In this test I take a sealed woofer and place it in a Honda Accord and then in a Honda S2000. The first is a midsize hardtop sedan, the second is a small soft top roadster.
Test Car 1: My 2001 Honda S2000, sub on the passenger seat, FR shifted up 6db, BLUE
Test Car 2: 2002 Honda Accord 4dr, sub in the trunk facing up, FR shifted up 6db, RED
Reference output: woofer tested on bare floor in the parking lot in BLACK. See how well it matches the manufacturer's white sheet here:
http://www.seas.no/images/stories/pr..._datasheet.pdf
Gear:
Seas CA26RFX
Sealed 1 ft3 box, stuffed
180w @ 8ohm max power
Design
Sine sweeps are ran though the woofer from my car. I keep the output fixed across all vehicles. I have a long wire which enables me to drop the sub in any car, or anywhere.
First the sub was tested outside of a car, on the concrete floor facing up. The mic position is about 10 inches from the woofer. I average about 15 sweeps. This frequency response graph is the BLACK line in ALL graphs that I will post for this test.
All the in car tests are from the driver position. The sub placement varies by car, I can test one location or more.
Finally, all output levels are matched at 60hz. I do this so that enthusiasts can get a good idea of the shape of the FR graph compared to reference and other cars. I will also state the output shifts I do that are needed to match at 60hz. All graphs will therefore overlap at 60hz by construction.
My comments (what I see):
Compared to the Accord, the s2000 has very little gain under 27hz. The smaller interior probably has a lot to do with that. Above 27hz there is a decent amount but the sub box was in the passenger seat in my car whereas the sub box was in the guy's trunk for the Accord test. Quite a bit of output is lost with distance yet they tie more or less around 30hz.
The trouble with this method is that the distance from the mic to the sub box is almost always different. When I test the sub outside I have the mic 10 inches from the cone. In my car the sub box was on the seat so half a foot away. In the Accord it was over 6 feet away. Relative output levels can be thrown off like that but I suspect the transfer function remains the same shape wise. That is why I chose to match the frequency response curves at 60hz by shifting the plot. This way it's easy to see the changes in cabin gain.
Big thanks to Circa40 for donating the box, Duckymcse and JT Audio for tools and hosting over on DIYMA.
This should be useful for simulating subwoofer boxes in WinISD. More importantly if you combine the transfer functions from these tests with the simulation you can get a very good idea of the final frequency response of your subwoofer.
Now what about your comments?
areas of improvement
what startles you
what did you expect vs. what you see
etc.
Test Car 1: My 2001 Honda S2000, sub on the passenger seat, FR shifted up 6db, BLUE
Test Car 2: 2002 Honda Accord 4dr, sub in the trunk facing up, FR shifted up 6db, RED
Reference output: woofer tested on bare floor in the parking lot in BLACK. See how well it matches the manufacturer's white sheet here:
http://www.seas.no/images/stories/pr..._datasheet.pdf
Gear:
Seas CA26RFX
Sealed 1 ft3 box, stuffed
180w @ 8ohm max power
Design
Sine sweeps are ran though the woofer from my car. I keep the output fixed across all vehicles. I have a long wire which enables me to drop the sub in any car, or anywhere.
First the sub was tested outside of a car, on the concrete floor facing up. The mic position is about 10 inches from the woofer. I average about 15 sweeps. This frequency response graph is the BLACK line in ALL graphs that I will post for this test.
All the in car tests are from the driver position. The sub placement varies by car, I can test one location or more.
Finally, all output levels are matched at 60hz. I do this so that enthusiasts can get a good idea of the shape of the FR graph compared to reference and other cars. I will also state the output shifts I do that are needed to match at 60hz. All graphs will therefore overlap at 60hz by construction.
My comments (what I see):
Compared to the Accord, the s2000 has very little gain under 27hz. The smaller interior probably has a lot to do with that. Above 27hz there is a decent amount but the sub box was in the passenger seat in my car whereas the sub box was in the guy's trunk for the Accord test. Quite a bit of output is lost with distance yet they tie more or less around 30hz.
The trouble with this method is that the distance from the mic to the sub box is almost always different. When I test the sub outside I have the mic 10 inches from the cone. In my car the sub box was on the seat so half a foot away. In the Accord it was over 6 feet away. Relative output levels can be thrown off like that but I suspect the transfer function remains the same shape wise. That is why I chose to match the frequency response curves at 60hz by shifting the plot. This way it's easy to see the changes in cabin gain.
Big thanks to Circa40 for donating the box, Duckymcse and JT Audio for tools and hosting over on DIYMA.
This should be useful for simulating subwoofer boxes in WinISD. More importantly if you combine the transfer functions from these tests with the simulation you can get a very good idea of the final frequency response of your subwoofer.
Now what about your comments?
areas of improvement
what startles you
what did you expect vs. what you see
etc.
Last edited: