The great header debate - simplified?

So I've been reading about header design and exhaust scavenging and so forth for a while now and I came to the realization, I might read and study for a year and still learn more stuff everyday - there's just so much that goes into designing headers!
My request is simple. Do I want shorty or long tube headers? The vehicle in question is an awd, dohc v6 with an automatic transmission. Mostly DD but the occasional track event/hill climb/time trial. Should I be mostly concerned with maximizing airflow and increasing my rev range or concentrating on low end torque due to my drivetrain losses (probably close to 24% when the awd and automatic tranny are taken into account)? Future mods will most likely just consist of exhaust and some sort of custom intake and dyno tuning due to almost no aftermarket support for my X-Type.
I'm not concerned with making the most horsepower possible or using the latest fancy anti-reversion whatever or mocking up my own (due to space constraints and just general lack of talent), just looking for the nitty gritty - which will provide the best all around performance (not just high performance but all around street improvements)?
My options for off the shelf headers that MIGHT work are the stock Taurus Duratec headers, MSDS Mazda 6 (short tube) headers and CP Engineering long tube headers for the Mazda 6. I'm tempted to go with the MSDS set because I had them on my SVT Contour and Marty is a pretty good guy to deal with but the design of their y-pipe is a little crazy. http://www.msdsinc.com/ho816.html http://nautilusperformance.com/performance-automotive-parts-store#ecwid:category=2181184&mode=product&product=9153734
I like the looks of the CP Engineering headers but I'm not sure how they will work with the AWD setup on the Jaguar. http://www.cp-e.com/products/mazda-6s/mazda-6s-long-tube-header-system
What are your thoughts?

You want the long tube header for sure, but probably anything is better than a stock cast iron manifold (if that's what it has).

Everything will be a compromise of some sort. You want to get the one with the least amount of compromise, and that should be the long tube.

I read a rule-of-thumb somewhere, that the primary tube should contain the same volume as the cylinder.

In reply to mguar:

I think there's more to take into consideration than JUST equal length headers, there's a simple case of less restriction may be better than stock. For example, here's what the stock headers look like:

Just about anything would be better than this - the question is for my application, which one do I go with?

Zomby, care to elaborate a little more on what you said? Why are long tubes less of a compromise than the short tubes? Short tubes have their place in the performance world just as much as long tubes, don't they?

kb58, that's an interesting rule of thumb I'd never heard that before. I wonder how much truth there is to that?

I vote long tube too. I think shorty exhausts are popular because they are cheaper and easy for manufacturers to package/ship. Every real race header I've seen was long. The exceptions being when space was limited.

chknhwk wrote: Zomby, care to elaborate a little more on what you said? Why are long tubes less of a compromise than the short tubes? Short tubes have their place in the performance world just as much as long tubes, don't they?

Short tubes exist because sometimes it's impractical to fit long tube headers, also they are cheaper and easier to make. For any given equal amount of R&D, long tubes will make more power. Of course, a well designed short tube will make more power than a crappy long tube.

Cam tuning can also play a major role - In SR20 engines, longer duration and increased overlap have a bigger impact than header design - in general...strictly speaking about headers on SR20s though, longer = better

If you've got a musical sort of mind - headers work on resonance. You want to match the resonant frequency of the tube with the frequency of the pulses. Then you get a standing wave with low pressure at the exhaust port and voila, your scavenging gets a big bump. Short pipes are high frequency, long pipes are low frequency. And thus long tubes tend to be happier at lower rpm. Generally, you do want the longest tubes you can manage to fit.

Packaging drives most designs, of course. Trying to fit a long enough pipe can be a challenge. The 30" tube I used for my Targa Miata header had to be tied in knots to make it work but I also got a 14 hp increase over a Racing Beat unit. And you could hear it, too. At certain engine speeds on the dyno, the exhaust note would change. Same with the equal length header I built for the Seven, it's one of the most musical engines I've ever driven.

10' of exhaust tubing packed in the same space as a factory manifold!

well ive driven 6's with the cpe headers. and i myself run the msds altho with a custom y pipe and modified subframe. on the dyno there is only a few hp diff same with tq. the cpe's tend to have a really raspy sound while the msds shorties tend to keep that same lovely duratec sound we all love. since the jag is a vvt motor (i think) expect some rasp around 4500 to 5200 rpm. however u can truthfully expect 15 to 22 whp out of any header you select. also nautilous performance makes some that are a lil different from the msds look about the same but the routing is ever so slightly diff and supposedly are a lil better. looking at your routing as it is now the ypipe from the mazda6 will not work. and the fusion headers run down at the back of the motor with a diff style oilpan. my advace would be the bank1 header from either and a bank2 header from the mazda6 ford fusion ect. look into buying for a ford escape.... however i do know they dont fit with the awd drivetrain.... i did come up with a solution for that however it is a big of work but can be done.

do you have the ability to modify one of the headers purchased and fab up your own custom y-pipe? ill try to detail how it can be done. however on bank 1 side of motor which accesorys are there... is the alternator or anything else hanging over the manifold when mounted? not a big deal if it is thats what spacers are for.

You also want to be aware of piping diameter... excess diameter slows the charge down.

Further more, grouping the pipes correctly might offer you a few options (not knowing the V6 in question) with longer tube headers you might find a combination of 3/2/1, or 3 groups of 2 into 3 groupings of one, and finally into a single pipe leading away that benefits the entire rpm range

also of note the shorties actually produce better lowend tq than the longtubes. with what is available.

Keith, those look incredible but that's getting a little more in depth and involved than I'm willing to go right now.
KatyB, would you happen to have pics of your setup? I'm assuming you're running a Mazda6? When you talk about this 'solution' for the awd drivetrain what exactly do you mean? Off hand would you know if the Nautilus Performance headers sound the same as the MSDS? I'm tempted to stick with the short tubes because I don't want to lose that awesome Duratec sound...

Keith wrote: If you've got a musical sort of mind - headers work on resonance.

I recall a MM&FF article years back where they tested various types of shorty headers on the 5-liter Mustang. They tested equal length shorties, which looked similar to the ones you posted where it's a tangled mess of tubes, and compared them against unequal length units, which had wildly varying tube lengths but were simple, straight designs.

There was no real difference in performance from either! They theorized that any gains to be had with equal lengths were negated by all of the numerous and tight bends required.

They ALL made more power than the factory setup, though, which is telling. If you're unfamiliar with the 5.0s, they all had shorty headers from the factory, but the tubes were fairly small and they had many kinks and dents in the tubes for ease of assembly on the line - having to use open-end wrenches to tighten the headers just will not do in a production-line environment.

So, the end result is, if the stock arrangement is bogus, ANYTHING you can do to uncork things will help, even if it's not anywhere near equal length. Flow first, theoretical tuning second.

Now, where things get complicated is on more modern cars (ahem LS1) where the factory manifolds/headers are actually very well designed, and have a merge collector style exit and entrance to the close-coupled cat. Have heard way too many stories of people losing power with long-tubes unless a very free flowing exhaust is used to discount things. The merge is key, as it seems to decouple the exhaust system from the header. The exhaust can be restrictive and quiet but the header doesn't "see" it. Remove the merge and go with old-school style straight collectors, and now the header is "coupled" to the exhaust and it needs to be free-flowingly loud.

Another thing is that low-overlap cam timing is a lot more insensitive to exhaust tuning than something more rowdy. Look at the manifolding used on the new 5.8 Shelby engine as an extreme example. 650hp from a decently high revving V8 with some boost on top, and the exhaust manifolds are the same goofy looking boxes with five holes in them that Mod motors have used since the beginning. Mild cam timing doesn't care.

My tangled mess of tubes picked up 14 hp over a Racing Beat unit, so it wasn't as bad as you might think :D

Not to mention is looks berkeleying awesome. That is worth at least another 14 hp.

Keith wrote: My tangled mess of tubes picked up 14 hp over a Racing Beat unit, so it wasn't as bad as you might think :D

It's also a whole lot less cramped than an equal-length shorty for a Fox. Your bend radii look to be at least twice as large. Let's see if I can find an example...

Lots of tight bends and immediate turns right out of the port. Blech.

And also note that your header has a decent collector, not just wedging four tubes into a flange.

6" radius for most of the bends, if memory serves. I had one or two little ones where I had to go to 4". And yes, a decent Edelbrock collector. Of course, after I built it I saw a professionally built E-prod header with much nicer packaging. Never thought of sweeping them around THAT way...

Just goes to show you can't generalize that twisty is bad, though.

Here's one of the shorties I built for the LS1 MG. In this case, I didn't have room to package a lot of tube or even get equal length so I just went for low restriction.

Sometimes, you do what you have to do instead of doing the ideal.

In reply to chknhwk:

i do not have pics but my setup is very similar to how the 06 to 08 m6 headers are set up. yes the nautilous will produce the same sound. the issues ive seen with the awd drivetrain had to do with the power takeoff actually being in the way of the header coming down. i havent had the time or money to prove my solution but what i was planning on was 2 headers for bank2 spacing the alternator out chopping the very bottom of the header off for the one on the bank1 side after the collector and fabbing up a custom ypipe which would allow me to clear the power takeoff. i havent done actual measurements just have looked at (ok so i layed under my tribute for 3 hours after trying to put headers on it going wtf!!!! then that hit me eventually ill try it.

Vizard did a nice bunch of studies of headers over the years, which many header people did/do not like. Ended up disproving most of the myths.

The object of the game is to get the exhaust out of the engine, so the intake charge can get in next. That's it. The rest is fluff. Lots of entertaining theories on the best ways to do this, almost all with a grain or two of truth to them. But the basic fundamental is flow. Don't lose track of that.

A smooth pipe flows better than lumpy one. Gobs of weld in the path are not good, nor is a smoosh crimp or other unsmooth events.

The worse area of nonsmooth for most headers is the collector. Where the pipes come together. Just smash them into a larger pipe and you're likely to not flow much better than an oem lump of a manifold. The collector design is a great area to see how well the overall header will flow. Things like pacesetter are terrible here, while Stahl is wonderfull.

Bends are never a good thing for flow, but an engine bay tends to demand the pipes be bent around things to fit. That's ok. Less bends is better, smooth bends are better, larger radius is better. But, the effect is overall minor.

Equal length has become a holy grail, and in almost every dyno study I've seen, it gains virtually nothing. It's important if you want you trumpets to all be in tune in the band, but far less so for an engine that's just trying to get gasses out of the cylinder. Same the rotational patterns of flow in the collector. Oh the theory is real, and the effect can actually be measured. But it's a trace effect. There are other things far more important, like overall gas flow.

Still worried about long verses short pipes? The resonant frequencies of a short pipe are higher than those of a long pipe, so a long pipe matches lower engine rpms, while a short pipe matches higher engine rpms. The significance of this is completely overwhelmed by flow importance. How the pipes are welded, how they are run into the collector, and how the collector tapers the flow.

Well, that's my take on it.

foxtrapper wrote: smart things

I've never dynoed different headers, never built my own, but from my background at school and my understanding of the situation, this is dead on.

In reply to Keith:

Those are beautiful

Want to make me a set for a Olds 403 and squeeze them down to a size that will fit in a freakin G Boddy? Oh and I need them by next week..I'm kidding. Would be nice to get some functional ones for that combo though.

Why not build the correct headers.. send me the bore stroke #cylinders and camshaft timing events Duration, intake opens intake closes, exhaust opens exhaust closes and I'll run the formulas to get you correct size, primary length and collector length.. Once you know that you can either select something close or make something correct.

Chknhwk, being a former Contour SVT owner, did you ever spend time on any of the Contour forums? Do you remember many years ago when Sean Hyland Motorsports sold longtube, equal length headers for V6 Contours but nobody ever bought them because they were 3 times as expensive as the MSDS short tube headers? For a while they were the holy grail of Contour mods, and it was assumed you'd gain a huge amount of power from them. A couple of people finally threw down the cash and everyone was excited to see what they would finally do. As I recall, they didn't gain any power over the MSDS headers (which did give a decent gain over the stock manifolds, even those with 'gutted pre-cats'), and everyone was glad they didn't spend the extra grand for the longtubes. Since the X-type V6 is extremely similar to the Contour V6, I doubt you'd get much better performance from longtubes. Find something that fits and go with it.

Long tube/short tube, Ugh, I hate such a generality. But the Higher the rpm power is made at the shorter the length of the primary can be..(also the larger relative diameterwill be) 18,000 rpm in Formula 1 is realtively short but large diameter) while 6000 rpm can be done with (depending....... ) much shorter.

Then it becomes a matter of where you want power to come at.. For example let's say you want max power to be at cruise (in order to get max fuel mileage) That rpm might be say 2800 RPM then you adjust the length and size to reflect that lower RPM. Or if your goal is to make peak power for racing then you adjust to the sort of operating RPM you race at.. (say 3500-6500 rpm) Then the diameter of the tubes will be larger and shorter than a header designed for cruise RPM.

GREAT CARE MUST BE USED IN NOT GETTING THE EXHUAST PIPE TOO LARGE IN DIAMETER! The difference between say 1&1/2 and an 1&5/8ths makes a real differeance in power. Too large diameter will cause a significant loss in power gain..