Hi,
I want to reproduce single cycle waves (below 200Hz) using a 12" driver (either woofer or subwoofer). As I understand the time domain behaviour of drivers don't allow single cycle sines. With a real life driver, the oscillations take time to build and time to die which is fine for music but not for my application.
Following are some of the ways to circumvent the problem
1) Use a servo controlled driver, servo control has good time domain behaviour.
2) Use multiple smaller drivers, say 8 of 5" drivers, smaller drivers will inherently have better time domain behaviour.
3) Use a driver with very high BL, the high BL will generate lot of back EMF, the high back EMF will allow the amplifier to control the driver better yielding better time domain behaviour. Of, course, very high BLs damp the lows, but thats not a problem since EQ before the amp can bring up the lows, as shown by Don Keele in his paper http://www.xlrtechs.com/dbkeele.com/PDF/Keele (2003-10 AES Preprint) - Nom vs True Eff High BL.pdf
What are your thoughts?
Thanks and Regards,
WA
I want to reproduce single cycle waves (below 200Hz) using a 12" driver (either woofer or subwoofer). As I understand the time domain behaviour of drivers don't allow single cycle sines. With a real life driver, the oscillations take time to build and time to die which is fine for music but not for my application.
Following are some of the ways to circumvent the problem
1) Use a servo controlled driver, servo control has good time domain behaviour.
2) Use multiple smaller drivers, say 8 of 5" drivers, smaller drivers will inherently have better time domain behaviour.
3) Use a driver with very high BL, the high BL will generate lot of back EMF, the high back EMF will allow the amplifier to control the driver better yielding better time domain behaviour. Of, course, very high BLs damp the lows, but thats not a problem since EQ before the amp can bring up the lows, as shown by Don Keele in his paper http://www.xlrtechs.com/dbkeele.com/PDF/Keele (2003-10 AES Preprint) - Nom vs True Eff High BL.pdf
What are your thoughts?
Thanks and Regards,
WA
This should do it, no drivers.
Eminent Technology TRW-17 Rotary Subwoofer | Sound & Vision
Eminent Technology: home
Eminent Technology TRW-17 Rotary Subwoofer | Sound & Vision
Eminent Technology: home
Last edited:
Cutting off a sine wave is the same as clipping, look at the sharp angle at cuttoff. This means you need hi frequency to achieve this. So smaller speakers would work better.
Yes, a 1Hz sine burst has a wide spectrum, with a sinc envelope.
https://www.ietlabs.com/pdf/GR_Appnote/IN-105 Frequency Spectrum of a Tone Burst.pdf
https://www.ietlabs.com/pdf/GR_Appnote/IN-105 Frequency Spectrum of a Tone Burst.pdf
cbdb is right - the driver's bandwidth (meaning how extended the high end is) determines how fast it can response to a transient like an impulse or other fast rising waveform. Smaller drivers will be better in general, but its not the size (e.g. cone diameter) but rather the high frequency extension that is the reason for that.
Also, keep in mind that the crossover will also band limit the driver. This means starting out with a driver with a wide bandwidth and then limiting that bandwidth with a low pass filter in the crossover will be much the same as using a "slower" driver that has less raw high frequency extension to begin with.
Frequency and time domain responses are not independent.
Thanks all, for this thread.....it has my mellon spinning.
I'm very aware that transient response equals frequency response.
And that perfect time domain response equals perfect mag (and phase) response across the full audio spectrum. Heck, it's been one of my audio mantras.
But until this thread, the idea never hit me that a driver couldn't replicate a single pulse of a single sine wave....ie one and only one, 360 degree period.
(Not a repeating burst ala the GenRad paper. Thx for that btw )
Anyway, i'm wondering just why can't a sub accurately reproduce say one single 50Hz sine wave.????
I hear the comment about sharp cutoff at start and finish...but ?????
I've been compelled to figure out how to test this.
Gonna try a single-period sine burst as soon as some parts come in...
I'm very aware that transient response equals frequency response.
And that perfect time domain response equals perfect mag (and phase) response across the full audio spectrum. Heck, it's been one of my audio mantras.
But until this thread, the idea never hit me that a driver couldn't replicate a single pulse of a single sine wave....ie one and only one, 360 degree period.
(Not a repeating burst ala the GenRad paper. Thx for that btw )
Anyway, i'm wondering just why can't a sub accurately reproduce say one single 50Hz sine wave.????
I hear the comment about sharp cutoff at start and finish...but ?????
I've been compelled to figure out how to test this.
Gonna try a single-period sine burst as soon as some parts come in...
Most tone burst testing uses several cycles, and often with an envelope.
https://pearl-hifi.com/06_Lit_Archive/07_Misc_Downloads/Shaped_Tone-burst_Testing.pdf
https://pearl-hifi.com/06_Lit_Archive/07_Misc_Downloads/Shaped_Tone-burst_Testing.pdf
Another good paper, thx again.
And yep, i'm familiar with sine bursts, and shaped sine bursts, having run the CEA-2010 test on subs.
What has me so intrigued, is the question of ....does the start and stop of a sine wave constitute the need for extended response, past the particular frequency of the sine wave.....
Took me all day to figure out just how to pulse a single sine wave....
only to find out i need a gizmo type amplifier to hook to my arb function generator. At least i found the gizmo amp cheap on ebay....
I'm a few days away from making some thumps lol
And yep, i'm familiar with sine bursts, and shaped sine bursts, having run the CEA-2010 test on subs.
What has me so intrigued, is the question of ....does the start and stop of a sine wave constitute the need for extended response, past the particular frequency of the sine wave.....
Took me all day to figure out just how to pulse a single sine wave....
only to find out i need a gizmo type amplifier to hook to my arb function generator. At least i found the gizmo amp cheap on ebay....
I'm a few days away from making some thumps lol
I dunno...doesn't voltage start at zero at the beginning of a sine wave?
How is that asking for 'maximum velocity instantly ?
How is that asking for 'maximum velocity instantly ?
The more abrupt the start and stop of the envelope, the wider the frequency bandwidth.
A sine with a rectangular window is very different in spectrum from an infinite sine.
Hamming and Hanning windows, less so.
WINDOWS | Chapter Three. The Discrete Fourier Transform
Last edited:
Sure, but for a 50Hz sine wave, how great is that rate of change starting at zero ?
The wave has a whole dang 20ms to make one period ..yes?
The wave has a whole dang 20ms to make one period ..yes?
I mean, forget envelopes, windows, and such....
How does a true single sine wave come out?
It will take time to accelerate and so won't be at maximum velocity at zero crossing, there will always be an envelope, you can't ignore it.
A sine wave "starts" at t = minus infinity, and "stops" at t = plus infinity.
Anything of a lesser duration is not a sine wave, and has a window, so the spectrum
is altered from a single discrete frequency.
A finite length waveform of any type is mathematically a transient, and is not steady state like
a true sine (or other infinitely repetitive waveforms). Finite length waveforms also have a wide,
continuous spectrum containing all frequencies (to some degree, depending on the windowing),
and not just discrete frequencies.
Last edited: