SSI's / M&K Sport Muffler vs. 1 3/4" Headers / Megaphones

[Jeff Higgins]

Another data point in my endless quest for 3.0 liter MFI nirvana...

I ran the car on the dyno yesterday with a set of borrowed 1 3/4" European Racing headers and cookie cutter megaphones. Mt "normal" exhaust consists of SSI's and an M&K sport muffler. The motor is a late small port 3.0 liter running on S/RS (36 mm throttle bodies) MFI with JE 10.5:1 pistons, DC GT2/102 cams, Electromotive twin plug ignition set at 25 degrees total. Here is a dyno chart from a couple of years ago on the SSI/M&K exhaust:



Here is yesterday's runs on the headers/megaphones:



As you can see, the motor lost significant torque and horse power on the headers/megaphones. That, and the A/F ratio in the low and mid range completely tanked off the bottom of the chart, down to 9:1 or richer. Granted, no special effort was made to re-tune the MFI pump to run with this set-up, but looking at the chart, I'm not even sure it could be.

"Reading between the lines" on the new chart, a couple of things become apparent, to me at least. First, it looks like volumetric efficiency suffers a great deal in the low to mid range. MFI is a "dumb" system, and will deliver fuel based on throttle position and rpm only. It does not "see" intake air volume or cylinder volumetric efficiency. In other words, if the cylinder is filling with less air at a given rpm/throttle position than it would normally be, the MFI pump doesn't "know" that, and simply pumps fuel based on the two parameters it does "see". The result is a very rich condition, as shown in the dyno chart. Yes, it runs a bit rich in that range with the "normal" exhaust anyway, but the apparently decreased volumetric efficiency of the header/megaphone combination really accentuates that.

The second thing that becomes apparent is that the motor gained no top end hp. The peak power moved up the rpm range a bit, but is actually less than it was before. I found this surprising. I fully expected to give up bottom end and mid range. But top end as well? It looks like my particular combination is not "exhaust limited", but is either intake (36mm S throttles), port (small port 3.0 heads opened up to 38mm intakes), cam (DC GT2/102) or valve (stock SC valves) limited.

Oh well - lesson learned. "The whole is greater than the sum of the parts." My 3.0 liter "torque monster" concept appears to be better served with an exhaust that keeps velocities and therefor scavenging up in the lower to mid range, and isn't limited on top by the SSI's and sport muffler.

My wife (bless her patient heart) understands none of this. All she knows is that she gets to keep the heat and the muffler. She's a happy gal...

[island911]

So, louder is not always better. Who knew?

fwiw, I wouldn't call it worse volumetric efficiency, but rather just worse evacuation/scavenging. -- when I hear "volumetric efficiency" I think only about the efficiency of the intake stroke. ...anyway, I get what you're saying. --thanks for the post/info.

I'll bet your neighbor will be happier too.

[Steve@Rennsport]

Hi Jeff,

Fun stuff, eh???

911 engines are FAR more sensitive about exhaust systems than most people know and these must be matched to the displacement and RPM operating range for each application.

In your case, those 1.75" headers were much too large for your mild 3.0 and the megaphones simply kill the mid-range without any return on top since the engine's configuration (camshafts) isn't set up for that.

A better choice would have been 1 5/8" headers and a good dual-inlet sport muffler or race muffler system that retained the "crossover".

Nonetheless, you received a good education and thats always worthwhile.

[Jeff Higgins]

Fun stuff, indeed. I am still very much in the learning curve on this, so it is a lot of fun to try different things. In this case, Brooke was good enough to loan me the headers and megaphones "for the common good", adding to our shared data about these motors. I think we both expected a little gain on top, and a bit of a loss in the mid range, but nothing like this.

It would be fun to try the combination you outline, Steve. One of these days, if I can scare up a set of 1 5/8" headers, I just might. It occurs to me now (in hindsight, of course) that maybe I should have brought a muffler along so I could do another set of pulls with the 1 3/4" headers and a sport muffler. I would have had another data point, and maybe a better idea which did the most "damage" - the big headers or the megaphones.

Oh well, maybe some other time. For now, though, the headers/megaphones are going back to Brooke, and I'll be cruising in quiet sport muffler bliss. And my wife might even start riding with me again.

[Steve@Rennsport]

Originally Posted by Jeff Higgins

It would be fun to try the combination you outline, Steve. One of these days, if I can scare up a set of 1 5/8" headers, I just might. It occurs to me now (in hindsight, of course) that maybe I should have brought a muffler along so I could do another set of pulls with the 1 3/4" headers and a sport muffler. I would have had another data point, and maybe a better idea which did the most "damage" - the big headers or the megaphones.

Those headers were much too big for your motor so no matter what muffler you tried, the results would have been the same. There are 4 key things: header diameter & length, secondary length, balance pipe, and muffler design. Any issues with these variables can really enhance or kill the power. Further, these also make big changes to the AFR's from 3K RPM to redline.

For now, though, the headers/megaphones are going back to Brooke, and I'll be cruising in quiet sport muffler bliss. And my wife might even start riding with me again.

Put her on a monthly retainer and she might,...

[teenerted1]

so a 2in-1out is in your future

[Scooter]

Originally Posted by teenerted1

so a 2in-1out is in your future

Not if I have anything to say about it.

[Steve W]

my input, fwiw, your midrange AFR went way too rich, which is why you lost so much midrange torque. That rich condition and loss of torque carries to the upper rpms with a lower peak reading. I don't know anything about space cams, but if your afrs were corrected to around 13.0, I'd bet you would make the same midrange torque as before, and see about 10 hp at 7k.

[Jeff Higgins]

Originally Posted by Steve W

my input, fwiw, your midrange AFR went way too rich, which is why you lost so much midrange torque. That rich condition and loss of torque carries to the upper rpms with a lower peak reading. I don't know anything about space cams, but if your afrs were corrected to around 13.0, I'd bet you would make the same midrange torque as before, and see about 10 hp at 7k.

I'm sure it would have helped, Steve. I think it still would have lost power, though. It would have been a fair amount of work to re-tune where the stylus rides on the space cam to optimize the results, with many, many tuning runs and readjustments of the springs. Either lots of dyno time or lots of open road time with my LM-1. The latter would have landed me in jail before I ever made enough runs to get it adjusted.

No, I think Steve Weiner's assessment is right on. My motor is built as a relatively mild 3.0, meant for good mid range performance on the street without having to ring its neck to get it. The heads/cams/valves simply will not flow enough, nor will the motor rev high enough, to require such large headers. As a result, when these overly large headers are installed, exhaust port velocities drop to the point that the escaping charge is no longer helping to scavenge and re-fill the cylinder behind it. Even if I optimized the A/F in the lower end and the mid range, the cylinder is still filling less. There is simply a smaller charge going "boom", so there is no way the torque could remain the same.

Interesting stuff, this engine building/tuning. I did build this motor with very specific intentions, to be a strong mid-range, long lived street motor. The entire intake tract, head porting, cam selection, and exhaust system were chosen to be on the "smallish" side to keep port velocities up in the mid range. Running these big headers would have required everything upstream be enlarged as well. With that would come the need to spin it up higher, to take advantage of that increased flow potential.

So, essentially, I changed only one component from an otherwise fairly well matched system, and it no longer worked all that well. There is a synergy between these components that must be respected when selecting and assembling them. I suspect, for example, that a mamoth set of PMO 50's, or 44mm RSR stacks on my MFI, would have the same detrimental result on overall power. Or throwing a set of GE100's in it. One miss-matched component, either oversized or undersized, will upset the ballance of the whole package. This exercise was a good demonstration of that. More so than I thought it would be.

Either that of my wife paid off the dyno operator...

[Steve@Rennsport]

Originally Posted by Jeff Higgins

There is a synergy between these components that must be respected when selecting and assembling them. I suspect, for example, that a mamoth set of PMO 50's, or 44mm RSR stacks on my MFI, would have the same detrimental result on overall power. Or throwing a set of GE100's in it. One miss-matched component, either oversized or undersized, will upset the ballance of the whole package. This exercise was a good demonstration of that. More so than I thought it would be.

You are a wise man Mr. Higgins,.....

Truer words were never spoken.

[kiesan]

All I know is when I followed this, umm, thing to the dyno test I was convinced my car was suddenly a hybrid and running on battery. I literally could not hear my own car. And its a 987 Boxster 3.4 with catless headers/noisy pipes, etc

[rswannabe]

Hey Jeff,

If you want to you can borrow the 1 5/8" stainless exhaust that I am currently running on my 2.8 to see what that does for you. This is the GHL built (I think) system I got through Steve and it even has heat! I am planning on pulling it off sometime in the not too distant future to put the 1 3/4" headers I lent you onto my 2.8 to see how my motors likes them. I suspect my big port (41mm), big cam (RSR sprint), and big carbed (PMO 46's) motor will respond a little more favorably then yours ;-). Can easily optimize the carbs for them as well with some jets and venturies. I see a dyno tuning session at CarbCon coming my way.

[MrPerkles]

Interesting post Mr Higgins,made me think ive got my exhaust system about right

[tobluforu]

I'm now on my 4th exhaust system for my 72, that's right 4th. First I had a bursch tripple which was loud as hell. Then I bought a used "stock" muffler and opened up the tip so it was 3 inches. Then I built a magnaflow style exhaust with Glass packs; then I bought an M&K and threw out the stock muffler because it was a rust bucket and now I have the dansk sport. Well hell, my math is wrong because that is five. I have yet to find out how the dansk is as I'm installing cams in the car and ran into some issues. But the best muffler for all around power on my bastard of a car is the magnaflow style muffler that really uses a basani muffler with glass packs with the stock a close second-stock had better low end by far, basani-good low end great mid, no top end really as the car had some ground down T cams.
Mufflers on these mfi cars really do make a huge difference.

[rswannabe]

Yes, these MFI motors definitely respond to muffler and intake changes. Here is a link to some testing I did with my 2.2S MFI motor in my 911 back in 2003 (befor ethe motor migrated into my 914).

Individual K&N filters on MFI - Page 3 - Early 911S Registry Bulletin Board

The testing results start on page 3 of the thread.

[glassp]

Originally Posted by tobluforu

I'm now on my 4th exhaust system for my 72, that's right 4th. First I had a bursch tripple which was loud as hell. Then I bought a used "stock" muffler and opened up the tip so it was 3 inches. Then I built a magnaflow style exhaust with Glass packs; then I bought an M&K and threw out the stock muffler because it was a rust bucket and now I have the dansk sport. Well hell, my math is wrong because that is five. I have yet to find out how the dansk is as I'm installing cams in the car and ran into some issues. But the best muffler for all around power on my bastard of a car is the magnaflow style muffler that really uses a basani muffler with glass packs with the stock a close second-stock had better low end by far, basani-good low end great mid, no top end really as the car had some ground down T cams.
Mufflers on these mfi cars really do make a huge difference.

Gidday toblefloro,
i have put headers on my MFI 2.4 and of course the thermostst no longer works !!!
any ideas how to make it work again and retainmy headers ??
thanks
Pete

[911st]

Interesting post. Expecially the part about a 'cross over'.

Jeff, I think you are right on about balance. With a 36mm intake and a 35mm exhaust ports if you compair there cross senction to the curtain area under the valve seats you look to be port limited on the exhaust and intake side. Cams can help a little but the straws might be on the small side compaired to the size of the glass.

Spit balling:

I am guessing a 1.5" header is about 33mm ID, a 1 5/8's about 36mm, and a 1 3/4's about 39mm.

As to the large headers w Megs: Starting with the primary tubes, going from a 35mm exhaust port to a 39mm ID tube is about 20% larger. Thus, the exhaust flow is going to have to imidealty slow down about 20% right at the exhaust port. When the gasses slow down there is turbulance and an increase in back pressure.

Then there is the megaphone. A proper length secondary can improve savaging. Going to a megaphone mostly forgoes such opertunity. On top of that its ID expands so quickly it might also increase turbulence and increase back pressure further. Seems counter intuitive I guess.

It seems that it might be that both the larger tube and the lack of good secondary tube add up to not maintaining exhaust volicity and killing any scavanging potental.


I have a lot to learn about exhausts but the cross over thing has me still puzzeled.

On an even fire 911 style motor, a cross over should increase pressure before its conection point and reduce it after. Thus if the secondary tube size is small (ID and/or length) it might help.

It just seems that placing a cross over close to the collectors would slow the cylinder evacuation and increase refersion. On a flat 6 even fire motor, each side has no or very little overlape of exhaust events. Tie them togeather with a cross over and you have two exahaust valves open at the same time on the common manifold which could increase the pressure outside the exhaust valve from another cylinder that is at its peak just as the other is closing. Thus, it seems like a cross over should hurt a 911 motor.

This cross over thing is somthing I am not understanding and I hope Steve or someone can enlighten me about what tech is at work.

What I am suspecting is that the secondary ID and length might be more important than we might think. I am wondering if a cross over effect might be more of a fix for a less than ideal secondary tube sizing.

One of the advantages of a cross over is it dose reduces how loud it might be.

No mater, if somthing works-- it works!

Jeff, I suspect if you were looking for any further improvement, if you could go to 38mm stacks with a 1 5/8's header and clean up the exhaust port ID to match -- your motor might have the potental to breath up to 10% better (36 v 38mm section delta's) and might even see up to a 10% bump in power. I made amost the same HP through 36mm 2.4S stacks on my 2.8 MFI twin plug mod-S cam motor with stock 2.4T headers and sport muffler.

[911st]

Just read one resource that says that secondary tube length can be more important than primary. Some parameters they note"

Better to small primary than to large.

Primary ideal length about 32 to 38".

Secondary should be about 12-20" before there is a significant increase in size (megaphone, muffler, or atmosphere I assume.) That this is key to TQ production.

Read another referance that for a 3.0 6 said 1.5" for street, 1 5/8's for a rally motor, and 1 3/4 for a fill race. For best full TQ curve about a 32" primary tube was best.

Anyone know the primary and secondary tube length on SSI's?

[campbellcj]

Interesting topic (I know it's a couple months old - found it in a search). I'm contemplating running my next track event with megaphones instead of the M&K "R" 2in/2out muffler I have been using for a while and really like. Engine is a twin-plugged 2.7 with Weber 46's, 1-5/8" headers and makes peak power at 7300 rpms; therefore, I am thinking it will respond well in the top-end to an open exhaust. Does this sound right? I believe the M&K "R" design is very free-flowing and lightweight relative to many other 911 mufflers so I may be chasing nits here.