Great thread! Here is a twin turbo set up for your consideration and feedback :)


http://photos.smugmug.com/photos/765220169_Bgp82-XL.jpg

No doubt it sounds great and sounds much better than the B&B I had.

I remember when I got my new stainless headers from Marco and how I couldn't wait to replace the short B&B header with my new custom equal length MM header. I spent all night ripping out the B&B and carefully installing the new set. Within a week, I went back to the dyno to see the improvements.

Initially, I found the spool was slower, but from 5000 rpm on, it felt good. When I got to the dyno, it was confirmed. I lost power up to 5000 but from there on, wholly crap it took off, nearly straight up. My problem was that I hardly ever drive at those RPMs on the street. But it sounded mean.

About 6 months later, I was in FL at a builder's dyno where the customer had his 3.3L CIS built and he replaced his B&B header and put 1 3/4 inch equal length pipes. The motor was at the builder not to initially be built, but to find out why the car was not putting out more power (owner spent huge $$). I went back home to GA and I later found out from the builder that they found the problem - the header was too big at 1 3/4in primaries. When they went to 1 1/2 short pipes, the power came up big time and they found they needed a different turbo (turbonetics) to handle the new-found wide power band (2800 to 7000 rpm). They also told me that the back pressure (the pressure between the head and turbo) also increased (sorry, can't remember the pressure they measured, but I think was around 8 psi?????).

So, even though I don't know all the physics and thermal dynamics of building a header, I know what I've seen and learned from my own experiences - short, direct pipes that keep heat and velocity up to a properly chosen turbo works best. I wish I would have kept a record of all those dyno's I did because data is the only way to argue these points.

Great advice. As far as performance goes, I wouldn't buy anything for my car that wasn't tested and proven.

I knew the answer to that story before you got to the end. 1.75" primaries are WAY too big.
If the pipes aren't the right size for the application they won't work. Sounds like yours were too big as well. Velocity vanishes quickly if the ID is too big.

Is there any resource or rule of thumb for primary and secondary sizes per boosted HP?

Brian, what size secondary size did you choose?

Anyone know a stock 930's secondaries and J pipe ID or OD is?

Primary and secondary ID are critical to velocity. I use 1.5" primaries and 2" secondaries up to 450WHP and 1.625"/2.25" beyond that. You also have to take a look at the rest of the build and see how/where the power is made to size the headers appropriately.

The stock Euro system is thick walled and I have not cut one of these in two to check the ID. The early Euro/USA system has very tiny secondary pipes.

For my 3.3L CIS, yes. For EFI (which was planned within a year) the size is just right for a 3.4L. In fact, I welded on a T4 flange to accept a bigger turbo - GT35.

Still, the problem (for me) is that they are equal length for a mostly street application. But I think this thread is inspiring me to make my own short pipes while the motor is out.

On a NA motor with SC/Carrera cams above 200whp one starts to benefit with a change to 1 5/8's primary tubes.

Is there a way to find out where a turbo will benefit in a jump up from 1.5" primaries?

Any third party expert info on this or rules of thumb?



If not, can we calculate it some way from the NA info above?

My gut tells me that if at 1 bar if we are pushing up to twice the NA air through the intake it might be about about 2 times or 400whp +/- but this may be far from, far from true.

If it is true and a 1 5/8's is about 20% larger that might take us up to about 480whp.

And if 1 3/4 is about 40% larger than 1 1/2 that might be good up to about 560rwhp.

This dose not mean we can not make much more hp, only that there might be some HP to be gained on the top end if there is not something else limiting.


Again, this could be far from true.


Anyone have any data points on primary tube sizes?

Brian,

Sorry, I did not mean to jump over your reply. Thanks.

No problem Keith. This is one instance where applying N/A flow theory will hurt you. Size the headers too big and you lose velocity. It's a balancing act (litterally) between back pressure and velocity.
The twin turbo pictured at the top is a nice setup. They could have taken it one step further and terminated the collector at the turbo. You can also eleminate the waste gate plumbing by using internal wastegates like the K16's have.

My 934 style exhaust also has 1.5" primaries and 2" secondaries.. Just installed and not tested yet. But good to hear my numbers are about right.. :)

Those internal wastegates usually dont respond well to higher output tuning..

The new tial MVS 38s probably are the ticket on a fairly high output TT engine.
(new mvs 38right/old tial 38 on left))

2 of these probably weigh less than half of one porsche waste.. SmileWavy

http://forums.pelicanparts.com/uploa...1263710842.jpg

I agree that external WG have advantages but there are packaging considerations as well. Internal waste gates are adjustable which can help offset other weaknesses. The above system has the WG venting back into the muffler so may as well use internal unless just to big/powerful.
I also use dual TiAL 38mm WG on higher HP systems that must have precise WG control. Have made several of these and really like the packaging and control.

[QUOTE=Ringmaster;5129509]Great thread! Here is a twin turbo set up for your consideration and feedback :)]

This is beautifully made, and can take advantage of pulse conversion due to being a twin turbo. The problem is that the individual pipes dump into a pretty big collector, which lets the pulses expand before they get to the turbo.

Also, the individual pipes can and should be MUCH shorter and straighter.

Tuning individual pipes to a certain length is done for scavenging effect. For scavenging to work you need to have a fairly large amount of overlap in the valve events. The idea is that a low pressure suction pulse will travel back up the tube and arrive at the exhaust valve just as the piston has gone past TDC and is started on the intake stroke. The low pressure pulse will still let exhaust gas flow out of the engine, even though the piston is now going down. Obviously, such a situation can occur only at a very narrow RPM, and needs a cam with a lot of overlap to work really well. If you are like most, you have the classic turbo cam, which has exceptionally low overlap, so forgetabout scavenging.

The other reason that longer equal length pipes are used is for momentum tuning. In this concept the mass of the exhaust gas gets moving due to the high pressure in the cylinder on the exhaust stroke, and the low atmosphere pressure. Once the exhaust valve closes, the exhaust gas will still keep flowing down the pipe because of its momentum, as long as it doesn't run into any interfering pulses. It will empty out the tube and make the pressure nice and low for the next exhaust event. For us the problem is obvious, we have a turbine in the way, so instead of low atmospheric pressure, it encounters something quite a bit higher, which greatly reduces the momentum effect.

Make the pipes shorter. This will improve your time to make full boost. Since a setup like this can push a lot of air through the engine, the diameter of the pipes is OK, but no bigger. If your goal is in the 450 HP or less, they could be smaller. The diameter balance for you is: 1) too big and the exhaust pressure expands in the tube instead of the turbine at low speeds - no good. 2) too small and your monster motor will have too much backpressure.

[QUOTE=Speedy Squirrel;5130566]
Great post and thanks for the feedback. To quickly address a couple of your comments...

-The cams are large Schrick race units with lots of overlap
-The engine package makes over 700hp
-It was built as a semi-race engine designed to operate effectively between 4000-7500rpm

More later when I have some time to type :)

re: primary and secondary sizing.


Spit balling again.

I am not suggesting we just use N/A flow quantities for the header sizing directly. That would not make sense. Just that we interpolate from the N/A numbers if we do not have any solid numbers for turbo'd motors.

For example, if we can flow say twice the air through the intake using pressure. Maybe the same applies to the exhaust side as, at least as a starting point.

As a point of reference 450whp might be compared to 225rwhp on a N/A motor. This seems to fit well.

I had no issue making 217whp through the stock 1.5" headers on my 3.2 Carrera and a sport muffler would have put me at near 220whp.

This worked well on my car but testing by SW confirmed that steping up a size would good for about 10 hp.

Not saying the relationship will be as direct but there may be a strong correlation as long as there is not something else that is going to create a restriction like the turbo hot side or compressor moving out of its efficiency range.

Before anyone making near 450whp runs off and buys a set of larger headers they might note that if the motor is still a 3.3 with SC cams how much might this hurt where the achieve full boost.

If one increases the size of the primary tube about 20% and the secondary tube about 28% velocity is going to slow down around 22% or so. Also the volume of the headers will increase about the same 22% or so.

This could push boost response back in as much as 600rpm (using 3000rpm as the starting point).

Now, bump the displacement from 3.3 to 3.5 and boost response should come in sooner by about 200rpm.

Also, I bet that if a 91/2 3.3 turbo comes on full boost by 3000rpm, a 93 3.6 that uses the same turbo will probably come in at about 2750rpm.

Going to a 1 7/5 from a 1.5" on a 3.3 and boost response could move back as much as 1200rpm or so. Great maybe for a race car if it has the heads & turbo to support it.

Thus. it might be better for most of us to stay with a smaller size to keep drive-ability even if 10-20 hp might be had with a larger size.

Just a thought.

Oooh, that looks lovely! I'd love a system like that on mine, although i'd have the wastegate pipes dumping straight out just for the added noise on WOT. Would the muffler part bolt straight onto a stock 993, are the turbos in the same position? Is this a custom job or is it available for sale off the shelf?

Sounds like all of this back and forth about equal length vs. short path is moot then. It also saves me $2500 :D

Scavenging is not the primary function of equal length headers on a 930 engine; having a tuned or tunable system is the point. It's all about efficiency and velocity. There is neither in a short tube or log system, you are just dumping exhaust at the turbo. If a tuned system is not important to the consumer then a short tube or log header is a much less expensive choice. The Chinese e-bay offerings are dirt cheap.

Interesting concepts, one picks and chooses basic engine principles that increase volumetric efficiency once the turbo is bolted on. Why?

Typically, if one is hotrodding an N.A. engine, yes they employ these principles to effectively use the mass of air flowing in both the intake and exhaust system to deliver more than 100% VE. In effect it’s mild supercharging on the intake and extracting on the exhaust. These are the same concept but acting in different areas of the engine system. They are beneficial acting independently, or combined.
The fact they are at their best when combined (tuned) doesn't negate their potential postive effects across the entire power range.

Suction (such a stimulating term!). Just using it regarding your engine compartment can lead you down wrong paths. It’s best to say absolute pressures. When you consider absolute pressures the lure of the word suction goes way and you are only looking at slugs of air, N.A. or turbocharged.

What changes once you place a turbo in the flow path? If you say the turbine represents backpressure then every N.A. street machine hotrodder with mufflers and tailpipes are only kidding themselves?. They should save their money because you can only see benefits when exhausting into open atmosphere? But then open atmosphere is 14.7#, isn’t 14.7# real pressure................

Yesterday I was playing on a Turbo calculator website. I put in the factory data for the '77 930 Turbo motor (no guessing the intercooler eff.). I was somewhat surprised to find the assumed VE had to be dropped to only 75% to make all other knowns work.

Obviously this is low for a performance design like the 911. It shows how much was left on the table to meet all other requirements like emissions, fuel economy, noise, etc while still producing more power reliably. In many ways a diamond in the rough compared with say the MFI 2.7RS.

This exercise also shows expensive hardware only marginally improves HP with no changes to improve VE. The best turbo and best intercooler will add little compared with good old basic hot rodder VE improving techniques of intake tune and balance, port/valve sizing (flowing), cam timing/lift and exhaust tune and balance.

This does not say the most efficient compressor and CAC are not extremely important. They are, after the high VE gets that charge in it needs to be cool not simply a high BOOST pressure.

I hear keeping the exhaust events in tact is one of the goals of a turbo exhaust. If that is the case, then some length to the equal length primary tubes might make for some benefit.

This is part of the reason a split scroll system works well. By matching up cylinders that exit the motor the events are more able to hit the turbine wheel in tact.


As copbait73 brought up, equal length also helps reduce the differences in residual pressure that will be present when the exhaust valve closes at each cylinder.

Seems even if we have some 20psi of pressure resident in the exhaust headers at full boost this dose not compare to the pressure and speed the gasses come out of each cylinder.

The pressure in the headers is going to appear in waves and with waves there are high pressure parts and low pressure parts.

If the exhaust valve closes during a high pressure part of the wave, less exhaust is going to exit. Low pressure, more will exit.

If this is the case, it dose not matter if each cylinder is getting exactly the same amount of fuel, there will be different levels of residual exhaust remaining in each cylinder and this will effect cylinder fill.

If so this could result in some cylinders running leaner than others and result will be seen with a variation in temperatures per cylinder.

Maybe we do not want to think of it as savaging like on a N/A motor but there seems to be something to this beyond letting the exhaust events just exit in a more orderly manner.

On a street motor where the duty level is modest this is probably not a bit deal. Under hight duty like in competition, it could become a very big deal.

The TT set up is very nice except the WG's coming back into the exhaust. That is not recommended and if has to be done should re-enter as far away from the turbo as possible. In this case just before the muffler.


For those that follow that are trying to decide between two or one I came across this from Garrett about single v double:



TurboByGarrett.com - FAQ's

FYI, I just measured a 930 J pipe. Thus:

1.5" primaries per Brian.
2" secondaries.
2.25" Y pipe to turbo.
K27/T3 flange 4 sq inch nominal, est. 3.5 sq inch actual.
T4 flange apx 6sq inch or est. 5.5 sq in actual.

These are nominal or OD sizes.

Here is how a stock system seems to play out:

The secondaries seem to have a ID cross section of about 1.8 times the primaries. As we have an exhaust event once every 240 deg per bank that is 1.5 exhaust events per revolution. An SC cam I believe keeps the exhaust valve open for about 220 deg. Thus, there is some reserve for exhaust expansion and collection at the secondary tube.

Then the two secondaries enter the Y or J pipe on a 930. The Y pipe is about 60% of the size of the two secondary tubes. Thus, gasses will have to be accelerated significantly as they enter the Y/J pipe.

The 2.5" Y pipe than supplies the K27/ T3 flange is mostly the same size at the turbo inlet so the velocity entering the turbo hot side dose not change.


On most after market 930 systems they seem to keep the primary to secondary ratio. However, the area of the two secondary tubes where they enter the turbo is much larger than the entry to the K27/T3 flange by close to 1.6 times. Thus, the gasses need to start accelerating before entering the turbo.


A couple of thoughts.

I wonder if we should be working from the turbo back and look at the changes in velocities?

On a stock 930 the Y/J pipe is going to see the highest velocities of any section of the headers and thus be the area that might be most likely to create back pressure. Doing something to retain the heat along that run might be something to consider as that would reduce back pressure to some degree. (wrap, coatings, larger tube?)

On most after market 930 systems the transition just before the turbo might be something to look at as gasses have to accelerate about 60% before they enter a K27 flange do to the cross section difference.

Putting a WG at that point is only going to add unnecessary turbulence and make the WG circuit unreliable. Moving the WG cricut away from the transition up into the turbo should make for an improvement.

A T4 turbine connected to two 2" secondaries moves most of any changes in velocity into the turbine housing as two 2" tubes have the cross section of about 1.1 times the hot inlet size. Seems like a good fit.

Running larger that stock 930 tubes to any turbo using a K27-7200 hot side (HF/K29 quick spool...) is probably not going make for much improvement as the turbine will probably not accept the increased flow potential anyway and WG sizing will probably become even more important. But that is just a guess.

Not an expert, just trying to figure this stuff out.

My 930 has very little Turbo lag. has special Muffler system and what turbo?

Can be seen here:

Cars For Sale - 1980 Porsche 930 EURO Turbo Coupe

Can't tell what 3K turbo is installed but probably a K27 7200 with a 964 support bracket, because that bracket doesn't look modified.

The exhaust headers with heat, and the intercooler are B&B with the single out muffler.

It looks like the crankshaft seal behind the flywheel is (still) leaking from all the oil all over the bottom of the bell housing.

On ebay:
Porsche : 911 : eBay Motors (item 280451069404 end time Jan-20-10 13:34:45 PST)

http://forums.pelicanparts.com/uploa...1264005418.jpg

How fun it that!

Fair comments re my headers. FWIW, I have tried dump-to-atmosphere vs plumbed-back WGs on a few cars and unless the exhaust system is undersized there isn't usually a large power difference. For my purposes I prefer the car to be a bit quieter @ WOT, hence the WGs dumping back into the exhaust stream.

Here is a photo of some iteration of 993 GT2 EVO race headers. They are large diameter, long tube and appear to be equal length. If I didn't mind driving around with ear plugs in I would do something similar on my car http://forums.pelicanparts.com/suppo...leys/devil.gif


http://forums.pelicanparts.com/uploa...1265257364.jpg

Yes, those are factory GT2 Evolution headers and they are gorgeous!

I do not think the primaries could be much shorter and accommodate a nice collector and waste-gate circuit.

Interesting how they carefully transition the WG's for efficient exiting.

I wonder--if we are expecting a substantial amount of exhaust gasses exiting the WG circut at WOT, if the same principles for reducing turbulence and promoting good flow should be extended to it to.

Usually the WG systems seems more like an after thought on most set up with us just sticking a 90 deg junction off the secondaries may or may not have an effect.

Could this offer some additional performance especially with the smaller turbine quick spool turbos that tax the WG even more.

Just a thought.

Any shorter and they wouldn't be equal length primaries. Equal length was considered important so here you have them.

The importance of flowing W/G has been recognized since the early 70s, mainly from work on Turbo Indy cars in the late 60s. I have a Crane Cams Turbo book from '75 showing it. This was made a science in the early to mid 80s with F1 Turbo motors. Also, running "spec" turbos CART champ car systems have beautiful header design with extra special attention to W/G flowing. You find very large piping off the secondary pipe going to equally large W/Gs.

As an aside, probably the best application of a W/G is 30degrees tangential on the turbine housing ported just upsteam of the volute throat. This was seen on many of the later F1 cars. Impossible to package on a 930 but interesting as best practice and you see the principle applied to the secondary pipes on the system above.

I learned something pretty cool today about WG's.

Thx!

Ok, back with some info from the first spring ride, track the car fairly often and I feel the 4-speed gearbox is the biggest setback with these cars on the track and a broader powerband was on the wishlist.. and I got it :)

(Euro exhaust figures above)


And with the stainless headers

0.5 bar at slightly under 2400 rpm
1.0 bar at 2700rpm

I was hoping for some improvement but this is almost unreal.. :eek: :)

Great info!

To be honest, I did not know changing from euros to shorties would make that much difference in boost response. Impressive.

That is about 400 to 500rpm sooner.

Puts .75 bar at about 2550rpm with a K27-7200 turbo.

What cams, intercooler, timing?

Any chance for a before and after dyno run?

My next set of headers, given the chance, will be Brian's new equal length headers for there quality, better efficiency and ability to make more HP. They are consistent with my goals and have several solid advantages over shorties that are important to me.

However, this thread is about equal v short headers and Zeb 930's info has given us a data point that can be used for comparison in place of a back to back.

Looking at the before and after Dyno for the equal length headers (ELH) over euro headers, it looks like the ELH made boost about 150rpm to 200rpm sooner compared to the shorties making boost about 400 to 500rpm sooner.

Equal Length dyno compassion before and after: http://forums.pelicanparts.com/911-930-turbo-super-charging-forum/461519-latest-project-headers-14.html#post5195908

It also looks like peak (.75 bar) boost was reached at about 3100rpm with the ELH's compared to about 2600rpm. I do not think a comparison of boost points would be directly comparable or valid as they may be other factors that may create a variance (street v dyno; observation v dyno boost curve; and, possibly different cams). I just do not think there whould be 500rpm between ELH and shorties.

Thus, it looks like it might come down to the shorties making boost maybe 200 to 300 rpm sooner than the equal length but the equal length making more HP.

Me, I will take the quality and HP.

Standard cams, standard timing(29), powerhaus fullbay IC, no dyno figures unfortunately.

If I where to buy a set of new headers I would also go with brian's for sure, they are a work of art! (brian helped me with turbo tank, hose, connections and gaskets, thanks again)

Got these cheap from a friend about 2 years ago and thought i give them a try. And they worked way better than expected.. :)

(And not to forget it probably made the car 20-30 pounds lighter in the rear)

http://forums.pelicanparts.com/uploa...1270658788.jpg

http://forums.pelicanparts.com/uploa...1270658798.jpg

zeb930

Did you fabricate the fresh air box or buy it? If you bought it, where?

Thanks

Love the uncommon color of your car.

Keith,
The RPM graph was off on my dyno run by several hundred RPM. What I see from my car is a threshold reduction of ~500RPM over the Euro J-Pipe setup. This is with a K27-7200. Full boost is also ~500RPM sooner. I am getting a lot of client feedback now that supports what I am seeing. Beyond that you also see a very significant increase in off-boost torque that translates directly to driveability. You really need to drive it to fully appreciate the change. No dyno comparison will be necessary at that point.

100% homemade.. :) Found the ducts at the local hardware store, I knew the placing was good from the nato-Ruf BTR car.

Would have made a similar tube version as the Ruf if I had the room but my full-bay IC made me do a narrow version.

If you decide to do a fresh air intake remember to make a couple of small holes in the bottom of the airboxlid for water evac.. And a lid for the tube if you expect to park it in extremely heavy rain/ washing SmileWavy

http://forums.pelicanparts.com/uploa...1270690653.jpg

http://forums.pelicanparts.com/uploa...1270690683.jpg

Very similar to the one used on the Group B Turbo from the early 80s.

Check the "Group B info and picturs wanted" thread in this section.

Do you remember the original use for that duct? I can't recall seeing one before.

That would save me some upcoming work.

Thanks

zeb930,
Is that PVC ventilation ducting?