounds significantly different and more aggressive (don't have any sound clips yet).

I haven't really thought much about producing these but I suppose I could do a one time small qty run if the interest was there. I'll have to think about it though because the time involved will be significant and we'd have to pick 00-05+ headers to base if off... it's the most rare of the header types iirc.

On a different note, I just got back from a nice round trip drive from LA to San Diego. I had a lot of fun (drone free too!). I'm really liking this exhaust a lot. I even gave a 996 911TT a run for his money ;) I flew past him with the valves wide open and he gave me a thumbs up.
 
Last edited:
the videos in the first post are marked Private...I can't veiw them. :frown:
 
If you go to Super Car Sunday this upcoming weekend, there will be plenty of cameras rolling to post your new notes. So, you coming?
 
I saw today that HKS built a 3-stage exhaust. They are using an electric valve.

If you're interested in this technology you can read more about it here. Unfortunately, they are selling this for $12,800USD

481127_10151074538128528_1180798292_n.jpg


I don't know what they were thinking about the tips but the concept and technology works.
 
Not sure where you are with the control method of this project but I have an idea that may work well for you.

You will need one of these:
http://www.oreillyauto.com/site/c/d...eyword=egr+vacuum+solenoid&pt=02144&ppt=C0018

a three possition switch and for the purpose of testing and easy programming a ZAVT-1 added to your ZT2 wideband.
http://www.zeitronix.com/Products/zavt/zavt.shtml

If you ever decide to sell this thing then the ZT2 and the ZAVT-1 would be replaced with a Zublin Engineering RPM window switch.
http://www.zublin.com/nosrpmswitch/model_ns-300.jpg

Here is how it would work, the three possition switch will allow for a valve open, closed, or auto setting from the dash. The switch in open mode would activate the solonoid and the vacuum would open the valve. the switch in the closed possition would disable the solonoid and prevent the valve from opening, the auto mode would allow the programmable RPM switch to open and close the valve at any RPM setting on the controller.

For the purpose of testing and easy programming your ZT2 can be used to trigger the solinoid at any RPM you choose and can be changed easily from a laptop connected to the ZT2 using the alarm trigger feature in the software.

I can draw up a system design and generate a parts list if you want but this will give you a good control option to run with that we can refine more with testing.

Dave
 
Back
Top