Header Test - Examination of Hytech and comparison to standards.
Thread Starter
Honda-Tech Member
Joined: Feb 2000
Posts: 4,049
Likes: 2
From: Snowwhitepillowformybigfathead
Header Test - Examination of Hytech and comparison to standards.
I'm reminded of some of my earliest training in argumentation. My Grandpa on one occasion baited me with "What's the best car? Cadillac? Ferrari?" I responded with the question "What do you want it to do? Go around a track fast? Carry four passengers in comfort? Haul a ton of rocks?"
Discussion of the subject of headers (or any other component for that matter) often runs into similar obstacles. So it seems to me that in starting the next installment of this ongoing serial drama that I should stipulate that I don't believe in the idea of "the best" in practical terms, and feel that it is generally unhelpful. In previous posts I've restricted my use of superlatives to words like "good", "acceptable", and "satisfactory". This is not representative of low standards on my part, but rather a recognition of practical limitation. Still, a lot of people want "the best", and there's always somebody who'll sell it to them. What can I say but "More power to them".
Limitation is the governing principal in the conduct of my development and now testing. Before going any further, I'd like to thank Louie for trusting me with his Hytech header.
In the course of this testing I learned a lot about the peculiarities of dyno testing and as a result I have some redressing to do. Earlier this summer I installed the Mugen 4-1, Mugen ECU, and my race exhaust and went to Loynings to dial in the fuel pressure. After the CART support race I swapped the borrowed Mugen out for my own Comptech 4-2-1 and returned to Loynings to determine whether or not the Mugen was worth keeping. The Comptech produced much lower numbers and I reported this on H-T. This turns out to have been BAD INFORMATION. My latest testing resulted in these two headers performing VERY close to each other. This immediately got our attention. A review of the ambient conditions in the previous testing showed them to be very close. Our theory is that over the few weeks following that last race the pump gas in my tank lost significant spunk. This theory is supported by Loynings experience and that of several others I consulted. I'm informed that changes in tire pressure, alignment, and other small factors can add up to significant differences exaggerating, masking or hiding entirely actual performance differences. I've come over time to accept numbers that "sound right". This habit is more dangerous than I thought, and I will be more careful in the future. My apologies to Comptech.
I lost some interest in doing this test following Chris' post of his results with CTR pistons, JDM 4-1, and apparently stellar tuning by DynoSpot. While being wary of the risk in accepting his numbers, they tend to demonstrate that relatively cheap good 4-1's can be used in the production of good performance - that being my central point in the last episode. But by this point I was committed. I think the testing proved to be worth my time and money for educations sake.
The test was also used to examine the usefulness of the Mugen ECU which has never on this board received a reasonable analysis. Simply put the ECU evened out the A/F over the useful operating range (for the 4-1 and traditional 4-2-1), allowing effective simple fuel tuning using fuel pressure alone. The result of this is additional power at both extremes of the powerband. It's appeal to a prospective user could depend on how much screwing around they want to go thru to get their tuning in the ballpark, and the hard value they put on their time. The design purpose of the Mugen ECU is use on nearly stock motored endurance racers with conventional headers. That it's usefulness decreases when you employ hotter cams and such ought not be reason to dismiss it entirely since that's not what it's made for.
About tuning in general: I find a lot of well informed/experienced professionals don't like the usual inexpensive tuning devices. This dislike arises in part from a rejection of the inelegance of these devices, and in part from a recognition of their functional shortcomings. I did not become an expert on this subject in the course of this exercise, but I recognize that successful tuning is not trivial. The failure of the majority of relatively mildly hopped up and tuned B18-series motors to produce numbers much in excess of what cars like mine produce stands as pretty good support for this contention. The complexities of tuning for maximum performance without damaging the motor in the process are doubtlessly beyond a significant percentage of those who feel themselves well qualified. This is particularly important here because my testing confirmed what John has said: the Hytech header screws up the A/F curve. The Hytech header (and similar long primary 4-2-1's) MUST be used in conjunction with as sophisticated a tuning setup as you can practically apply to get all the power out of it. This can be spun either positively or negatively depending on your perspective and/or motivation.
How much power is that and where is it? Well, that's where I got a surprise. The Hytech performed significantly better than standards at the top end (7500 up). It performed worse than the standards in the middle of the powerband (6500-7200). This is correlated with the A/F situation. Total area under the curve in the useful powerband was greater with the Hytech. Running the curve off the soft bottom as tested would provide the seat of the pants dyno with satisfying feedback - potentially misleading in many instances, though not here. With tuning the Hytech wouldn't feel as good, but would make more area under the curve. This result (top end advantage) has been corroborated by others with testing experience with these type headers. So, contrary to expectation, the Hytech produced more on top without loss on the bottom (presumably, when tuned), rather than more on the bottom without loss on top. Performance below VTEC was not considered and is of no interest for my purposes.
I am obliged to say that with more sophisticated tuning all of the headers can probably make more power. What will make how much, and how they would stack up is a question for another test - maybe YOU can do that test. It seems pretty clear to me from my results and from talking to others that these funny long primary 4-2-1's CAN/DO produce usefully good results on these motors.
I understand that perfect A/F isn't everything, and that tolerance of the kinds of spreads seen in my testing is fairly common. A highly satisfactory solution to this and correlated horsepower deficits within the powerband may or may not be achievable - whether thru tuning or further development of the long primary 4-2-1. I would personally prefer the flattest possible torque curve in the useful powerband, and would generally be willing to forgo small increases in the area under the curve in preference for more linear response to throttle. Every setup is likely to have it's own particular curve once tuned, and selection of a known package is the only real alternative to costly development and replacement scenarios in pursuit of a desirable shape.
The idea that one particular curve is the best, is of course unsustainable. Determining a situational best is difficult. Sacrifice top end for grunt off a key corner? Where in the rev range are you spending most of your laps WOT? As you run thru the gears the answer is typically "in the top half" since aero drag increases exponentially. It's not easy, and few of us at this level will be employing different engine packages for each track,or for different weather conditions anyway. But we all like to believe in something. Believing it doesn't make it usefully beneficial - unless you've got a lot of latent psychological performance.
The idea of transient response as it's been touted remains an issue. I drove the car around a bit with the Hytech and find little in the response to tip in or any other part throttle phase of operation to convince me that there's anything significant to this religion. I will say that the effect of a steeply rising horsepower curve on a drivers perception, as mentioned above, is the most likely basis for sustaining this belief. There is of course the idea that the motors recovery to the WOT curve during the shifting sequence is a place where this dividend is paid, but that's a microscopic benefit that's hard to prove unless your standard is a car that bogs perceptibly. And I firmly believe that a racing driver wants a soft linear response to the return to throttle in the cornering sequence. If we are properly asking everything of our contact patches, we must have that soft linear response.
This is the first time that results this clear on this type of header have been observed by Loynings. Typical testing on motors as hot as Formula Atlantic Toyota motors, and as mild as Formula Ford Kent motors has not demonstrated superiority and adoption of use has typically been the call of drivers. Very good drivers are as likely to choose either a 4-1 or a 4-2-1. Loynings own testing has typically shown optimized 4-1 design to be superior on those motors.
In discussing these matters with the most informed people available to me, the inescapable conclusion is that THE ANSWER to the question of what works best lies at the end of a path of testing that stretches damn near to infinity. Reasonable satisfaction is the most rational goal. Why these things do what they do would require wearing out a motor on the engine room dyno. If my relationship with the B18 lasts as long as I think it will I may get around to that testing in a year or two. That will also depend on my relationship with George Washington.
Here are the graphs and tables (take care as you read the graphs - the scales are not all the same, and confusion is possible. The report generation did this to us - sorry):
Note: disordeR posted enlargements here: https://honda-tech.com/zerothread?id=354338
ECU Comparison graphs:
Comptech:
Mugen:
Hytech:
Summary graph - Stock ECU:
Detail table - Comptech: http://www.imagestation.com/picture/...0/fcf63cc3.jpg
Detail table - Mugen: http://www.imagestation.com/picture/...6/fcf63248.jpg
Detail table - Hytech: http://www.imagestation.com/picture/...f/fcf6320f.jpg
Summary graph - Mugen ECU:
Detail table - Comptech: http://www.imagestation.com/picture/...3/fcf6319d.jpg
Detail table - Mugen: http://www.imagestation.com/picture/...0/fcf63303.jpg
Detail table - Hytech: http://www.imagestation.com/picture/...0/fcf632cc.jpg
Here are some of the test particulars:
98 Integra Type R, motor stock internally, Mugen ECU, B&M FPR, 2.5 race exhaust with Magnaflow in the resonator position and Dr Gas SpinTech at the back. Test was run with heavy tall street tires on CRV steelies. Dyno was Superflow. No cam timing adjustments were made and fuel pressure was left at 60 psi for all runs. Timing was left at whatever it is (16-18 degrees). Each combination was run three times in succession and then averaged. The rollers were marked each time we rolled back to swap headers, and the rear restraining strap lengths were left fixed.
We committed some errors in testing too. After running the third setup (the Hytech), we were so distracted by the results that we failed to return to our initial setup and run again to establish a margin of error. I also ran out of time and didn't get real instrumentation installed, so we didn't maintain tight control over water and oil temp. But with 50-ish ambient and consistently timed testing, I think that conditions were probably close enough for my purposes.
Scott, who thinks a set of 48mm Webers and an Electramotive Crank Fire setup wouldn't be much harder to make work than a Hondata or AEM EMS.........can you tell that I'm real excited by the prospect?
Oh, and here's a bonus shot - a photo of a truly innovative header:
Larger version of the dyno graphs: https://honda-tech.com/zerothread?id=354338
[Modified by RR98ITR, 6:37 PM 12/8/2002]
Discussion of the subject of headers (or any other component for that matter) often runs into similar obstacles. So it seems to me that in starting the next installment of this ongoing serial drama that I should stipulate that I don't believe in the idea of "the best" in practical terms, and feel that it is generally unhelpful. In previous posts I've restricted my use of superlatives to words like "good", "acceptable", and "satisfactory". This is not representative of low standards on my part, but rather a recognition of practical limitation. Still, a lot of people want "the best", and there's always somebody who'll sell it to them. What can I say but "More power to them".
Limitation is the governing principal in the conduct of my development and now testing. Before going any further, I'd like to thank Louie for trusting me with his Hytech header.
In the course of this testing I learned a lot about the peculiarities of dyno testing and as a result I have some redressing to do. Earlier this summer I installed the Mugen 4-1, Mugen ECU, and my race exhaust and went to Loynings to dial in the fuel pressure. After the CART support race I swapped the borrowed Mugen out for my own Comptech 4-2-1 and returned to Loynings to determine whether or not the Mugen was worth keeping. The Comptech produced much lower numbers and I reported this on H-T. This turns out to have been BAD INFORMATION. My latest testing resulted in these two headers performing VERY close to each other. This immediately got our attention. A review of the ambient conditions in the previous testing showed them to be very close. Our theory is that over the few weeks following that last race the pump gas in my tank lost significant spunk. This theory is supported by Loynings experience and that of several others I consulted. I'm informed that changes in tire pressure, alignment, and other small factors can add up to significant differences exaggerating, masking or hiding entirely actual performance differences. I've come over time to accept numbers that "sound right". This habit is more dangerous than I thought, and I will be more careful in the future. My apologies to Comptech.
I lost some interest in doing this test following Chris' post of his results with CTR pistons, JDM 4-1, and apparently stellar tuning by DynoSpot. While being wary of the risk in accepting his numbers, they tend to demonstrate that relatively cheap good 4-1's can be used in the production of good performance - that being my central point in the last episode. But by this point I was committed. I think the testing proved to be worth my time and money for educations sake.
The test was also used to examine the usefulness of the Mugen ECU which has never on this board received a reasonable analysis. Simply put the ECU evened out the A/F over the useful operating range (for the 4-1 and traditional 4-2-1), allowing effective simple fuel tuning using fuel pressure alone. The result of this is additional power at both extremes of the powerband. It's appeal to a prospective user could depend on how much screwing around they want to go thru to get their tuning in the ballpark, and the hard value they put on their time. The design purpose of the Mugen ECU is use on nearly stock motored endurance racers with conventional headers. That it's usefulness decreases when you employ hotter cams and such ought not be reason to dismiss it entirely since that's not what it's made for.
About tuning in general: I find a lot of well informed/experienced professionals don't like the usual inexpensive tuning devices. This dislike arises in part from a rejection of the inelegance of these devices, and in part from a recognition of their functional shortcomings. I did not become an expert on this subject in the course of this exercise, but I recognize that successful tuning is not trivial. The failure of the majority of relatively mildly hopped up and tuned B18-series motors to produce numbers much in excess of what cars like mine produce stands as pretty good support for this contention. The complexities of tuning for maximum performance without damaging the motor in the process are doubtlessly beyond a significant percentage of those who feel themselves well qualified. This is particularly important here because my testing confirmed what John has said: the Hytech header screws up the A/F curve. The Hytech header (and similar long primary 4-2-1's) MUST be used in conjunction with as sophisticated a tuning setup as you can practically apply to get all the power out of it. This can be spun either positively or negatively depending on your perspective and/or motivation.
How much power is that and where is it? Well, that's where I got a surprise. The Hytech performed significantly better than standards at the top end (7500 up). It performed worse than the standards in the middle of the powerband (6500-7200). This is correlated with the A/F situation. Total area under the curve in the useful powerband was greater with the Hytech. Running the curve off the soft bottom as tested would provide the seat of the pants dyno with satisfying feedback - potentially misleading in many instances, though not here. With tuning the Hytech wouldn't feel as good, but would make more area under the curve. This result (top end advantage) has been corroborated by others with testing experience with these type headers. So, contrary to expectation, the Hytech produced more on top without loss on the bottom (presumably, when tuned), rather than more on the bottom without loss on top. Performance below VTEC was not considered and is of no interest for my purposes.
I am obliged to say that with more sophisticated tuning all of the headers can probably make more power. What will make how much, and how they would stack up is a question for another test - maybe YOU can do that test. It seems pretty clear to me from my results and from talking to others that these funny long primary 4-2-1's CAN/DO produce usefully good results on these motors.
I understand that perfect A/F isn't everything, and that tolerance of the kinds of spreads seen in my testing is fairly common. A highly satisfactory solution to this and correlated horsepower deficits within the powerband may or may not be achievable - whether thru tuning or further development of the long primary 4-2-1. I would personally prefer the flattest possible torque curve in the useful powerband, and would generally be willing to forgo small increases in the area under the curve in preference for more linear response to throttle. Every setup is likely to have it's own particular curve once tuned, and selection of a known package is the only real alternative to costly development and replacement scenarios in pursuit of a desirable shape.
The idea that one particular curve is the best, is of course unsustainable. Determining a situational best is difficult. Sacrifice top end for grunt off a key corner? Where in the rev range are you spending most of your laps WOT? As you run thru the gears the answer is typically "in the top half" since aero drag increases exponentially. It's not easy, and few of us at this level will be employing different engine packages for each track,or for different weather conditions anyway. But we all like to believe in something. Believing it doesn't make it usefully beneficial - unless you've got a lot of latent psychological performance.
The idea of transient response as it's been touted remains an issue. I drove the car around a bit with the Hytech and find little in the response to tip in or any other part throttle phase of operation to convince me that there's anything significant to this religion. I will say that the effect of a steeply rising horsepower curve on a drivers perception, as mentioned above, is the most likely basis for sustaining this belief. There is of course the idea that the motors recovery to the WOT curve during the shifting sequence is a place where this dividend is paid, but that's a microscopic benefit that's hard to prove unless your standard is a car that bogs perceptibly. And I firmly believe that a racing driver wants a soft linear response to the return to throttle in the cornering sequence. If we are properly asking everything of our contact patches, we must have that soft linear response.
This is the first time that results this clear on this type of header have been observed by Loynings. Typical testing on motors as hot as Formula Atlantic Toyota motors, and as mild as Formula Ford Kent motors has not demonstrated superiority and adoption of use has typically been the call of drivers. Very good drivers are as likely to choose either a 4-1 or a 4-2-1. Loynings own testing has typically shown optimized 4-1 design to be superior on those motors.
In discussing these matters with the most informed people available to me, the inescapable conclusion is that THE ANSWER to the question of what works best lies at the end of a path of testing that stretches damn near to infinity. Reasonable satisfaction is the most rational goal. Why these things do what they do would require wearing out a motor on the engine room dyno. If my relationship with the B18 lasts as long as I think it will I may get around to that testing in a year or two. That will also depend on my relationship with George Washington.
Here are the graphs and tables (take care as you read the graphs - the scales are not all the same, and confusion is possible. The report generation did this to us - sorry):
Note: disordeR posted enlargements here: https://honda-tech.com/zerothread?id=354338
ECU Comparison graphs:
Comptech:
Mugen:
Hytech:
Summary graph - Stock ECU:
Detail table - Comptech: http://www.imagestation.com/picture/...0/fcf63cc3.jpg
Detail table - Mugen: http://www.imagestation.com/picture/...6/fcf63248.jpg
Detail table - Hytech: http://www.imagestation.com/picture/...f/fcf6320f.jpg
Summary graph - Mugen ECU:
Detail table - Comptech: http://www.imagestation.com/picture/...3/fcf6319d.jpg
Detail table - Mugen: http://www.imagestation.com/picture/...0/fcf63303.jpg
Detail table - Hytech: http://www.imagestation.com/picture/...0/fcf632cc.jpg
Here are some of the test particulars:
98 Integra Type R, motor stock internally, Mugen ECU, B&M FPR, 2.5 race exhaust with Magnaflow in the resonator position and Dr Gas SpinTech at the back. Test was run with heavy tall street tires on CRV steelies. Dyno was Superflow. No cam timing adjustments were made and fuel pressure was left at 60 psi for all runs. Timing was left at whatever it is (16-18 degrees). Each combination was run three times in succession and then averaged. The rollers were marked each time we rolled back to swap headers, and the rear restraining strap lengths were left fixed.
We committed some errors in testing too. After running the third setup (the Hytech), we were so distracted by the results that we failed to return to our initial setup and run again to establish a margin of error. I also ran out of time and didn't get real instrumentation installed, so we didn't maintain tight control over water and oil temp. But with 50-ish ambient and consistently timed testing, I think that conditions were probably close enough for my purposes.
Scott, who thinks a set of 48mm Webers and an Electramotive Crank Fire setup wouldn't be much harder to make work than a Hondata or AEM EMS.........can you tell that I'm real excited by the prospect?
Oh, and here's a bonus shot - a photo of a truly innovative header:
Larger version of the dyno graphs: https://honda-tech.com/zerothread?id=354338
[Modified by RR98ITR, 6:37 PM 12/8/2002]
Honda-Tech Member
Joined: Jan 2002
Posts: 772
Likes: 0
From: bay area, ca, us
Re: Header Test - Examination of Hytech and comparison to standards. (VTECxSiR)
so what were the hp and torque gains from each header, im sooo fawking tired right now i cant read it all so ur post would go very appreciated.
Honda-Tech Member
Joined: Nov 2000
Posts: 11,967
Likes: 0
From: Charlotte
Re: Header Test - Examination of Hytech and comparison to standards. (Bloodwynch)
Cliffnotes maybe?
Thanks for giving your time and effort for this test Scott.
Will
-who dreams of a expensive header and a PowerFC
Trending Topics
Honda-Tech Member
Joined: May 2002
Posts: 11,209
Likes: 2
From: Tampa, FL
Re: Header Test - Examination of Hytech and comparison to standards. (Hooch at 8400-RPM)
...Thanks for the time it took to do the right up and report all this info.
-Dave, who fears the hytech guys response.... (edit: not because Scott said anything bad about the hytech.. just waiting for the require "you needed more fuel tuning and cam gear tuning")
[Modified by Dave-ROR, 11:10 AM 12/7/2002]
Honda-Tech Member
Joined: Jun 2001
Posts: 4,514
Likes: 1
From: Vancity, b.c, CANADA
Re: Header Test - Examination of Hytech and comparison to standards. (disordeR)
Nice to see a tech post in the ITR forum again. Thanks for all your hard work, it is appretiated.
Honda-Tech Member
Joined: Feb 2000
Posts: 2,383
Likes: 0
From: Richmond, VA
Re: Header Test - Examination of Hytech and comparison to standards. (mafay2)
For bigger pics of the graphs, COPY AND PASTE
http://www.imagestation.com/picture/...f.jpg.orig.jpg
http://www.imagestation.com/picture/...4.jpg.orig.jpg
http://www.imagestation.com/picture/...f.jpg.orig.jpg
http://www.imagestation.com/picture/...4.jpg.orig.jpg
Honda-Tech Member
Joined: Sep 2000
Posts: 5,276
Likes: 0
From: Sears Point, CA
Re: Header Test - Examination of Hytech and comparison to standards. (xjohnx)
Thank you Scott...regardless of the results, I am glad that you did this test. You are correct in the additional tuning, but most any combination could result in additional power with additional tuning. Thank you for your detailed write up.
Austin
Austin
Honda-Tech Member
Joined: Jan 2000
Posts: 826
Likes: 0
From: Eastern Coast Rocks, NY, USA
Re: Header Test - Examination of Hytech and comparison to standards. (owen_the_soyboy)
Wow, that was well worth the time in reading. Thanks for taking the effort much appreciated.