turbo lag, short v equal length headers.
 

Stole the following data point from another thread that pretty well displays to potential effect of smaller volume exhaust header compared to a more free flowing equal length system.

I suspect volume of the header system, including the wast gate pipe, has a direct correlation to turbo response.

The larger the volume the more area that has to be pressurized.

There will always be examples where design or other factors overcomes.

This example is a stock restrictive compared to a race system so the HP levels in the upper ranges might be discounted.

It may be a better designed shorty system could better approach the equal length depending on the design. However, the equal length has the advantage of better timing the exhaust events which reduce back pressure over a shorty system.

The same principle seems like it would apply on the intake side as the amount of volume that has to be pressurized.

This data means nothing without pictures of the systems tested. OEM is the stock J-pipe system. I canot imagine that OEM J-pipe outperforms any header nomatter what the design.
So we need to see pictures.

Sure it dose.

However I did not realize it left out that this is on a Saab motor, not a Porsche. I thought that info came over with the quote but missed it. Apologies.

It is just a comparison of a smaller volume inefficient stock cast iron Saab header against a built equal length system that seems to display that less volume spins up the turbo faster.

This is not the last word in header design, just a point of reference as there are always other factors that have effect.

What is most interesting is the stock unit on the same motor made more HP early in the power curve over the built equal length system.

This is just a data point.

It's a dilemma.
Less resistance to flow means more gas velocity at the turbine inducer and faster spool-up.
Equal length headers yield better off-boost performance but IMO it's hardly worth the effort since a turbo engine spends so little time off-boost.
My 993 heat exchanger based system is the shortest possible primary path possible. Boost threshold is around 2500 rpm so time spent off boost is negligible.

The drawbacks of long and equal length are that they introduce more surface area, and more volume. The energy to the turbo is a combination of temperature and pressure (don't use the "E" word anybody). Longer cools the exhaust much like a radiator, and also takes longer to pressurize when you go from low flow to boost unless ...

you have a divided turbine housing, and you have connected our 1,2,and 3 to one side, and 4,5, and 6 to the other. Then you can convert the pressure of the pulses to energy in the turbine, which is a beautiful thing. In practice it is not usually achievable in a boxer engine. The three equal length pipes have to come together at one side of the divided turbine housing with no single pipe collector in between; otherwise the pulse conversion efficiency is greatly reduced.

Faced with these problems, most engineers turn the dial the other way and make the manifold as short and small as possible, abandoning equal length. In this way the thermal losses and volume are minimized.

Having spent thousands of dollars of my own money and having looked over the shoulders of lots a great motor builders over the past 10+ years, I would follow the above without a second thought if I were to do this all over again...

Don,

Sorry but I am not sure I follow what you said.

If you can expand it would be appricated.

Thx

Keith, Don said that after pissing away untold sums of money and hanging out with tuners that know even more than Speedy, that the shortest primaries possible are the hot set-up.
Theory has it's place on the board but there's no substitute for BTDT.
Thanks Don.

I should also have mentioned that another solution is to divide the turbo - make twin turbo's. Then you can have the optimal setup, short equal length manifold with high pulse conversion.http://forums.pelicanparts.com/uploa...1263046619.jpg

I agree that it is meaningless data in relation to a porsche 930 OEM exhaust versus custom equal lengths. Apples and Oranges.

I like technical threads in general, it is just a shame that nobody has put all the aftermarket exhaust possibilities back to back on the same 930 and dyno and afternoon. Probably never gonna happen.

If it was a Honda Civic, probably the information would be readily available, lol.

That is the easy and cheap way out. It is a matter of returns and effort/time per dollar. The significance of any gain has diminished returns as the effort to get that gain is increased.

There is also the consideration of long tubes vs short tubes when it comes to equal length. Reducing the total header volume while retaining equal length primaries has its merrits in the proper application. This is more commonly seen more in the twin turbo systems than in the single turbo systems, largely due to cost. Optimal header design is application specific. For the stockish 930 masses unequal length short tube 304 headers are the norm. That makes them the best compromise for the dollar that folks are willing to spend, it does not necessarily make them the best design for the application.

In the above Saab example we need to see what the OEM and aftermarket systems look like. The aftermarket system could be a miss-match while the OEM system is designed quite well for that engine.

That has been on my to-do list for some time. It's a big effort and cash outlay. Also if I do it there is percieved conflict of interest.

I think most people would trust you on the results.

I general nothing is is going to beat a well designed equal length header system for getting the exhaust out in the most efficient way.

A well designed exhaust of any type should bump peak TQ and lessen the fall off after peak TQ.

Thus, creating potential for to increase the average HP potential "under the curve".

Header volume seems to be to trade off with all Turbo set ups as to boost onset response no mater if equal length or not. Same thing is true on the intake side. A larger intercooler has to be pressurized before the motor is going to see full boost.

I suspect the WG volume is part of that volume. Most get excited with the short pipes to the turbo but do not think about the WG circuit volume. I would love to see someone fill up a stock system with water and a sortie system and measure the difference. It might not be as much as we would like to think.

As to normaly asperated differences pre-boost, even at idle there is back pressure in the headers as the turbo is creating boost. I think my resident boost at idle was about .2 bar if measured at the intercooler. Thus, we are probably not going to see any scavenging effect nor reduction in back pressure unless we make improvements in back pressure at the turbo (Zork).

There, may be some magic in if the exhaust hits the turbine wheel with consent pressure or in waves. I have not clue about this.

It is pretty well accepted that on a 4 cylinder motor that using a split turbine improves response but that with the more cylinders this advantage seems to fall off to the point there is not any gain with a V8.

What interests me is using equal length and a less restrictive turbine and balancing that the best AFR and ignition set up. But I am more interested in DE's.

If I was commuting I would want higher compression, a small turbine, lowest restriction muffler that I could live with, and smallest volume headers possible (including WG), and good heat. Probably a 993HE based set up.

I suspect there is nothing really wrong with the stock 930 set up as seemingly proven by the Monster CIS build where he gets well over 600hp with them.

Still, as well noted above, this is all application specific.

In some cases making an improvement to an exhaust system that should make a significant improvement is limited or reduced because of restrictions somewhere else.

Just my two cents and not worth a penny more.

A couple of statements I must take on:

“Equal length headers yield better off-boost performance but IMO it's hardly worth the effort since a turbo engine spends so little time off-boost.”

Really, you must be something to see in traffic or is this statement based on driving strictly on the racetrack? If you reflect on normal driving you will notice you are on boost maybe 10% and full throttle much less. What does your engine like best off boost, equal length or something other?

“Must reduce volume in the exhaust”…….

The words that best describes what a turbine desires most is no expansion before reaching the turbine followed by maximum expansion through the turbine. In real terms, the difference in volume causing boost delay between equal length systems and Other is mice nuts when you look at the sweep volume of 3.3L at 2500RPM.

Dynos are great tools however their readings will never guarantee who reaches 100mph first.

I thought the same thing as soon as I read that. While I am reluctant to dispute with Squirrel and DonE (their posts are always spot-on it seems), it seems to me that in a mere two crank revoultions the pistons will push 3.3 liters of air into the exhaust system. It doesn't seem intuitive that a wee bit more volume due to routing choices could be material or not made up in one 4-cycle event. Pulse/wave etc is a different story perhaps, but that varies by RPM and cam architecture and is further complicated with a turbo sitting on the end of the system.

I wonder:

A turbo is always on boost. Even at idle.

There is vacuum on one side of the throttle plate and boost on the other at idle and cruse.

On a turbo the exhaust pressure before the turbo can run to about 3 times the boost level.

Thus, at one bar we might see up to 45psi in the headers!

If header pressure is at say 3psi at idle and will end up at about 45 psi at 1 bar boost, that is a lot of pressure that has to be built up in the exhaust before ones boost target is achieved.

Knowing this, it seem that it makes sense that header volume might have a notable impact on lag?

As such on a turbo car it would not seem that we should expect equal length or any headers to do anything at, for, or against low rpm (before 2000rpm) with the back pressure created by the turbo at idle.

Such back pressure might be more restrictive than any muffler a normally aspirated car might ever see.

On top of that "nearly zero overlap cams" do not seem to care very much about the header design except if it is overly restrictive. Especially early in the rpm range.


If we are dumping exhaust out the WG it seems we are making much more energy than we need to run the turbo at the desired boost.

If this is true are we concerned with a little heat loss due to how much surface area there might be from one header design to another?

On a turbo optimized for ultra HP operation with a large non restrictive turbine wheel, retaning as much exhaust energy as possible at the headers and turbine due to heat loss might be more important.

On a normally aspirated motor with a long tail pipe I can see how having exhaust gasses cool and increase back pressure a PSI or so might hurt peak HP.


Lastly, why can an equal length of proper design make more HP?

I suspect headers make more HP on a NA motor for two reasons. Less back pressure and savaging.

Can we expect any scavenging from turbo headers?

If we are operating at say 45psi at one bar boost I would think it might be minimal a best.

However, equal length should allow the exhaust events to exit in a more orderly manner. If so, this should help reduce back pressure just as it dose on a normally aspirated car.

If we can reduce the pressure in the headers from say 45psi to 43 psi, that is about a 4% reduction in back pressure.

That should allow for 4% less exhaust pressure remaining in the cylinder when the exhaust valve closes. That should intern allow some level of additional intake fill during the intake cycle...

I am not an exhaust expert, just trying to better understand.

Dose any of this make any sense?

Sorry, I didn't read past the comment the turbo is on boost at idle. This is not true. Depending on the size of turbo, some are barely turning. You have no flow, that's the function of the throttle.

Are you sure?

I used to have up to .2 bar if I measured before the throttle plate with my K27-7200.

Just asking.

Is this off to:

I know you understand turbos well.

I thought I would chime in just for fun. Brian, please correct me if I am wrong.

The header pipes being equal is not as important as the length if they are correctly "tuned". By tuned, I mean that they are at such a length that when the engine is at peak design HP RPM, the exhaust gas pulse or wave is just starting to exit the pipe as the next exhaust valve is opening to create another pulse. Because as the pulse exits the pipe it "pulls" some of the remaining exhaust from inside the pipe due to the rapid expansion of the gases as they leave the end of the pipe. Think top fuel dragster exhaust pipes. The header pipes can be combined into a single pipe as long as this single pipe is shorter than the individual cylinder pipes.

Now, if you can properly tune the overall header pipe length (this number should be known for a stock 911 engine) to within about 1/8" of optimum and then make all of the header pipe lengths equal, you will have accomplished something that will enhance the performance of the engine.

In order for this to work, the pipe diameter must fall within a certain range, generally agreed to be about 1 5/8" to 2". If the pipes are too small, the gasses cannot be pulled efficiently. If the pipes are too big, the effect is diminished. A collector cone or megaphone on the end can help too.

So, tuned header pipes, with equal lengths, about 1 5/8" in diameter is most desirous.

Now enter the turbo... The tuning length is not as critical because there is not a much larger area for the gases to rapidly expand at the end of the header pipes. The equal length is good because this keeps the time between pulses consistent. The pipe diameter is still pretty important because you don't want to create restrictions; rather you want to keep the ID as consistent as possible to keep the flow going to the turbo without changing the velocity (which happens when the pipe ID changes at a constant flow of exhaust).

I think for us turbo drivers having the shortest equal length headers possible will help direct the most exhaust heat and pressure energy to the turbo. After the turbo, the length is not too critical because the pulses have been "chopped" by the turbine blade and just need to get out of the way. This is facilitated by a large diameter exit pipe with minimum bends, say around 4". Of course most jurisdictions require a muffler, and maybe even a catalytic converter. Clearly these impede the efficient exhaust system, but what are you gonna do?

Thanks,
Mark

Seems like everyone thinks "on boost" means full boost.
When I push the accelerator just a little more than in granny mode, I'm getting 3-5 psi of boost. Acceleration is brisk but not warp speed. That's what I'm referring to.

Yes, positive regarding idle.

Regarding pressure differentials:

Actual Porsche 1978 published data; @ 6000 ERPM, intake .8bar (11.6#), exhaust manifold roughly 1.2bar (17.4#).
Intake is roughly .8 bar at 3000 RPM (W/G regulated). Pressure differential goes exhaust positive at 4300ERPM.

Porsche has always been a high performance installation, with that slower boost response, but lower exhaust backpressure before and downstream of the turbine.
When you have installs typical of DET cars they have much higher exhaust system backpressure (downstream of turbine) therefore higher negative differentials to compensate.

Sorry to jump in late guys, I would like to hear some opinions (nice and mean) on these headers I had made. Im not going to cry as I did not spend a whole hell of a lot on them.
I have since moved the passenger side wastegate mirror image to the driver side..
http://forums.pelicanparts.com/uploa...1263093868.jpg
http://forums.pelicanparts.com/uploa...1263093901.jpg

Are those for a CIS car???? 'got some big pipes there...

How do you like their performance?

Paul-

I apologize for making a point at your expense. Yes, I know no sane person does this. But you are still taking about acceleration which at any rate, 100% or something less, is still a small % of the total driving cycle. Most Porsches do not have pressure/vac instrumentation, that's unfortunate as these show how the car responds N.A. then boosted.

I've owned my equal length pipes for quite a while now and for street use, they're a waste of extra money I spent. I also find it odd that nearly all header builders consider "equal length" being just back to each collector, not to the turbine. If they aren't perfectly equal to the turbine, who cares - a bit of a challenge with most of our turbos being off-set.

Will have some exhaust spoolperformance info later this year. Have changed from stock euro to a short length 321 system. No other changes made except all oil-lines from thermostat and rearwards.

With euro exhaust and K27-7200 + HKS "doubleacting" boostcontrol

0.5 bar at around 2900rpm
1.0 bar at around 3100rpm

Hope to see some improvement.. Collectors are far from optimal but I got the system from a friend fairly cheap..
Will let you know the numbers for the stainless later on (lots of snow) SmileWavy

http://forums.pelicanparts.com/uploa...1263096793.jpg

Everyone is getting very close to the correct understanding. Let me make a few brief comments to help guide things along:

1.(gskrken) Your thought experiment is in many ways exactly right regarding the effect of addtional header volume. You just have the conditions wrong. When we go from low boost to full boost we do not just have to fill up additional volume, we have to pressurize it. Until we meet the pressure demand of the turbine, the volume continues to need "more filling". Thus, even a relatively small volume can require many liters of exhaust gas compressed into it before the turbine energy requirement is met. When you consider that a .2 second improvement in time to full boost is noticable in the seat of your pants, I think you can see how this can make a difference

2. (lucittm) Your basic grasp of gas dynamics is very good. The individual pipes must be arranged so that, in our case, the pulses are spaced 120 degrees apart. This naturally occurs when each bank is grouped together. An open end reflects back a negative pulse (low pressure). If this negative pulse arrives at the exhaust port just as the exhaust valve is opening, it increases flow out of the cylinder, just as you more or less stated. For all cylinders to be doing the same thing at the same time, the lengths of the tubes have to be the same, as you pointed out.

The condition that we have (and that you recognized) is that we have no open end to reflect a pulse back. It is neither open nor closed. Any pressure pulses are expanded through the turbine and used to compress air. This is the fundamental fact that, unique to turbo engines, we have to recognize. We do not reflect back ANY negative pulses. Without any reflected pulses the length of the tubes becomes non-critical, except from the standpoint of volume that needs to be filled, and surface area for heat loss.

There is only one way to use the pulse energy in a turbo engine, and I mentioned it in my original post. If the individual cylinders of one bank are connected to one side of a divided turbine housing, and the other bank likewise, the pressure of each pulse can be expanded in turn through the turbine, measurably improving its efficiency. This picture is about right (there should be a wastegate on each side, and the divider wall is melted away. It should go all the way to the turbine flange):

http://forums.pelicanparts.com/uploa...1263098931.jpg

One last word, about temperature. Temperature lost is energy lost. How critical it is depends on how much is thrown away.

They are for an efi car- they work pretty well, still never drove the car with a proper tune with them installed, but I can say that the car dropped 300 rpm of spool time, so I guess they are an improvment over the last set (B&B's modified for a T-4 flange and single V-44 wastegate).
I am doing a diffrent injection system next week which I will post a thread on and at which point I will see how it does on the dyno.

Well if I may chime in here, on my 3.8tt the headers on her are custom equal length setups - they wouldn't be on the car if they weren't vital to the amount of sheer torque that the motor produces as well as top end breathing...but that being said it makes for an EXCELLENT street motor.

Look at the 935s, the setup i'm moving to now to be exact to place the turbos at the rear of the car...they're designed for high rev and single purpose use, track suited so to speak.

So the header debate is great but keep in mind, its about the use of the car and motor and what is needed from it that will designate the header design.

Just my $.02

Kris,

For what it is worth.

For a sortie I love you headers.

You kept the WG volume at a minimum, great collectors, modest turns where possible...

Do not know about tube size. Looks like a T4 style turbo which has interest to me.

;)

Here is a peak at a Protomotove Equal length.

Short as possible primary's but still equal length, equal secondaries, short WG pipe. Probably very expensive.

This guy is one of the best at getting the most 'efficiency' possible out of a turbo motor.

http://forums.pelicanparts.com/uploa...1263142795.jpg

There is probably small differances as to pre boost TQ, boost response, and efficiency, back pressure... between well designed and sized shorties and equal length system as it relates to a 930.

What maybe 100 or so rpm difference as to lag and a couple of poneys here and there. Again, on a well done for the application system.

If off idle a motor can only make 40 hp there is not much we can do with just the headers.

Other factors like turbo selection, intercooler volume, AFR, timing, compression probably make make and much or more of an impact.

Which is why I built mine equal length to the turbo.

The unequal shorty headers should take their "design" one step further and eleminate the collectors. Just use a step-up log design and reduce the exhaust volume to absolute minimum.

thanks man- the turbo is a Pt-67 ball bearing which is basicaly the a gt37r- I only used twin 44 wastegates because I already had a single 44 and bought a second cheap. It will perform well when I get a better injection system and have a better tune. This is what I did with the wastegates instead of having them asymetrical... It is also divided up to the turbo for future plans of a 40r with a divided housing.
http://forums.pelicanparts.com/uploa...1263154369.jpg

Yes, an equal length system is best to have equal secondaries.

Or-- keep each side a split system including the turbine.


Kris, love your car but your tail pipe looks a little small! ;)


Suspect that the state of tune, turbo choice, and what happens after the turbo is probably more important than short v long, assuming a solid header design for both.


My votes for what it is worth:

Best equal lengths, Brian's system.

Best shorties w heat, M&K converted 993NA's.

A solid contender is the stock late 930 HE system for quality and originality.

OK - different animal (twin turbo). On the tt projects I've been part of, I would submit you couldn't tell the difference between short vs equal on a street car (day to day driver). Equal pipes really come into their own at high rpm but give up some down low. I witnessed this directly when I replaced my B&B pipes with a set of Marco Manzie equal pipes and found I traded performance from one end of the power band to the other. It wasn't a bad trade, but I do most of my driving under 5000 rpm.

Brian - have you made a genuine equal length header yet? If so, do you have a picture? I would really like to see how you handled the turbo placement. Might be in the mood for a new header ;)

I like zeb930's setup. It is basically the 934 arrangement with no divided turbine housing, heat exchangers, and a single wastegate. Hope he defrosts early over there so we can see what it does.

930GT40's is interesting because of the dual wastegate locations. I would jet-hot coat that sucker to keep the heat in and call it good. Do you think the slip joint will keep loose over the life of the system? I think those things get stuck after awhile, which can lead to cracking.

I know that the equal lengths are a waste on an undivided housing, so I wouldn't roll with RarlyL8's setup myself.

But what about RarlyL8's equal length primary and secondary headers with a divided flange Gt40? ;)

Yes Don my headers are true equal length from the port to the turbo. The turbo rests in the stock position. I would not have gone to all the trouble for nothing. Take a look at any high end header system vs the Chinese stuff and you will see a distinct trend toward equal length. You will also see superior metals being used.

http://forums.pelicanparts.com/911-930-turbo-super-charging-forum/461519-latest-project-headers.html?highlight=headers

It should also be noted that split volute systems are being built as well as twin turbo systems using my headers. A component design allows for options to suit the application.