headers

[bseballkid1990]

I just recently put a h22 in my 93 accord, I'm looking for some decent priced headers. I was thinking of tri-ys but which would be the best looking for mid range power

[djcaz_aom]

So, you are buying them for looks? Why do you need more than one?

When I buy a header, I buy it based off of my engine's needs to get to my target HP goal. I have never bought one simply because it looked good.

[Jarrett]

I think there's supposed to be a sentence break between "best" and "looking". I thought the same thing at first. It's sad how even though everyone was taught correct punctuation and grammar, no one seems to care to use it.

OP, any tri-y header you buy for the H22A now is going to be a replica of one that was produced previously. I have yet to find anyone that you can buy a new header from that is based on an original design, with the exception of Bisimoto. From what I've seen, SMSP, Vibrant and RMF don't make their header anymore. There are plenty of knock-offs out there, though. They're all going to make decent power, but will likely not have near the quality of collectors as the originals. I have not looked at one personally, though.

Unless you can find a used one, I don't know what else to buy these days that's not a rip-off of someone else.

[bseballkid1990]

Yeah there was suppose to be a comma. I'm on my phone lol. I was looking at some kidd racing this guy had on his prelude but looking at other brands. Which is best though between tri ys, 4-2-1, 4-1

[CyborgGT]

4-1 promotes a narrower band of top-end power, 4-2-1 is better for mid-range, but generally provides a broader powerband. Another thing to look at is the length of the pipes before they merge: a "long tube" header is going to be better for mid range torque, while a "short tube" helps in the higher RPM. Intake manifolds work similarly. Also, when the diameter of the header pipes are stepped up at a point between the head flange and where they merge, or they have what are called "anti-reversion chambers," that helps decrease back-pressure (when the exhaust pulses merge at the collector, some of that energy is sent backwards towards the exhaust valves). Just in case you didn't know, 4-2-1 and "tri-y" are the same thing. It simply refers to the fact that there are three "Y" shapes in the system due to the three merge collectors.

Despite being a knock-off, I've heard the build & materials quality on the High End header is pretty good. If you can find one, that is; Logic Motorsports that sold them seems to have disappeared. Though it requires either a traction bar or a notched front cross-member to fit. A good way to make decent power on a budget would be to find the generic 4-2-1 design by DC Sports, Greddy, Megan Racing, etc and have the final collector enlarged to 2.5". Even the stock tubular header should be pretty good with a collector enlargement; honestly, I don't see the shiny aftermarket manifolds flowing noticeably better - they do look a lot nicer, though, if you care at all about aesthetics.

- - -

Why do you need more than one?

There are four cylinders, hence four headers - or heads, or beginnings or starts - to the exhaust system. Four sources of exhaust gas. Just like how people call the intake manifold pipes "runners" and not "a runner" despite them being a single cast piece. Many old European cars actually had separate pipes you bolted up to the individual exhaust ports on the head/s.

[djcaz_aom]

There are four cylinders, hence four headers - or heads, or beginnings or starts - to the exhaust system. Four sources of exhaust gas. Just like how people call the intake manifold pipes "runners" and not "a runner" despite them being a single cast piece. Many old European cars actually had separate pipes you bolted up to the individual exhaust ports on the head/s.

If you look at the OP, it sounds like he wanted to buy more than one and for looks.

[DAKOTADEMON7]

If you look at the OP, it sounds like he wanted to buy more than one and for looks.

Maybe he wants it to look like a V8. Lol.

[CyborgGT]

I've actually been looking at my build plan, and trying to make it less expensive without cutting corners or setting my goals lower, so I can get it running quicker. Instead of going aftermarket with the header, I'm working on my Spoon primaries with OEM downpipe. Going to get a Vibrant merge collector, turbo flex pipe, and flange, and have a shop weld in the bungs I need for sensors.


2" inlet dia. (perfect for stock), 2.5" outlet.

*ED. - for an example of the cost of modifying an OE downpipe, I wrote up a price list from ImportRP.com. They've usually got the best prices, so I go through them a lot.

82.22 Vibrant 10355 stainless merge collector 2" in/2.5" out
39.72 Vibrant 60804 stainless turbo flex 2.5"
04.16 Vibrant 1197 stainless EGT sensor bung
07.94 Vibrant 1194A stainless 02 sensor bung x2 (3.97/ea)
21.85 Vibrant 1482S stainless 3-bolt 2.5" flange
04.16 Vibrant 1462 stainless 3-bolt 2.5" gasket
10.98 Vibrant 1450A stainless hardware pair x2

171.03, plus the cost of weld labor. I'm also going to need some straight pipe to fit the sensors, but I'm not going to order 5 feet of it; maybe I'll just get Vibrant's UJ bend I wanted anyway to provide scrap bends in case I need them for the rest of the system.

[illinois_erik]

I just recently put a h22 in my 93 accord, I'm looking for some decent priced headers. I was thinking of tri-ys but which would be the best looking for mid range power

I like the header I bought, looks good, claimed gains, very good build quality,decent price. everything i wanted
Only problem Ive come across with it, is the clocking of the O2 bung, it puts the sensor right onto the steering rack, I remedied this with a stainless steel elbow that was bored out (theyre usually used for downstream sensors on highflow systems causing CEL's) However I think if i had gotten one of those universal ITR 2.5" testpipes with the slip fit sections, I couldve used that O2 bung with no problems. The header uses the ITR 2.5" graphite donut collector.

http://www.highendautosports.com/item.php?id=144

[CyborgGT]

^ Very nice. I was looking for Logic and not High End and found nothing, so I thought they were gone. That price is too good for how those are supposed to perform.

I know that in a Prelude the front crossmember needs significant notching unless you've got a traction bar...

[johnl]

4-1 promotes a narrower band of top-end power,

My understanding (and I'm not disagreeing with what you've written, just elaborating);

More or less, usually. 4-1 tends to create a narrower power band, but it's more the length of each individual pipe that determines where in the power band the most affect will occur. Short pipes in a 4-1 will tend to create a narrower power band higher in the rpm range, longer pipes will tend to create a narrower power band lower in the rpm range. A 'peakier' engine either way, though 4-1 headers tend to shorter pipes to maximise high rpm power at the expense of lower rpm power.

We need to keep in mind that a 'tuned length' pipe is beneficial between X and Y rpm, but tends to be counter-productive outside that range, i.e. rather than creating a low pressure wave to arrive at the exhaust valve as the valve opens, a high pressure wave will arrive, with the opposite affect to that which is beneficial to cylinder filling.

Of course all pipes have a 'tuned length', even short branch 'log' type manifolds. It's just that with such manifolds the 'tuned length' is so short that the rpm typically never get high enough for the tuned length affect to come into play (the engine just won't rev that high). As a result, older engines with such manifolds have some degree of reversion occurring at all rpm, which doesn't help power, and the main reason why longer branch headers used to have such a dramatic affect on power output (when log style manifolds were common), but less so these days where stock exhaust manifolds are usually somewhat more sophisticated (the stock Honda manifolds are reasonably good).

4-2-1 is better for mid-range, but generally provides a broader powerband.

Yes. The short pipes work at higher rpm, but the additional longer pipes add another tuned length that extends the power band width. The second pipes do lessen the 'focus' of the first pipe length, so the beneficial effect of the tuned length of the first pipes is lessened, but the power band increased.

Another thing to look at is the length of the pipes before they merge: a "long tube" header is going to be better for mid range torque, while a "short tube" helps in the higher RPM. Intake manifolds work similarly.

Agreed, as above. It's all about the time it takes for positive and negative pressure waves (moving at supersonic speeds, depending on gas temperature) to move along the length of the pipes. The shorter the pipe the less time it takes for a negative pressure wave to move back along the pipe to arrive at the opening exhaust valve, which suits the shorter exhaust valve opening time intervals at higher rpm. With a longer pipe it takes more time for the negative pressure wave to reach the opening exhaust valve, which suits the longer time intervals at lower rpm.

Also, when the diameter of the header pipes are stepped up at a point between the head flange and where they merge, or they have what are called "anti-reversion chambers," that helps decrease back-pressure (when the exhaust pulses merge at the collector, some of that energy is sent backwards towards the exhaust valves).

My understanding is that the 'step' in diameter (smaller port ID, larger pipe ID) lessens the severity of a positive pressure wave as it arrives at the port / header junction (and thus at the opening exhaust valve beyond that). This is the 'anti-reversion' bit, it lessens the severity of positive pressure 'reversion' that occurs outside the tuned pipe rpm range, reducing the power loss that occurs outside the 'tuned' rpm range.

I don't think it strongly (at all?) increases the strength of the desirable negative pressure wave as it arrives at the opening exhaust valve (it might, I could be wrong), but is more to mitigate the negative affects of the unavoidable positive pressure wave, the net effect being to slightly widen the power band without lessening peak power (or at least minimally, if at all).

[CyborgGT]

The big thing that attracts me to the Bisimoto header is that it combines 4-1 with long tubes. Just from looking at it, I would assume it offers a great powerband. Sounds like Bisi's leaving the Honda world to go with high-dollar manufacturers so they can improve their image though, so sadly their headers look to be going the way of HyTech and SMSP.

[wagon-r]

Only problem Ive come across with it, is the clocking of the O2 bung, it puts the sensor right onto the steering rack,

My muffler shop just asked me where I wanted the O2 bung. They were happy to add it anywhere on the header or exhaust pipe.

[wagon-r]

The big thing that attracts me to the Bisimoto header is that it combines 4-1 with long tubes. Just from looking at it, I would assume it offers a great powerband. Sounds like Bisi's leaving the Honda world to go with high-dollar manufacturers so they can improve their image though, so sadly their headers look to be going the way of HyTech and SMSP.

Its the quality control - or lack thereof - that is likely sinking the Bisimoto header. Too many of them just leak far too much.

[johnl]

My muffler shop just asked me where I wanted the O2 bung. They were happy to add it anywhere on the header or exhaust pipe.

Ideally an O2 sensor should be as close to the cylinder as possible, which promotes quick heating of the sensor (an issue for passing emissions testing, but not a lot else, though there may be a slight benefit for economy as the ECU will go into closed loop just a little sooner). In practice it's better to place the sensor where (or just after) all the pipes collect into one pipe, so that the sensor is getting the average of the O2 content from all cylinders.

My sons old Prelude (MY89) has two stock sensors, each one located just past where two primary pipes join but prior to where the two secondary pipes converge into one main pipe. This gets the sensors closer to the cylinders, but adds the cost of an additional sensor (and the ECU would have to be able to recognise two individual signals from two separate sensors).