Dear
Mr. Coleman,
I
am writing to take issue with your assumptions in the SCC June 2000 Sentra SE-R
cam article. But before I go into further detail, let me say whom I am. I am not
an automotive engineer or any sort of engineer for that matter. I am an auto
enthusiast of two years and I have familiarity with the SR20DE engine. I am
also a social scientist. I have a vast understanding of the philosophy of
science and the “tricks” of statistical analysis. I have done my fair share of
studies to know that the research paper’s assumptions have a direct impact on
the results of the article itself. The assumptions that you start with,
determine to a large extent the results that you get.
I
believe that the assumptions that you start with for your testing are flawed
and this leads to flawed results. Permit me to explain.
On
page 258 of the article you state “We did all the tests on our 91 engine which
has the best cams and best head of the bunch. In an attempt to simulate the
performance of the later engines, we also tested the stock 98 cams in our 91
head. This may not be exactly representative of the 98 engine’s performance,
but it is the best we could do.” What does “may not be exactly
representative" mean? Does that mean that the “98 cams in a 91 head”
are closely representative or totally not representative? I take
it that you mean that they are closely representative, because on page
260 you instruct the reader to look at pages 265, 267, 269, 271, and 273 in
order to approximate the power of the tested cams on late model SR20DE engines.
“If your SR20-powered steed is newer than that [post-1993], the second two
charts are a better approximation of what you should expect.” So your
basically telling your readers to extrapolate the results of the aftermarket
cams tested in your "98 cams in a 91 engine” to all the post 1993
SR20DE engines.
What
you are saying is that the “98 cams in a 91 engine” is close in power output
to post 1993 SR20DE engines. But your baseline results for the “98 cams in a 91
head” show 128hp and 124 lb-ft @ 6400 and 4800 rpm respectively. And the 91
SE-R that you tested the cams in had a Hotshot header, advanced timing, HKS
cat-back exhaust, and an Injen intake. This car earned you 13-14 hp in a
previous Sentra SE-R article. A 128hp @ 6400 rpm from a late model SR20DE does
not sound right.
Independent
dynamometer testing (http://www.se-r.net/car_info/dyno/index.html)
on a 1998 Sentra SE with a K&N drop-in filter produced 123.6 hp @ 6200 rpm
and 121.6 ft-lb @ 4800 rpm. Assuming that the K&N filter produces 1-2hp, it
is safe to assume that a late model SR20DE stock engine produces around 120-121
hp at the wheels. Am I to understand that a Hotshot header, HKS exhaust, Ingen
intake, and 21 degree timing produce only 7-8 hp in a post 1993 SR20DE engine?
That is highly unlikely.
Dynamometer
runs of post 1993 SR20DE powered vehicles equipped with modifications similar
to your 91 SE-R taken from http://www.se-r.net/car_info/dyno/index.html
show the following:
It
is clear from the above that the “98 cams in a 91 engine” that you created does
not approximate the power output of late model SR20DE engines.
I
know that comparing dynamometer runs is not advised due to variations in
temperature, altitude, and dynamometer machine (most of these examples are from
a Dynojet machine similar to the one you used). But I do doubt that these
variations will produce differences of 8-12hp between dynamometer runs.
Moreover, you compare different dynamometer readings in the MR2 vs. Miata
article, so I will give myself the same leeway.
The
problem is in the assumptions that you started with. You assume that the 91-93
SR20DE engines have the best cams and the best head. While the former is
correct, the latter assumption is false. 91-93 SR20DE cams do produce a
dynamometer proven 5-6 hp on a late model SR20DE engine. But 98 cams in a 91-93
head reduce the power output of an SR20 engine. This is because Nissan
redesigned the 95-99 head and intake to work well with the reduced cam profile
of later model SR20 engines. Thus, while Nissan reduced the cam profile they
modified the head and intake to keep the power output at or around 140 hp @
6400 rpm at the flywheel. Evidence for this comes from your July 1997 Project
200SX SE-R article. A Bosch chassis dynamometer found the power of the 1995
200SX SE-R to be 137 hp @ 6700 rpm at the flywheel. This is not far off from
Nissan’s claim of 140 hp @ the flywheel for late model SR20DE engines.
I
believe that “98 cams in a 91 engine” creates a fake baseline for the
power output of a late model SR20 equipped vehicle. This fake baseline, in
turn, affects the cam test results. When the aftermarket cams were tested from
the fictitious baseline, they seem to create a higher power output than they
actually would in a “real” late model SR20DE engine. For example, the JWT S3
cams (the most widely used by SR20DE powered vehicles) produced 25.9 hp @ 7000
rpm from around the 115 hp @ 7000 rpm fictitious baseline. I am pretty certain
that had you tested a "real" late model SR20DE powered vehicle with
similar modifications to the 91 SE-R, then the power output with the JWT
S3 cams would have been around 128-130 hp @ 7000 rpm. The 26 hp @ 7000 rpm gain
would have been reduced to about 14-15 hp @ 7000 rpm. The 14-15 hp gain is more
in line with what the JWT S3 cams would produce on a late model SR20DE vehicle
at 7000 rpm.
I
really appreciate the testing that you do for the SR20DE community. Honda
dominates the import aftermarket scene. It is refreshing to see Nissan tests in
SCC. But I have to believe that SCC erred on this one. In your testing
procedure you treated the SR20DE head and intake as a constant and the cams as
the variables. In reality, however, both the head/intake and the cams are
variables. This means that you cannot put 98 cams in a 91 head and
approximate the power output of a late model SR20DE. You need two SR20DE
powered vehicles, a 91-93 model and a 94-99 model. If you did not have the
resources to include a 94-99 SR20DE powered vehicle, then it would have been
appropriate to drop the “98 cams in a 91 head” from the testing procedure. As
currently constructed, the article gives a false impression about the power
gained from these aftermarket cams on a late model SR20DE vehicle.
___________________________________________________________________________________________________________________________
Update:
Recently
I had my SE-R power output tested on a dynamometer. The car had the following power
modifications:
Take
a look at the dynamometer results here: http://www.geocities.com/n_dahi/dyno.htm
As
you can see the maximum power output was 147.1 hp @ around 6300-6400 rpm (third
run). Assuming that the maximum base power output of post-1993 SR20DE engines
is around 119 hp at the wheels, I gained around 28.1 hp at the wheels from my
stock base. That gain came with all the bolt-on power that
you can add to an SE-R (except ECU). Had SCC's testing been correct, I should
have gained 26 hp from the JWT S3 cams alone. That is
clearly not the case. If I accept SCC's testing, then I gained 26 hp from the
JWT S3 cams and 2.1 hp from all the other bolt-on modifications.
That is simply unreasonable. It is more reasonable to deduce that I gained
12-14 hp from the cams and 11-13 hp from the rest of the bolt-on modifications.
12-14 hp is what most enthusiasts would gain from the JWT S3 cams on a post-1993
SR20DE engine.
It
is clear to me that their "98 cams in a 91 SR20DE engine" is not a
close approximation to the power output of a "real" post-1993 SR20DE
engine.