NA1 Midrange Torque - OEM Manifolds vs Headers

[RedCarsGoFaster]

Hi all,

It's always bothered me a little that a Honda VTEC engine should be entirely capable of making 100bhp/litre, but that the 3-litre version in the NA1 doesn't hit that magic number.

Replacing the cast-iron OEM exhaust manifold with tubular headers seems to be the best first step to set about righting this wrong, but the price ranges are huge. RFY and Fujitsubo make works of art that I'd be proud to hang on the wall but might make my bank manager unhappy; OBX make an almost suspiciously inexpensive offering that from reading feedback I'm not too keen on . . .

Pride seem to make a very reasonably priced header to a decent design, and actually include dyno charts, which I applaud them for!

However, the before and after dyno chart shows up an interesting point I hadn't expected. Not so much for the 'after' curve, but for the 'before' - there appears to be a great big lump of torque present between 3500-4500rpm with the stock manifolds that then goes away once the headers are installed.



Satisfying my OCD by installing headers to push the peak power up is all very well, but on the street I use the midrange torque from 3500-4500rpm far more than I'd ever use peak power - on the basis of these curves I might actually lose out on that deal!

So the question - is the Pride dyno chart actually representative? Is there some special secret sauce in the OEM manifolds that causes the apparent bump in torque in the midrange?

Inputs welcome!

 

[Honcho]

In a word- backpressure. All high-flow exhaust headers and mufflers greatly reduce exhaust gas backpressure, which permits the gas to flow with much less resistance through the system. This is a race car-derived technique and pays dividends at high rpm and consequently high exhaust flow conditions. There is a tradeoff, which is reduced low end torque and poor throttle response. When I went from a high-flow exhaust back to the factory muffler on my 91, I was amazed at how much better the throttle response and pickup was at low rpm. By tuning the length and diameter of the primary header pipes, you can attenuate some of this effect, but not all of it. Generally, longer pipes will improve low-mid torque and shorter pipes will improve high-end power. Below is a good exemplar of some of the popular header makers and where they decided to focus in the power band:

 

[RedCarsGoFaster]

Thanks Honcho - fascinating insight! Guess I'll have to have a bit more of a think . . .

 

[goldNSX]

Headers definitely have a benefit in the mid-range over the iron cast manifold. The 3.2 headers are not perfect but also better.
Dynos showed that clearly with my (former) Taitec.
Most of the torque was gained in the mid-range, low-end was like before, the gains in high-end were there too but not as good as in the mid-range. I've had a quite conservative (not loud) exhaust back then.

As long as you get rid of the iron cast manifold I don't think there are big differences between the different headers.
The combination of headers/exhaust/cats will decide on further improvement (completely tuned system).

 

[Old Guy]

I would be a bit suspicious of that 'before' torque curve which shows a 'great big lump of torque' centered around 4000 RPM. Unless I can't read, that curve is suggesting that the C30A torque peak is occurring around 4500 RPM. The factory 'advertised' torque peak for the C30A is at 6500 RPM. Advertised numbers are subject to marketing enthusiasm so use with caution; but, the fact that the torque peaks out at 4500 RPM on that dyno run suggests that something is 'up' with the engine or the dyno has some instrumentation issues. You can find a number of dyno curves on various threads and they all show some lumpiness; but, not like that curve.

 

[RedCarsGoFaster]

I would be a bit suspicious of that 'before' torque curve which shows a 'great big lump of torque' centered around 4000 RPM. Unless I can't read, that curve is suggesting that the C30A torque peak is occurring around 4500 RPM. The factory 'advertised' torque peak for the C30A is at 6500 RPM. Advertised numbers are subject to marketing enthusiasm so use with caution; but, the fact that the torque peaks out at 4500 RPM on that dyno run suggests that something is 'up' with the engine or the dyno has some instrumentation issues. You can find a number of dyno curves on various threads and they all show some lumpiness; but, not like that curve.

. . . that's a really good point, the torque peak is in completely the wrong place. Don't know how I missed that one!

If you draw a straight line between 3000 and 5000rpm then it both looks plausible and suggests headers are the kind of improvement I'd hope for. Perhaps time to reference other stock dyno sheets for curve shapes . . .

 

[mwagner10702]

I have a full Pride setup in my car. Headers, high flow cats and their lightweight exhaust. Based on my butt dyno I definitely felt noticeable improvement in mid-range performance. Fitment was spot-on and the sound is glorious. It's really the Goldilocks of exhaust systems IMO. Not too loud, not too quiet, the way an NSX should sound - and (most importantly) no drone.

Are there possibly better performing headers out there? Probably, but where the net marginal performance gain over something like a Pride header might be able to be measured on a dyno, in real-world driving (and this is where it counts) would likely yield imperceptible change.

I would have no hesitation recommending a full Pride system.

 

[DRIFTER]

I’m running SOS headers, Random Technology Cats, to an ARK DTS exhaust. It’s been probably 15 years since I put my exhaust on but there is definitely more mid range torque after swapping over the headers and exhaust. It was a very noticeable difference threw out the power band.

When SOS was selling them they did a bunch of tests with different headers. I don’t remember any of them losing mid range torque. They even tested the OEM 3.2 headers on the 3.0 and compared them to a bunch of different popular brands at the time. I searched for the old thread but couldn’t find it.

 

[stuntman]

Something is wrong with your baseline. The torque curve shape and peak aren't right. Also, headers tend to make more torque across the board with no significant losses over the stock manifold.

 

[RedCarsGoFaster]

Thanks guys - really appreciate the input!