It's actually more like cabin gain for a given vehicle; the vehicle is the biggest influence. The sub itself can have some interactions with that but much lower in level, methinks.
It would make an interesting article to research more how much the cabin gain varies from vehicle to vehicle. I could also move a sub around within the trunk or hatch, and see what happens.
Your comment inspires me to reach out to some friends who've researched this more. I also feel like the Society of Automotive Engineers and maybe the Audio Engineering Society published some stuff about this. Or maybe it was Tom Nousaine. If anyone knows, please comment, I'd like to re-read about this topic.
Thanks for the Linkwitz link. I don't know if I can run high SPL tones long enough to do that without the neighbors getting bugged I'll have to try it. But certainly the Maximum physical excursion could be measured. I had forgotten about that, so glad you mentioned it.
As for the AES standard, frankly, it's outmoded. That looks like it was based on a paper by JBL's Mark Gander. OK, "outmoded" is the wrong word, but I think folks are more comfortable with actual acoustic distortion. And I would tend to AVOID Fs for a few reasons:
1) The resonance does present a kind of worst case, which is why Mark chose it. But I want to focus on the basic magnet/surround geometry which limits ALL frequencies.
2) Testing at Fs generates a MUCH louder and more intense SPL, which again is a practical problem in my neighborhood.
3) I think readers are more comfortable seeing all the woofers tested at the same frequency. Testing at Fs would mean all the woofers at different frequencies. Testing at something like 10 Hz means they should all be well below the in-box resonance.
Last edited:
Only if you're doing the testing with the driver mounted in a box. You don't need to mount the driver in a box to text Xmax.
The point is NOT to mount it in a box.
"Testing at Fs generates a MUCH louder and more intense SPL"
No it won't. SPL is a simple function of surface area, excursion and frequency (and in this case acoustical cancellation). At 20-30Hz Fs that is typical you won't see that much SPL even at high excursion. Distortion will probably be way louder.
You can test at different frequencies but I'm not sure there will be a benefit. But best to verify.
Testing both at 10Hz and at Fs might be a good compromise.
Alternatively you can state excursion for 1 2 5 10 and 20% THD.
Should give an idea how "good" it will sound at any given volume.
For example brand X with less Xmax might be cleaner than brans Y with more Xmax, due to different construction.
"Testing at Fs generates a MUCH louder and more intense SPL"
No it won't. SPL is a simple function of surface area, excursion and frequency (and in this case acoustical cancellation). At 20-30Hz Fs that is typical you won't see that much SPL even at high excursion. Distortion will probably be way louder.
You can test at different frequencies but I'm not sure there will be a benefit. But best to verify.
Testing both at 10Hz and at Fs might be a good compromise.
Alternatively you can state excursion for 1 2 5 10 and 20% THD.
Should give an idea how "good" it will sound at any given volume.
For example brand X with less Xmax might be cleaner than brans Y with more Xmax, due to different construction.
Have you considered measuring the loudspeaker transfer function using constant current to determine Xmax and driver nonlinearity. You need a constant current source, be careful not to cook the voice coil. Set up a guage to measure cone displacement and a simple platform to stack weights on the cone without damaging it. Start with say 1/2 amp and measure the cone displacement stack enough weight on the cone to return it to the rest position, a few data points will give the compliance then just measure cone displacement vs current till you see marked non linearity which would be Xmax. The higher current tests should be done fairly quickly It is possible to monitor voice coil voltage and calculate the resistance and from that work out the temperature and keep it within safe limits most likely the speaker will reach Xmax before the voice coil reaches its thermal limit.
Clearly it is correct that there is a large amount of variance, but there is also a large amount of consistency, compared to, say, a typical home setup, if only because the materials and volume are relatively similar (right down to the listening positions) in the average vehicle compared to the average home.
Given that, I still believe there would be value in trying to measure cabin gain of various sub systems in a standardized cabin, with the dimensions (or the sample vehicle) clearly stated and unchanging from test to test.
This forms a point of reference to compare to the usual speaker criteria ... whether that be "typical" values given by a car audio manufacturer or from the anechoic derived values in current driver literature.
Not sure if this is the very article you were interested in, but Tom goes into it a bit here:
"Autosound Cabin Gain at Low Frequencies"; Tom Nousaine, Car Stereo Review March/April 1992 *pdf; download = 2.8 MB
All of Tom's stuff can be found at his site: Nousaine dot com
Last edited:
Clearly that is correct.
However, I still believe there would be value in trying to measure cabin gain of various sub systems in a standardized cabin, with the dimensions (or the sample vehicle) clearly stated and unchanging from test to test.
This forms a point of reference to compare to the usual speaker criteria ... whether that be "typical" values given by a car audio manufacturer or from the anechoic derived values in current driver literature.
Not sure if this is the very article you were interested in, but this might ring a bell:"Autosound Cabin Gain at Low Frequencies"; Tom Nousaine, Car Stereo Review March/April 1992 *pdf; download = 2.8 MB
All of Tom's stuff can be found at his site: Nousaine dot com
Well, I meant IN a (sealed) box, to avoid cancellation from the rear wave. In that case, the SPL would be much louder at Fs than at 10 Hz.
I'm not picturing testing the SPL without a box...a near-field measurement I suppose might get around the cancellation problem, but would ignore noise/distortion generated from areas of the cone not next to the microphone. What exactly did you have in mind?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.