Performance Test done - SSI vs Stock SC
 

I recently had new SSI exchangers with a Dansk exhaust installed on my 1979 SC. The Dansk is dual in dual out. I also had the injectors replaced at the same time along with the heater box valves and oil return tubes.

I've heard figures about the performance increase related to the SSI installation but wanted to see for myself. So first I purchased a GtechPro. I also bought dB Noise meter as the difference in sound was also a concern. I made audio recordings of the before and after sound which is interesting to listen to. I plan to digitize it in the future.

Here are the Stock (except my car started with a test pipe installed) results:

Time of day - dusk
Temp -82 degF

0-60 mph - 7.1 sec
1/4 mile time - 15.3 sec
1/4 mile speed - 97.2 mph

Noise figures at rear center of car, level with exhaust:

@1000 rpm and 3' - 87 dB
@1000 rpm and 10' - 76dB
@4000 rpm and 3' - 102dB
@4000 rpm and 10' - 97dB

With SSI and Dansk:

0-60 mph - 6.7 sec
Not enough nerve to take more multiple 1/4 mile runs. I'll save that test for the track.

Noise figures same location as above:
All are average figures.

@1000 rpm and 3' - 91 dB
@1000 rpm and 10' - 89 dB
@4000 rpm and 3' - 103 dB
@4000 rpm and 10' - 100 dB

Conclusion: 0-60 increase is about 0.4 seconds. For cars starting with Cat, the results should be even better. I also anticipate the 1/4 mile improvement to be noticeable due to the free breathing at higher rpms. The sound at idle is substantially louder and deeper. The noise level surprise was that the dB level did not drop off as much going from 3' to 10' feet as with the stock exhaust. During a trip to the store, it set off an alarm two cars over when I cranked it. The exhaust sounds great except for a slight drone from 2000-3000 rpm but you can stop this by down shifting. Above 3000 rpm and at idle, the exhaust sound is just what I expected, deep growl at idle and race car like at 5000 rpm and higher. I'll post pictures once they are available. Any thoughts as to why the sound level remains somewhat constant further from the car? Could it be the increased air velocity from the SSI/Dansk?

Thank you. Very nice.

------------------
87 Carrera Cabriolet

It is pretty hard to compare early tuned exhaust to the untuned appendages hung on smogged, CIS cars!

Without the recordings and the capability to feed them to spectral analysis software knowing the specific engine rpms at various stages in the recording, it would be difficult to be specific! In general, the factory 'Banana' muffler acts like a very large volume crossover pipe connecting the two banks of the SSI system that would boost the lower frequency components of the exhaust note. Because of even fewer baffles than the factory muffler, the Dansk sport muffler is even more likely to boost those low-frequency components in the exhaust note.

The most commonly used 'A' scale for making noise measurements, denoted by a reading of 'xxx dBA,' is fairly heavily weighted towards the low frequency end of the audio spectrum, and low frequency (less than 1000 Hz) sounds are attenuated less over distance in air than higher frequencies.

It sounds like you ought to be pleased with the results of the upgrade!

------------------
Warren Hall
1973 911S Targa
1992 Dodge Dakota 5.2 4X4 parts hauler

I wonder if it has to do with the lower frequencies of the lower tone. I think the lower frequncies of the deeper tones travel better than higher frequencies. I'm not a microwave techicen so I could be all wet. I have one too and I totally agree with the great overall sound.

------------------
Mark
The Beast
mark@hargett.com

One error in my posting, the speed figures are average. The noise figures are maximum.

Considering that you lost almost half second 0-60 with 2900lb. car, must have gained 10-15 HP. Is that a good guess, anyone?

Thanks for sharing your experiences.

The drone/resonance at 2000-3000 rpm is that very annoying? I often cruise at speeds that in fifth gear lies around those rpms. Can the car still be used as a daily driver or will the noise drive you nuts?

Please make the sounds available as wavs or MP3s! I'm very interested in hearing them.

Speeder,

Here is how I see the gain mathematically:

V = 1/2 * a * t^2
V= velocity
a= acceleration
t= time

this equation can be rearranged to be:


a = 2V / t^2


since in both runs the end velocity was the same (60mph) we can say:

a2 / a1 = t1^2 / t2^2

therfore

a2 / a1 = 7.1^2 / 6.7^2 = 1.12

This means the acceleration increased by an average of 12% through the power band. Acceleration is proportional to torque so you can say the torque increase by an average of 12%. Hp is torque * rpm, so again the Hp would have increased by an average of 12% thourgh the power band. If Hp had increased evenly throughout the power then you could say that given a stock 911 SC has 180Hp peak, the peak Hp increase by 180*0.12= 21.6Hp. This seems like alot, so I think what is happening is torque in the lower rpms is increasing more than 12% while in the higher rpms it is increasing less than 12%.

Autobonrun, did you richen your mixture after the SSI's were installed?

------------------
Tony
'78 911SC

[This message has been edited by tbitz (edited 08-28-2001).]

[This message has been edited by tbitz (edited 08-28-2001).]

[This message has been edited by tbitz (edited 08-28-2001).]

According to Bruce Anderson, SSI on a 3.2 yeild about 17 hp peak based on dyno testing. And the improvement on a 3.0 is an unquantified "more".

So more probably is around 20ish hp peak improvement, consistant with tbitz calculations.

Tony,

Nice try with Newtonian physics lesson ... but 0-60 times are hardly uniformly accelerated objects!

A gain of 10-15 hp would be believable, but that would fall short of your calculations.

A 20 hp gain as discussed is starting into fantasyland ... the 2.7 CIS engine lost EXACTLY 10 hp from '74 to '75 with the reverse procedure you did. And, port sizes became a severe limit with the 3.0 and 3.2 engines, and it took 9.8:1 pistons for the RS 3.0 to get 230 hp. And, quite a bit wilder cams than the old 'S' AND 10.3:1 compression to get 255 hp at 7000 rpm out of the SC RS!
------------------
Warren Hall
1973 911S Targa
1992 Dodge Dakota 5.2 4X4 parts hauler

[This message has been edited by Early_S_Man (edited 08-28-2001).]

Folks,

I got the equation wrong. The above equation is for distance not speed (ie: d = 1/2 * a * t^2).

The correct equation is:

v = a * t

Using the above and autobonruns numbers we get:

a2/a1 = 7.1/6.7 = 1.06

Which means the AVERAGE acceleration increased by 6%. This means the AVERAGE torque increase is 6%. If torque was flat over rpm then peak Hp would increase by 10.8Hp (180 * 0.06).

The above equation is for constant acceleration, as Warren pointed out. It is correct to say the AVERAGE torque increased by 6% over the run.


------------------
Tony
'78 911SC

[This message has been edited by tbitz (edited 08-28-2001).]

Didn't Bruce Anderson perform this upgrade on many, many SCs, and report that hp gains were in the 15-20 heighborhood? He was not guessing. He's got a dyno.

I'll be making this same upgrade soon, along with 20/21 cams. Perhaps I should shop for an early ('78-'79) set of heads because of their larger ports. Are they larger intake ports, or larger exhaust ports?

Anyhow, I'm hoping for an extra 200 horsepower, but I'll settle for 30-40. And I may get it. You guys are just getting me excited.

------------------
'83 SC

You can't expect reliable calculations of hp using 0-60 runs for data! It doesn't take a physicist to realize that!

If you want reliable results, the data must be in [b]ONE particular gear, no shifting involved, and over the same test terrain, i.e., the EXACT same stretch of road, same tempeature, humidity, wind, etc. I trust that you don't have the 'before' data that meets such criteria, so your calculations are pointless! Entertain yourself if you wish, but you can't calculate horsepower with useless data! Real simple fact.

------------------
Warren Hall
1973 911S Targa
1992 Dodge Dakota 5.2 4X4 parts hauler

After reading a little more on sound propagation, I think one of the reasons that the sound travels further with the Dansk is that the exhaust tips now point straight back rather than down. Also, the lower frequencies should propagate further for a given dB.

If this was my daily driver, I don't think I would install the exhaust just for the additional power. The looks and sound are great but it is louder. I don't think I would like to hear this everyday. Since I drive the car less than 2000 miles a year, it's a perfect change from my 740 BMW. The droning sound is not that bad at higher speeds (55 mph+)even if I let it get down below 3000 rpm. In traffic at low speeds, I tend to try to keep the car over 3000 rpm. I don't want to entice an officer to inspect.

As far as the performance, I think that 0.4 seconds or so improvement is reasonable. I never expected the 0-60 times to increase that significantly. Again, I anticipate that the 1/4 mile test should really improve. The gTechPro does perform a HP calculation, but again, I need to be on a track to run it. Although I didn't run a pre-installation HP test, the HP figures on the current system should be reliable if they remain consistent over three or four runs. When I ran the 0-60 tests, I did try to pick conditions as similar as possible (temperature, time of day, etc.) and ran them at the same location. My figures are obviously not exact but met my needs by giving me more than just a gut feel that improvements occurred.

[This message has been edited by autobonrun (edited 08-28-2001).]

Yes I did richen it. The original settings left the system so lean, it would not even idle with the SSI's. It had to be richened quite a bit. Strangely enough, the car appears to be getting better gas mileage now. Before the installation, I was getting 16 mpg around town. I'll know for certain once I fill back up. As little as I drive, I'm waiting for fuel prices to fall from the current $1.80 per gallon price for regular before I refill. On some things I'm still cheap.

By the way, is it normal to have some backfire on decel or does it mean my air/fuel mix is still off? It is not popping, more of a burble.

You may not be far off given the fact that I started without a Cat. For those starting with a true stock SC, a gain of 15-17 HP may be correct. I'll know when I make the HP measurement runs. It's good when theoritical comes close to actual, recognizing that both include some assumptions that keep the results from being exact.

I continue to doubt whether a lean mixture alone can cause exhaust backfiring. It can cause intake backfiring, for sure.

I said here a few times that my mechanic friends insist that deceleration backfiring in the exhaust means a leak in the exhaust system. At one of the joints, of course. I am not ready to doubt this, though at least some one has felt that a lean mixture did cause this kind of backfire.

Did the exhaust parts go to a machine shop for 'decking' to ensure they are flat and true? It has been my experience that they seldom are. Most need this machining, even brand new headers.

At least, if you do decide they need to be removed, you'll be glad they are the early style instead of the set you replaced.



------------------
'83 SC

Easy there Warren. What's you definition of "reliable data" +/- .003%? Granted more precise data than supplied by autobonrun exists. But this is a nice sanity check from someone who's not trying to sell us something. Besides if word gets out that variances can't be managed then the gig is up for Statistician and engineers everywhere. http://www.pelicanparts.com/ultimate/smile.gif

<HR><font color="#CCC4A8" face="Arial Black"><strong>'81 Platinum Metalic SC COUPE</strong></font></p>

You could measure 0-60 times down to the nanosecond, and it would still be usesless data! But, if you don't undertand ... it is pointless to explain why!

Just a clue -- no shifts , and no launches from a standstill ...


------------------
Warren Hall
1973 911S Targa
1992 Dodge Dakota 5.2 4X4 parts hauler

Two comments and yes, I am a statistician, of sorts. First, the most important part of a statistican study is the data collection. If you wait until you have collected the date before you hire a statistical consultant, you have made a mistake. So, Warren is correct in that if you want the most reliable results, you would consider starting with data that contains no shifting, for example. If possible.

Second, (and this is the other side of the story) while some researchers have the luxury of studying simple behaviors using reliable measurements (dyno results, load tests, etc) those of us in government and/or behavioral sciences MUST make due with data that at best resembles the real measures you are trying to predict. So, you do the best with what you've got.

Unless you're Superman, of course.

------------------
'83 SC

Warren,

I don't think the analysis is "pointless".

My calculations simple state that if it takes the car 6.7sec for 0-60mph after the mods and 7.1sec before, then the AVERAGE acceleration must have increased by 6%, and torque is directly proportional to acceleration, thus AVERAGE torque increased by 6%.

Autobonrun stated the times were average over multiple runs (thus reducing the errors caused by unequal shifting). He also stated that he tried to keep all the conditions the same, time, temperature, location, etc...I can only assume he tried to do a fair compairson.

Is the analysis exact? No, but you make do with the data you have.


------------------
Tony
'78 911SC

Yes, you do make do with the available data. But no, the 6% average acceleration increase assumption is not correct.

When data has known or estimable weaknesses, these can often be statistically mitigated. For example, if these tests involved one shift (from 1st to 2nd) and if that shift took 1 second each time, then you could adjust all figures down by one second. so the "before" would be 6.1 and the "after" would be 5.7. This would result in a 6.6% increase in acceleration, based on the old (6.1) time.

I'm not criticising the estimates already made here, just illustrating the kinds of adjustments that can be made to data in order to isolate the variable you are hoping to estimate. Other relevant questions in my opinion might involve whether the before and ofter 0-60 tests allow both engine configurations the same chance at showing off their power zones (that is, if the new setup makes way more power after 5500 rpm, and if 60 mph in 2nd gear is 5500, then the new setup did not get a chance to show its stuff). Also, does elapsed time really translate directly into horsepower? etc.

Okay, having said all that, I'd agree that the new exhaust seems to produce more power, and that 6% is in the ballpark in terms of the 'delta ET.' I'd consider this to be a conservative estimate.

------------------
'83 SC

Superman, you called it exactly; the backfiring was caused by an exhaust leak . When I rechecked the bolts, the ones between the SSI and Dansk had backed off from 1/2 to 1 turn. This brings up an important lesson learned. First, I agree you should take the new SSI and exhaust to a machine shop and have the flanges decked. I know for certain one of my flanges on the new SSI was not exactly flat, but I thought the gasket would make it up. Second, retorque the bolts on the entire system after a few days. I think the instructions said to do that, but I didn't listen. The system now makes a slight burble during decel which sounds real good and I never touched the fuel mixture.

Thanks for the advice.

Tony, pointless might be a bit far, however Warren has a valid point. During the zero to sixty run there are places where the acceleration is not increased at all, during the shift, and during the start (acceleration is determined by traction at that point). So sure you're giving average acceleration, but it's moot to convert that to torque because maximum torque at rpm isn't being always applied. I bet you would think it would be silly to calculate torque increases if the car was allowed to sit for 6.7 seconds before each run and that time was counted. Your average acceleration increase would be 3%. Is that valid? How about a 3 % increase in torque? Even if you subtracted the time it took to shift, you'd have to take out the time it takes to get in the same range of RPM after the start, that is used in second gear, otherwise you're getting the average torque from say idle to 6.5k and from 4 to 6k in the powerband. I think showing a .4 better 0-60 time is wonderfully helpful to those how wish to upgrade. But that measurement isn't satisfactory to be used to calculate torque increases. BTW it was quite silly to use d=(1/2)(a)(t)^2 for velocity. I don't mean to be rude. I'm just backing up Warren. I probably should have a bit more schooling before going off like that. I just started as a freshman at Berkeley majoring in physics.

Keith
79 930

I'm starting to be sorry that I brought up the HP/ 0-60 question. Plenty of dyno tests have been done on the SSI conversion on SC's showing something like 15 HP and similar torque gain in the middle (peak). I question the accuracy of anyone's self-tested 0-60 time anyways, but assuming that the times are 100% accurate how is this for scientific, Warren: 1983 911sc 0-60 7.0 sec. w/ 172HP. 1984 911 Carrera 0-60 6.1 sec. w/ 200HP (all factory figures). Same car, same trans gearing, same weight (approx.) That's more or less my data- math formula left me on the trailer, so to speak.

Satisfactory for whom? The guy has the kahunas to put up some calcs, using the minimal data given for some first order correlation. He simplified the model/math to a linear curve fit.
Warren states: but 0-60 times are hardly uniformly accelerated objects Ya, so . . .From what I've seen of 0-60 V-T graphs, a comparison of linear curve fits is not *that* "silly." Yes there is non-linearity in the curve. But the before and after curves are of the same car, same driver, and are likely to share the same non-linearity. So for a first order comparison it is not *that* "silly." It’s not like anyone’s life is dependent on the accuracy, nor was anyone using 8 significant figures. . .when it comes to mathematical models of mechanical measurements you can always make it more complex; or you can do something useful.
Oh, BTW I'm just backing up Tony.

<HR><font color="#CCC4A8" face="Arial Black"><strong>'81 Platinum Metalic SC COUPE</strong></font></p>

Yark,

Who cares about "peak" torque and horsepower anyway? Are you driving your car around town at exactly 5500rpm all day? I think for a street machine indicating an average increase of torque from idle to 6000rpm is more usefull than saying the torque increased 30% at 5500rpm.


Sorry for being "silly" when I used d=(1/2)(a)(t)^2 for velocity. I am not "majoring in Physics at BERKLEY". It has only been 10yrs since I last took a dynamics class. I did see my mistake and posted a followup with arguably the correct equation indicating my first error. Some folks would have probably just edited the orginal post to hide their error. I am human, I make mistakes.






------------------
Tony
'78 911SC

Well Tbitz, I hope you have learned your lesson.

Where the heck is rstoll when you need him?

Alright, i'll do it.

ALL YOUR AVERAGE ACCELERATION ARE BELONG TO US.

Anyway, I think the original intent of the post is verified. The SSI's appear to do more than just make more noise and make the underside of your Porsche shinier. Good enough for me!

All that data analysis jargon was giving me a seizure.




------------------
Kevin
87 Carrera coupe

Am I the only one that thinks this is totally unacceptable for something this expensive?

Emcon5, no you're not. It surprised me too.

I agree with Emcon5 and Mikkel, the fact that this flange was not flat surprised me and was counter to the excellent workmanship on the rest of the unit. It was off around one bolt hole where the SSI flanges to the muffler. If I recall, about 3/32". I've heard of this issue on other SSI's. I seem to remember somewhere that the reason given was that the flange would go back flat once the bolt was tightened up. However this is not believable as the other flanges did not have this offset.

With respect to the HP issue, as mentioned earlier, my overriding concern was 0-60 and 1/4 mile times. However, you all have my interest up, so I'll run some HP tests within a week or so using the GtechPro. It calculates HP by multiplying instantaneous speed times acceleration times weight and displays the maximum attained. Since accuracy of weight is critical, I'll have my car weighed with a full tank before making the runs. I'll probably only go up as far as redline in 3rd (maybe 4th) since I don't want to push my luck with the law. Since speed and acceleration are used, this device measures HP at the wheels, which of course includes any drivetrain losses as well as the effect of Cd. HP will probably be much less than expected.

Two questions: Is the stock figure of 172HP quoted on the 79SC, engine HP or HP at the wheels?

Do dyno results take into account Cd? I assume the resistance is varied to simulate acceleration, but does this resistance also include air drag which does get factored into the GtechPro's figures. If not, the Gtech readings should be less than a Dyno.

172 is the SAE net figure, at the flywheel ... corresponding to the factory's 180 hp DIN rating.

No, neither engine nor chassis dyno figures take into account wind resistance, one reason that EPA mileage figures are artificially high.
------------------
Warren Hall
1973 911S Targa
1992 Dodge Dakota 5.2 4X4 parts hauler

[This message has been edited by Early_S_Man (edited 08-30-2001).]

Island 911: I don't understand what "kahunas" it takes to try to figure out an average torque figure from not just "minimal data," but wholly improper data for that application. For the record, I didn't argue about converting to linear acceleration equations, because he's going for average torque. I'm not just saying it's okay with me for a rough estimate. It should actually be correct for average torque if drag is negated, even though instinanious acceleration is varied, because that's not what's being calculated. I just have a problem with shifts and starts. I have no problem with using such equations for, let's say, a 3000-6000 rpm in thrid gear test.

Tony: I think the integrals of torque charts across the power band are better indicators than peak, so Average torque would be useful. However, I addressed that if measured from a 0-60 time, you'd get some weird average of torque delivered in first from whenever he dropped the clutch to his shift point and in second from the shift to the rpm that 60 mph would be. This is because the rpm sweeps wouldn't be the same in first and second gear.

Keith
79 930

Haruuumph!

Yes Keith, I understand the problems you (and others) have with the model. And I remeber the pressures and expectations from school on using the "correct" equations. The point of which is to give you a feel of what's going on. But in the real world expectations change, and are driven by time and money. Assuming you make it thru and want to start working in engineering; you will find the error of the calculation you provide will be inversely related to the amount of time and/or money given to you. Simply put; high precision means high cost. . . .a quick no cost calculation means don't complain if it's "not good enough."

<HR><font color="#CCC4A8" face="Arial Black"><strong>'81 Platinum Metalic SC COUPE</strong></font></p>

Island: I understand the whole cost and time versus precision. I forgot to mention that my point was not to spend the time on the calculations. Time=money, so it wasn't really a no cost calculation, was it? I retract this if Tony enjoys physics equations :-) That isn't that far out of the realm of possibility though, I find certain procedures satisfying upon completion. If they have to do with cars, all the better. However, if I post any, I do hope, and believe someone will grill me at every opportunity, to keep me on my toes.

The first chapter in my Physics book is actually about very rough estimates. I was asked to estimate how much rubber is put into the air in the US every year because worn tire tread mostly ends up as air pollution. I got a number around 1.4 x 10 ^ 8 Kg per year. Since I was only given that the average tread depth is about 1 cm and that the density of rubber is about 1200 kg/m^3 and had to assume the volume of tread used on average per tire per year and the amount of tires in the field, I think I was pretty close to the real value of 3x10^8.

While we're on the subject of equations:

http://www.pelicanparts.com/ultimate.../girlsevil.jpg

Keith
79 930

Thanks for the clarification. . .so you’re a baiting pontificator? I say, as Monty Burn would say; Excellent.
(However, others on this board would say; There goes the neighborhood.)

On your rubber to air pollution calculations: You can’t just work on the premise that all of the tire rubber goes to air pollution; that would be silly. so your calculations are pointless. If you don't understand ... it is pointless to explain why!
Just a clue -- no rain, and no launches from a standstill leaving black-marks...

heh-heh-heh "There goes the neighborhood" I can hear the echos.
<HR><font color="#CCC4A8" face="Arial Black"><strong>'81 Platinum Metalic SC COUPE</strong></font></p>

Why shouldn't I be happy upon completion of a problem involving the calculation of the horsepower needed to overcome drag on a 911 at 140 mph for example? The tire problem existed in physics world, you know, the one where silly things happen, and frictionless icerinks pop up way too often. Our friend with the SC lives in the real one I believe. Why would I be a pontificator if I want people to let me know of my mistakes? That makes no sense, if I was a pontificator, I'd go on with my dogma and expect others to swallow it. Thanks for all the sardonic posts.

Keith
79 930

Early-S-Man---Warren, I would be interested to read your thoughts on the design theory of the Triad muffler. assuming that the first can is just an expansion chamber and the two mufflers are similar to regular glasspacks, it would seem that a similar setup would be fairly easy to fabricate. I have a cis 2.7 stock, with SS1,s and a buchered old 77 banana muffler.Shipping costs are a bit prohibitive to this part of the world!---Any one?
Thanks ----Bob