You are really missing the point. Would you please read post #17 again, it seems to be just a bit beyond your intellectual grasp so far.
B.
+1 to cbdb
With all the confusion, may I add a few words about "square waves".
By "square wave" we mean that when plotted on graph paper or when visual on a screen, it kind of looks like a square (or rectangular, really) shape.
Historically, for testing amps, it makes a great test signal on a scope because for it to look good, the amp has to be OK in phase and to pass equally sine waves from 1/10 the nominal repeating frequency to 10X that freq. Great test and experienced hardware testers (ahem, ahem) can tell a lot immediately by eyeballing the scope using two signals of 200 Hz and 2000 Hz.
In truth, a square or instantaneous (Dirac) pulse is not defined for a band limited system like audio (20-20kHz). In practice, it sort of makes a nice test but due to the scrambling of phases, not a true criterion of performance in home acoustics.
Moving from electronics to acoustics to human perception are other leaps. In the first place, as far as I can tell, there is no meaningful way to sense how good a square wave is on a scope from hearing it. Phase matters a lot in acoustics but not much in hearing.
B.
With all the confusion, may I add a few words about "square waves".
By "square wave" we mean that when plotted on graph paper or when visual on a screen, it kind of looks like a square (or rectangular, really) shape.
Historically, for testing amps, it makes a great test signal on a scope because for it to look good, the amp has to be OK in phase and to pass equally sine waves from 1/10 the nominal repeating frequency to 10X that freq. Great test and experienced hardware testers (ahem, ahem) can tell a lot immediately by eyeballing the scope using two signals of 200 Hz and 2000 Hz.
In truth, a square or instantaneous (Dirac) pulse is not defined for a band limited system like audio (20-20kHz). In practice, it sort of makes a nice test but due to the scrambling of phases, not a true criterion of performance in home acoustics.
Moving from electronics to acoustics to human perception are other leaps. In the first place, as far as I can tell, there is no meaningful way to sense how good a square wave is on a scope from hearing it. Phase matters a lot in acoustics but not much in hearing.
B.
If I understand correctly, the OP's company wants to create a fire extinguisher on acoustical waves principle. They already have a commercial product, that they want to develop further, and asking for (free) help from the DIY community.
The debate on the feasibility of reproducing an ideal 30Hz square wave does not help much to their goal IMHO. (A fire is not interested in wideband phase linear acoustical signals, but try to blow out a candle and you will have better idea...)
The debate on the feasibility of reproducing an ideal 30Hz square wave does not help much to their goal IMHO. (A fire is not interested in wideband phase linear acoustical signals, but try to blow out a candle and you will have better idea...)
sure, but at the the post #1 he didn't asked for perfect squarewave, didn't he? So he needs to define his target regarding output shape.
Then at post #4 he says: "currently using a 10in Sub" and asked to reduce the size under 8in but keeping the same output.
So looks like he don't need a perfec square waves.
He put 4 constrains:
1) tuned frequency is 30Hz
2) can't have an enclosure behind the driver
3) the SPL need to be 90db
4) size
So I said that if he wants a square wave he needs to sum some sine waves. I'm not designing by opinion but just giving a fact that can't be denied.
I can't help but reply, given i find so much contrary to my experiences.
First, a technical comment...there is no need for the DUT (device under test) to pass 1/10th the fundamental frequency to observe a good square wave.
The DUT simply needs to have flat frequency and phase, at the fundamental and up.
By good square wave, I mean something that maintains a solid square shape, with the inevitable wiggles across the plateaus due to a lack of high frequency fill in as frequency increases toward Nyquist. (digital signal)
My experiences are, that even with a band limited system confined to 20-20kHz, good looking dirac pulses and square waves are obtainable.
Certainly electrically; and with considerable attention to tuning, acoustically as well.
(It's probably worth noting, this does require either a full range driver, or linear phase crossovers.)
I can readily agree that I could never sense how good a square waves looks on a scope just by listening to the square wave through a speaker.
But I have found I eagerly look forward to checking out how square waves look, when I feel I have achieved a new level of sound quality with music.
I keep finding, over and over, square waves and impulse responses, look better and better ...as my ears are more and more pleased with the system's musical output.
The formula is simple....get mag and phase as flat as possible.
Music and measurements all come together.
Phase matters in hearing too, IME. It's timing, plain and simple.
I think lots of folks are finding this out as speaker technology keeps improving, with drivers' acoustic centers working together geometrically, and then combined with linear phase alignments.
It makes both on and off axis improve, with gains in clarity and dynamics.
Pro audio designs are improving fast.....
And so are good looking square waves
'ALL' that remains then....is to deal with the room, source material vagaries, and tonal preferences
Ok....back to blowing out fires....
Last edited:
A bit of a problem on well-moderated forums when a post says another person is very ignorant and the natural reply is unprintable.
I hope somebody will comment on the statement by mark100 I've quoted here and that comment may help readers choose sides.
B.
I apologize if I somehow gave offense...but honestly, I really don't see how..
I thought i was simply relating experiences in a non-confrontational manner...
But again, apologies if offense...
If you're looking for evidence to the contrary regarding my statement about 1/10th the fundamental response being unnecessary, I'd be happy to hear any.
I have heard others make the same statement, so I took the time to shoot some square waves though a scope, with high pass filters in place well above the 1/10th point.
I found I could produce good square waves with just the fundamental and higher frequencies.
This is consistent with the math behind square waves, afaict.
The only requirement is that the high pass filter doesn't effect the fundamental frequency and above.
Example: 500Hz square waves looked fine with a 475 Hz high pass filter.
(linear phase 100dB/quarter-octave)
I thought i was simply relating experiences in a non-confrontational manner...
But again, apologies if offense...
If you're looking for evidence to the contrary regarding my statement about 1/10th the fundamental response being unnecessary, I'd be happy to hear any.
I have heard others make the same statement, so I took the time to shoot some square waves though a scope, with high pass filters in place well above the 1/10th point.
I found I could produce good square waves with just the fundamental and higher frequencies.
This is consistent with the math behind square waves, afaict.
The only requirement is that the high pass filter doesn't effect the fundamental frequency and above.
Example: 500Hz square waves looked fine with a 475 Hz high pass filter.
(linear phase 100dB/quarter-octave)
Any over-courteous Canadian might reply, "Gosh mark100, maybe you are simply over-driving your circuits inadvertently and getting square clipping..."
Or someone might reply that your misunderstanding of square waves is supported by your "faulty" evidence.
To produce a 200 Hz wave-form that looks pretty close to squarish, the horizontal portion needs sine components with a far, far longer periods, or else it droops into something more triangular.
Likewise, to produce that sharp squarish rise (and fall), you need sine components with far, far shorter (fast rise) periods.
So as a rule of thumb, to make nice squarish looking oscilloscope traces, the signal passing through the DUT has to pass 1/10 to 10X components, in unaltered amplitude and phase.
Folks skilled at reading square waves can tell a whole lot more by eye and even pinpoint where there are shortcomings in the DUT. But as I've been saying, just a crap-shoot for a mic to pick up squarish looking waves in the acoustic air even if the speaker is excellent.
mark100, how about just signing-on to those last thoughts? It is harmful to readers to have false information endure on social media. Not to mention wasting people's time writing and reading posts like this one.
Perhaps my little sermon will be useful to OP in deciding if he needs a 30 Hz square wave or a 3 Hz sine wave or what it takes for fire treatment or whether a Rice-Kellogg cone driver is the best device to produce such a wave*.
B.
* not likely
Last edited:
With due respect, please have a look at the FFT of a square wave. You have the fundamental and even harmonics with decreasing magnitude. There is no component below the fundamental. You can use a brickwall filter below the fundamental, and it won't change anything.
In real word, it is difficult to apply a HP filter without affecting the phase. So you won't get a nice square wave after HP filtering even with a cutoff below fundamental, because the phase relation of harmonic components is smeared.
In real word, it is difficult to apply a HP filter without affecting the phase. So you won't get a nice square wave after HP filtering even with a cutoff below fundamental, because the phase relation of harmonic components is smeared.
This is a funny thread Where's the OP gone? He managed to get a nice looking square wave from his function generator, he may have got it to come out of his amp because it WAS ABLE TO PASS DC then he got lost.....
Meanwhile, because this is an audio forum, people started having irrelevant arguments
Where's the OP gone? He managed to get a nice looking square wave from his function generator, he may have got it to come out of his amp because it WAS ABLE TO PASS DC then he got lost..... Meanwhile, because this is an audio forum, people started having irrelevant argumentsWow, am I making a great fool of myself and owe mark100 and others a big apology? Or is "fundamental" meant differently?
B.
Last edited:
I've always thought 'fundamental' means the frequency being quoted...the one being measured.
Here's a shot of the 500Hz (fundamental) example I talked about, with a 475Hz hpf filter in place ( a near brickwall as mentioned by lcsaszar)
Pls ignore the plateau wiggles....this o'scope grab came from only a 48kHz processor and about 250ft of xlr cable..
And yep, another DIY audio thread true to form...goin' ever'where haha
Attachments
I am puzzled and wish others would post.
Below is a screen shot from Elliott, "Squarewave testing is a way to test many things at once, but you have to know what to look for. " And "Squarewaves also show you what happens with low frequencies. The slope of the (normally) flat peaks show the effects of bass boost and cut, and the amount of slope is an instant indicator that bass frequencies are either attenuated or boosted. "
Squarewave Testing
B.
Below is a screen shot from Elliott, "Squarewave testing is a way to test many things at once, but you have to know what to look for. " And "Squarewaves also show you what happens with low frequencies. The slope of the (normally) flat peaks show the effects of bass boost and cut, and the amount of slope is an instant indicator that bass frequencies are either attenuated or boosted. "
Squarewave Testing
B.
Attachments
Last edited:
Hi Ben, I'll shut up after this post and hopefully others will chime in.
This is just a try at cutting through what might be causing some confusion...
Yes, the downward sloping peaks show low frequency attenuation.....
but it shows attenuation in the passband being measured...
and the passband for a square wave is the fundamental and everything above it.
Nothing below the passband matters.
I think the reason response down to 1/10th the frequency of fundamental has been deemed historically important,
has been because that was a safe way to insure that no high pass filter was reaching up to/thru the fundamental....
which would of course cause a downward sloping square wave...
Elliot is such a great reference...best single source on info on square wave testing I've found so far...
This is just a try at cutting through what might be causing some confusion...
Yes, the downward sloping peaks show low frequency attenuation.....
but it shows attenuation in the passband being measured...
and the passband for a square wave is the fundamental and everything above it.
Nothing below the passband matters.
I think the reason response down to 1/10th the frequency of fundamental has been deemed historically important,
has been because that was a safe way to insure that no high pass filter was reaching up to/thru the fundamental....
which would of course cause a downward sloping square wave...
Elliot is such a great reference...best single source on info on square wave testing I've found so far...
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.