I posted this idea over at the Home Theater Forum last week, but I'd like to bounce it off you guys as well, and get some feedback.
Yeah, I know the correction tables for RS SPL meters go pretty low, but how 'bout this for a down and dirty DIY solution: An ELF mic.
Take any small midrange driver you have lying around and mount it in a very small enclosure (just big enough for the driver to fit). If you can find a sealed-back mid, you can skip the enclosure altogether. The parameters of the driver don't matter, and neither does the absolute size of the enclosure as long as it's tiny and well sealed (though it would probably be a good idea to include a pressure-equalizing pinhole) Attach thin wire leads and connect them to a voltage tester with an averaging function. Congratulations, you've finished your test mic.
The highish Fs driver in the small enclosure will give you a highish Fb of 200-500Hz. That's just dandy because this microphone will only be used in the linear region below its Fb, so the higher the Fb, the better. This idea rips off the ELF concept that makes use of the predictable 12 dB/oct roll-off below Fb, only we're making a microphone instead of a subwoofer.
You'll need to calibrate your ELF mic against some standard, like your trusty old RS meter. With the two mics close together, Play a test tone, say 100-120Hz, where both units are linear. (It should be at least an octave below your ELF mic's Fb, where the 12dB/oct roll-off behavior is in full swing.) Adjust the tone volume until the RS meter shows a nice round dB number. Note the number and the indicated voltage coming from your little ELF mic (peak or average, it doesn't matter). Now you're calibrated.
When you measure and plot data points, just add 12dB per octave below your calibration point. With reference to your calibration voltage, a doubling or halving of the voltage means a 6dB rise or drop.
If you're electronics savy, add a dual integrator to the ELF mic output to correct for the roll-off and save yourself some math. Add a meter and a set of calibrated switchable resistors for range, and you've got a self-contained unit.
A potential problem and a solution:
Problem: if the mic itself is vibrating (like picking up structural vibrations--and what doesn't vibrate at some frequencies?) motion along the driver/diaphragm axis will induce voltage because of the inertia of the diaphragm.
Solution: mount a second driver/diaphragm opposite the first wired in parallel in phase. Now any microphone motion along the axis of the voice coils will induce opposite (canceling) voltages.
PE sells sealed-back mids for <$5. Seems to me the simplest build would be two of those bolted face to face with 1-2" spacers between--no box, no fuss.
Unless Thiel and Small were lying, this mic should be nice and linear as low as you want to go.
Haven't tried it yet, but it seems like it should work well. What do you guys think?
Bill
Yeah, I know the correction tables for RS SPL meters go pretty low, but how 'bout this for a down and dirty DIY solution: An ELF mic.
Take any small midrange driver you have lying around and mount it in a very small enclosure (just big enough for the driver to fit). If you can find a sealed-back mid, you can skip the enclosure altogether. The parameters of the driver don't matter, and neither does the absolute size of the enclosure as long as it's tiny and well sealed (though it would probably be a good idea to include a pressure-equalizing pinhole) Attach thin wire leads and connect them to a voltage tester with an averaging function. Congratulations, you've finished your test mic.
The highish Fs driver in the small enclosure will give you a highish Fb of 200-500Hz. That's just dandy because this microphone will only be used in the linear region below its Fb, so the higher the Fb, the better. This idea rips off the ELF concept that makes use of the predictable 12 dB/oct roll-off below Fb, only we're making a microphone instead of a subwoofer.
You'll need to calibrate your ELF mic against some standard, like your trusty old RS meter. With the two mics close together, Play a test tone, say 100-120Hz, where both units are linear. (It should be at least an octave below your ELF mic's Fb, where the 12dB/oct roll-off behavior is in full swing.) Adjust the tone volume until the RS meter shows a nice round dB number. Note the number and the indicated voltage coming from your little ELF mic (peak or average, it doesn't matter). Now you're calibrated.
When you measure and plot data points, just add 12dB per octave below your calibration point. With reference to your calibration voltage, a doubling or halving of the voltage means a 6dB rise or drop.
If you're electronics savy, add a dual integrator to the ELF mic output to correct for the roll-off and save yourself some math. Add a meter and a set of calibrated switchable resistors for range, and you've got a self-contained unit.
A potential problem and a solution:
Problem: if the mic itself is vibrating (like picking up structural vibrations--and what doesn't vibrate at some frequencies?) motion along the driver/diaphragm axis will induce voltage because of the inertia of the diaphragm.
Solution: mount a second driver/diaphragm opposite the first wired in parallel in phase. Now any microphone motion along the axis of the voice coils will induce opposite (canceling) voltages.
PE sells sealed-back mids for <$5. Seems to me the simplest build would be two of those bolted face to face with 1-2" spacers between--no box, no fuss.
Unless Thiel and Small were lying, this mic should be nice and linear as low as you want to go.
Haven't tried it yet, but it seems like it should work well. What do you guys think?
Bill
Hi Bill
I expect this idea to work quite well and being cheap also.
But bear in mind that piezo-resistive pressure sensors are also quite cheap nowadays ......
What do you want to use it for ? Do you want to measure subwoofer frequency response, use it as sensor for some kind of MFB, detect tornadoes....... ????
Regards
Charles
I expect this idea to work quite well and being cheap also.
But bear in mind that piezo-resistive pressure sensors are also quite cheap nowadays ......
What do you want to use it for ? Do you want to measure subwoofer frequency response, use it as sensor for some kind of MFB, detect tornadoes....... ????
Regards
Charles
Thanks, Charles.
I don't know much about piezo solutions. Are they easy to implement and accurate at extremely low frequecies?
Detecting tornados sounds fun, but I was just trying to come up with a good way to test the extreme low-frequency performance of a new bass driver I'm soon to prototype. It's designed for very low Fs and very large linear excursions--like 6" P-P. I know it sounds ambitious, but I'm planning a small 2nd-order 2-driver box capable of 120+ dB (half-space) at 10 Hz.
Bill
I don't know much about piezo solutions. Are they easy to implement and accurate at extremely low frequecies?
Detecting tornados sounds fun, but I was just trying to come up with a good way to test the extreme low-frequency performance of a new bass driver I'm soon to prototype. It's designed for very low Fs and very large linear excursions--like 6" P-P. I know it sounds ambitious, but I'm planning a small 2nd-order 2-driver box capable of 120+ dB (half-space) at 10 Hz.
Bill
I read the subject, and while I waited for the page load, I was thinking to myself "just use an old speaker and measure the voltage induced by the pressure from the speaker that's playing".
So yeah, I think that'll work just fine.
The accuracy of the piezo at a given frequency depends on the size of it. They work similarly, though, so implementing should be pretty easy.
As far as accuracy at low frequencies, you need to choose the right piezo, as they will all have a given frequency response.
So yeah, I think that'll work just fine.
The accuracy of the piezo at a given frequency depends on the size of it. They work similarly, though, so implementing should be pretty easy.
As far as accuracy at low frequencies, you need to choose the right piezo, as they will all have a given frequency response.
- Status
- Not open for further replies.