We were recently contacted by Mike Raque, the owner of a 1996 Pontiac
Sunfire GT, who had purchased an Iperformance header for his Quad Four engine
to replace a RK Sport header. Mike explained that he had the car dyno tested
at The Dyno Shop in Louisville, Kentucky, on their DynoJet dynamometer, with
the RK Sport header on October 26, 2001 and then again at the same facility
on February 16, 2002, with the Iperformance header. Mike was nice enough to
make copies of the dyno tests available to us.
The dyno results were provided in the form of three different charts. The
first was a horsepower chart from the October 26, 2001 dyno session reading
from 3,000 RPM through 5,600 RPM horizontally and 0 through 150 HP
vertically. The second chart was a torque chart from the October 26, 2001
dyno session reading from 3,000 RPM through 5,600 RPM horizontally and 0
through 175 ft-lb vertically. The third chart was the combined horsepower and
torque charts from the February 16, 2002 dyno session reading from 2,800 RPM
through 5,600 RPM horizontally and 0 through 175 HP and ft-lb vertically.
Mike commented that he was extremely happy with the build quality, ease of
installation, and the "seat of the pants" improvement in performance that the
Iperformance Header provided when compared to the RK Sport header, but that he was
not as impressed with what he saw when he looked at the dyno charts,
specifically the peak horsepower measurement printed on each of the charts.
The October 26, 2001 test of the RK Sport header was indicated as producing a
peak horsepower of 135.6 HP, while the February 16, 2002 test of the
Iperformance header was indicated as producing a peak horsepower of 136.4 HP.
A one horsepower difference.
Mike commented that he knew the car felt like it had gained more than one
horsepower and was more than a little concerned that the performance
improvement he felt was not confirmed by the peak horsepower as printed on
the dyno charts.
Reading the Information
We took a look at the charts and immediately noticed several things. First,
that each of the three charts was in a different scale (as described above in
the layout of the charts horizontally and vertically, starting and ending at
different measurements). The appearance of the information on the charts is
deceptively similar, and considering the scale of the horsepower and torque
information on the chart would indicate a 6 HP or ft-lb difference in
measurement over a 1/8 inch change in position on the chart, it quickly
became apparent that it would be impossible to make any valid comparison of
the information on three different charts set to three different scales.
So, we had the information of the three charts scaled to match and put the
horsepower and torque curves for both the October 26, 2001 and February 16,
2002 tests on a single chart so that it could be easily and accurately
Once this was accomplished, a valid comparison of the results became obvious,
and what had been believed to be the "same curves" were found to be quite
different. In fact, the results from the two dyno sessions showed a
difference of two to four horsepower and four or five foot pounds of torque
across most, if not all of the chart.
Spikes and Speed Limiters
One mystery did still remain. Why does the October 26, 2001 test results show
a peak horsepower of 135.6 HP while the February 16, 2002 test results show a
peak horsepower of 136.4 HP?
Understanding this takes a very close look at the far end of the horsepower
curve and an understanding of how dyno testing is done as well as how the
Cavalier and Sunfire engine control system and how it limits testing on a
As described in our other dyno test writeups, dyno testing is done by first
parking the vehicle drive wheels on the rollers of the dynamometer, strapping
it down to the ground, and placing an inductive pickup on the number one
ignition wire (or trigger wire for the number one spark plug for vehicles
such as this with direct fire ignition systems). The vehicle is then
accelerated through first, second, and third gears into fourth gear, which is
used for the testing because fourth gear is typically the closest to a 1:1
drive ratio. The operator holds the engine speed somewhere just above 2,000
RPM, engages the dynamometer, and then accelerates in fourth gear up to and
just short of the maximum engine speed or redline.
For most vehicles, this is rather straight forward, but the Chevrolet
Cavalier and Pontiac Sunfire have a safety mechanism called a speed limiter,
that is built into the computer control system which cuts off the ignition
system when the vehicle reaches a speed of 105-108 MPH. This speed is easily
reached by these vehicles in fourth gear, and the vehicle will jolt rather
noticeably when the speed limiter kicks in.
When this happens while a vehicle is being tested on a dynamometer, the
action of the engine cutting off while the tires are spinning at 108 MPH will
result in the dyno rollers suddenly decelerating for a split second, until
the inertia of the rollers overcomes the engine, and the rollers will
accelerate again to a speed slightly less than before the speed limiter
This sudden deceleration followed by the feedback of the rollers overcoming
that deceleration results in a sharp and artificial peak at the very end of
the curves of both the horsepower and torque measurements.
This artificial peak can be seen at the very end of the curve just before the
line drops vertically to zero when the operator concludes the test.
If the speed limiter engaged at a point after the engine power curve was
already trailing off, this little artificial peak would not cause much
distraction, but, because the speed limiter on these vehicles engages while
the power curve is still climbing, just prior to actual peak horsepower, the
little artificial spike is picked up as the maximum horsepower measurement,
and while the October 26, 2001 test results showed the power curve peaking
just above 134 HP at a little below 5,400 RPM, trailing down to about 130 HP
after the speed limiter starts to engage, and then jumping 5 horsepower due
to the feedback of the rollers rapidly decelerating and accelerating, and
displaying the resulting artificial spike that is unfortunately and
inaccurately picked up as the maximum horsepower measurement, which the
DynoJet software prints on the corner of the chart as 135.6 HP.
This is why it is important for the dyno operator to determine what engine
speed the speed limiter will engage at and set the dynamometer to conclude
the test just before that speed limiter engages, to avoid this inaccuracy.
Dyno Chart Showing Performance Of Both The RK Sport Header and the
Click on The Charts to See a Larger Image...
So What Is The Performance Difference?
We would have liked to have seen a dyno run of the performance of the stock
manifold and down pipe, but Mike did not have the car tested before he put
the RK Sport header on, so analysis is confined to a comparison of the
performance of one header to the performance of the other header.
The RK Sport Header
After correcting for the false peak of running the car off the speed limiter,
the RK Sport returned a peak horsepower of 134.4 HP at 5,377 RPM. The
peak torque was 145.0 ft-lb.
Over the tested range from 3,300 RPM to 5,400 RPM, the RK Sport header
averaged 113.8 HP and 138.4 ft/lb of torque.
The Iperformance Header
The Iperformance header returned a peak horsepower of 136.4 HP and a peak
torque of 147.6 ft-lb.
Over the tested range from 3,300 RPM to 5,400 RPM, the Iperformance
header averaged 117.0 HP and 142.0 ft-lb of torque.
The Difference Between The Two
The Iperformance header measured a peak horsepower gain of 2.0 HP and a
peak torque gain of 2.6 ft-lb over the RK Sport header.
The Iperformance header measured an increase of average horsepower of 3.2
HP and an increase of average torque of 3.6 ft-lb over the RK Sport
Looking at the annotated power numbers at each 100 RPM increment, the Iperformance
header picked up a peak gain of 5.1 HP at 4,600 RPM and 5.5
ft-lb of torque at 4,400 RPM over the RK Sport header.
Further study of the annotated chart reveals that the Iperformance header
shows the most improvement over the RK Sport header through a broad section
of the middle of the power curve, roughly from 3,900 RPM through 5,200
RPM, where it picks up 3.5 or more HP and 3.5 ft-lb of torque at almost every
single 100 RPM sampling. This can be attributed to the Tri-Y design of
the Iperformance header, which produces a performance increase over a wider
range than the 4-1 design of the RK Sport header. 4-1 headers tend to produce
performance increases over a narrower band.
One conclusion seems very clear, that it is very important to get all of the
information and data being compared in the same scale and onto a single chart
so that it can be compared accurately.
Mike's "seat of the pants" assessment of the performance change was more
accurate than he believed, and that increase in performance with changing
from the RK Sport header to the Iperformance header is very easy to see in
the dyno testing results, when those results are presented with both sets of
horsepower and torque curves on one chart.
The Iperformance header shows peak gains of 5.1 HP at 4,600 RPM and 5.5
ft-lb of torque at 4,400 RPM over the RK Sport header. It is pretty
obvious to see that there is a rather serious difference between the
performance of the two products. How much of a difference is this in terms of
what the driver will feel? Consider this, the difference in power
between the Iperformance header and the RK Sport header is comparable to the
difference in power between a stock exhaust system and a performance cat back
exhaust system. This is a rather noticeable difference, and
represents a serious power advantage for those who are serious about
Mike informed us that he is more than happy to answer questions about his car
and the testing and can be contacted by email at:
We would like to thank Mike for providing his dyno results to us.
The Dynojet Shop that Mike Raque used for his testing can be contacted in
The Dyno Shop
7801 Carol Ave.
Louisville, KY 40219
TEL: (502) 964-2165