First speaker to reproduce a 60hz sq wave

[fpitas]

ASR is a tough crowd to impress. People would have questions, "shut up and take my money!" would be the last statement on the last page if it would ever happen.

Perhaps doing some market research was an honest intention but clickbaits aren't popular over here. It has been more than 24 hours now, has the OP left the building?

It's not a bad subject for a casual thread. The title of the thread is a bit misleading, though.

 

[audiofooled]

Yes, a lot of manufacturers on here are kind enough to answer many questions and successful in showing why their product works.

It is an interesting subject but not very easy to prove weather reproducing square waves are one of the key components of a good design, or just a byproduct. They can't do no harm, but as others have pointed, show us how audible it is (how it would translate to an actual performance with signals people actually listen to).

 

[Salt]

how it would translate to an actual performance with signals people actually listen to

Too much reduction, first: show us (in vision) how you may realize your claim.

 

[Mnyb]

Well it would be nice to see a full suite of measurement like a spinorama so that the traditional parameters that makes a good speaker also can be seen ?

And as others have hinted at give us tangible information we are curious ?

If op would provide good information and measurements I’m sure people would be less sceptical to the concept and more friendly .

The world of loudspeaker design has its fair share of products where the designers fetishise one parameter and gladly sacrifice real performance in other aspects to get there.
I just want to know if this another one of those or not

 

[thewas]

The world of loudspeaker design has its fair share of products where the designers fetishise one parameter and gladly sacrifice real performance in other aspects to get there.
I just want to know if this another one of those or not

From my past experience with similar "innovations" the post formulations and content (especially after concrete questions) give a quite unambiguous answer.

 

[MAB]

I mentioned that I had tried compensating phase to improve the rendition of a 60Hz square wave. Might as well finish this train of thought, if only to figure this out for my own sake.

Spoiler: 60Hz Square Wave from a Seas W26 woofer

I measured the Seas W26 nearfield as I did previously.

I note the lack of flatness to below 60 Hz. As I showed previously, the phase shift across frequency is more problematic and will lead to triangles instead of squares unless addressed.

I used rePhase to crush the phase to as flat as I can. I admit to being a newb at this, I wasn't able to do exactly what I wanted, and was in an iterative loop with phase and PEQ. Anybody who can educate me on approach is appreciated. The red trace is with compensation for phase and PEQ for frequency response, the blue is the raw driver's response in the sealed enclosure.

It's not as good as I hoped, but I was able to reduce the overall phase shift across the 3rd and 5th harmonic of 60Hz:

There is even a plateau out to ~400Hz with constant phase.

The resulting 60Hz square wave with the phase compensation and PEQ:

This is a good result.

Without PEQ, the waveform degrades:

Removing the phase compensation, but leaving PEQ produces a roughly triangular waveform as suggested in my simple simulation:

Phase is critical for this specific construction.

The raw driver with no phase compensation and no PEQ isn't much different than PEQ-alone. This is predicted by the simulation, the flatness of phase is more critical than flat frequency response.

As suggested earlier in the thread, add a crossover at 2 or 3 kHz to get rid of all of those phase and amplitude irregularities due to cone and suspension resonances.

Here is phase compensation, and PEQ, and 2kHz crossover at 48 dB/octave:

That's a pretty good square wave.
Obviously, a crossover helps. What a great side-effect!

A few comments:
This is only good for one MLP, a few cm in front of the woofer.
I can get good square waves continuously from 55Hz to about 200Hz. Above 200Hz I find sporadic frequencies that produce good square waves, no surprise since the phase drops steeply above 400Hz in my compensated run. Above ~500Hz all bets are off and I can't reproduce a square wave, again no surprise given the wild swings in phase as the driver resonates.
If I understood rePhase better (and what I am doing with phase compensation in general), and if I can flatten phase below 100Hz and above 400Hz, a larger useable range of continuous square waves is possible than what I was able to produce here.

 

[audiofooled]

I mentioned that I had tried compensating phase to improve the rendition of a 60Hz square wave. Might as well finish this train of thought, if only to figure this out for my own sake.

Spoiler: 60Hz Square Wave from a Seas W26 woofer

I measured the Seas W26 nearfield as I did previously.
View attachment 337861
I note the lack of flatness to below 60 Hz. As I showed previously, the phase shift across frequency is more problematic and will lead to triangles instead of squares unless addressed.

I used rePhase to crush the phase to as flat as I can. I admit to being a newb at this, I wasn't able to do exactly what I wanted, and was in an iterative loop with phase and PEQ. Anybody who can educate me on approach is appreciated. The red trace is with compensation for phase and PEQ for frequency response, the blue is the raw driver's response in the sealed enclosure.
View attachment 337872
It's not as good as I hoped, but I was able to reduce the overall phase shift across the 3rd and 5th harmonic of 60Hz:
View attachment 337896
There is even a plateau out to ~400Hz with constant phase.

The resulting 60Hz square wave with the phase compensation and PEQ:
View attachment 337875
This is a good result.

Without PEQ, the waveform degrades:
View attachment 337876

Removing the phase compensation, but leaving PEQ produces a roughly triangular waveform as suggested in my simple simulation:
View attachment 337878
Phase is critical for this specific construction.

The raw driver with no phase compensation and no PEQ isn't much different than PEQ-alone. This is predicted by the simulation, the flatness of phase is more critical than flat frequency response.
View attachment 337880

As suggested earlier in the thread, add a crossover at 2 or 3 kHz to get rid of all of those phase and amplitude irregularities due to cone and suspension resonances.
View attachment 337886

Here is phase compensation, and PEQ, and 2kHz crossover at 48 dB/octave:
View attachment 337884
That's a pretty good square wave.
Obviously, a crossover helps. What a great side-effect!

A few comments:
This is only good for one MLP, a few cm in front of the woofer.
I can get good square waves continuously from 55Hz to about 200Hz. Above 200Hz I find sporadic frequencies that produce good square waves, no surprise since the phase drops steeply above 400Hz in my compensated run. Above ~500Hz all bets are off and I can't reproduce a square wave, again no surprise given the wild swings in phase as the driver resonates.
If I understood rePhase better (and what I am doing with phase compensation in general), and if I can flatten phase below 100Hz and above 400Hz, a larger useable range of continuous square waves is possible than what I was able to produce here.

Please take a look at my post #53 . This is really a genuine measurement of a (my) loudspeaker reproducing a 60Hz square wave and captured by a microphone. It CAN be done. The real question is weather this is really important or just an eyecandy. Many of us like to take a look at pretty plots but one thing I can tell that this for sure was not taken at MLP. You can't make square reflections.

Another thing that I can tell for sure is that when designing my DIY loudspeakers and sub/bass module, I would not make any other compromises in favor of square waves. Because in the end, that's what it always is. A balancing act of compromises. My question to the OP would be about his.

This short article by an audio professional, Bob McCarthy of Meyer Sound may be interesting and give some hints about phase, impulse and amplitude response. If everything's right, my guess is that pretty square waves would follow, but from where?

 

[Matt_Holland]

Here is phase compensation, and PEQ, and 2kHz crossover at 48 dB/octave:
View attachment 337884
That's a pretty good square wave.
Obviously, a crossover helps. What a great side-effect!

You’re welcome.
Great result!

A thought: just as SINAD below a certain point with electronics is not audible, yet we still gain valuable knowledge from benchmarking products this way, is the loudspeaker square wave measurement an equally valid analysis tool, despite the results not necessarily being audible?

Practical issues would be to have a repeatable method for measuring so that fair comparisons can be made.

Maybe it’s just a bragging rights thing, and in reality tells us very little about a speaker’s performance and how it will interact in a room.

Edited for typo.

 

[Hayabusa]

Please take a look at my post #53 . This is really a genuine measurement of a (my) loudspeaker reproducing a 60Hz square wave and captured by a microphone. It CAN be done.

indeed you are audio fooled

 

[audiofooled]

indeed you are audio fooled

Yes. Look at the date I joined.

 

[gnarly]

Please take a look at my post #53 . This is really a genuine measurement of a (my) loudspeaker reproducing a 60Hz square wave and captured by a microphone. It CAN be done. The real question is weather this is really important or just an eyecandy. Many of us like to take a look at pretty plots but one thing I can tell that this for sure was not taken at MLP. You can't make square reflections.

Hi, the scope capture in #53 shows Ch 1 being measured. That's the line-level signal generator's output. Your mic must be Ch 2.
I say the above, because there's simply no way any acoustic capture can look that pristine. (I know cause I've made the same mistake using REW's scope .)

You can use either Ch1 or Ch2 as the trigger, but retry and see what you get capturing Ch2.

 

[gnarly]

I used rePhase to crush the phase to as flat as I can. I admit to being a newb at this, I wasn't able to do exactly what I wanted, and was in an iterative loop with phase and PEQ. Anybody who can educate me on approach is appreciated. The red trace is with compensation for phase and PEQ for frequency response, the blue is the raw driver's response in the sealed enclosure.

Nice work MAB. You are showing folks that good square waves are simply a product of having flat frequency magnitude response and phase, through the frequency range of the fundamental and several of the fundamentals' odd harmonics .

 

[ctrl]

This is only good for one MLP, a few cm in front of the woofer.

In just a few hours, you have not only repeated the "world sensation" of the initial post, but have provided more and more detailed information than @Fredrickj and have also confirmed that standard physics works (linear FR and linear phase response will reproduce square waves)

If the signal is not only picked up at the loudspeaker terminals but is recorded using a microphone (as you did), the temperature dependence of the speed of sound alone, for example, makes it difficult to repeat the experiment at arbitrary distance, even in an anechoic chamber.

Update: This statement is wrong, phase shift occur but the square wave will not be distorted - the replies from others pointing out the error are correct.
If the phase frequency response was linearized in spring at 20°C (293K), then in summer at 30°C (303K) temperature dependence of the speed of sound leads to considerable phase shifts, so that the square wave will show significant distortion. At a distance of 5m, the phase shift would be around 11°@100Hz and around 110°@1000 Hz (if I haven't miscalculated).

If one really want to get the best possible reproduction of a square wave (in human auditory range), one would have to go at least up to 20kHz, not only 1kHz (or even 420Hz), and then the temperature influences become even greater. So reproduction of a "cut off" square wave at one distance at one temperature is not a quality feature of a loudspeaker (but it is an indication of linear phase and FR - up to cut-off frequency - at the measuring point in space).

Therefore, as others have already said, the sparse information in the opening post does not bode well for an honest attempt to improve the listening quality of potential customers.

 

[RayDunzl]

Fourier Transform Speakers can not only produce 60hz square waves

My experience with square waves is that my speakers can produce a reasonable square at some frequencies, but not at others.

Could you post the result for some other frequencies?

Frequency sweep. Makes a nice square at some points.

 

[audiofooled]

Hi, the scope capture in #53 shows Ch 1 being measured. That's the line-level signal generator's output. Your mic must be Ch 2.
I say the above, because there's simply no way any acoustic capture can look that pristine. (I know cause I've made the same mistake using REW's scope .)

You can use either Ch1 or Ch2 as the trigger, but retry and see what you get capturing Ch2.

Ok, fair point, thank you. On REW scope, CH1 is a measurement channel, where my mic happens to be. In my setup CH2 (reference channel) contains nothing but noise. Here's "less pristine" version with all channels, including A+B math.

 

[gnarly]

My experience with square waves is that my speakers can produce a reasonable square at some frequencies, but not at others.

That's my experience too.

I got into square wave testing when i first started learning to use FIR to setup active DIY speakers. Here's a set from a 4-way main speaker made for use from 100Hz up.
Cherry picked no doubt Lol. I was excited then about the super results FIR could give haha.

Later, after I came to fully learn that square waves equal nothing more than excellent flat frequency mag and phase response, I said to heck with square waves.

I'd expect your Martin Logan's, with their lack of crossovers, can make for some pretty nice square waves at select frequencies.
As I'd bet you know, crossovers other than first order, bring too much phase rotation to let squares hold shape.

That said however, linear phase crossovers make squares possible....no matter how many crossovers (ways) are used, and no matter how high their orders.

Next time I do what I call a "perfect spot-tune" on a full-range 5-way, I'll try to remember to capture a big set of squares, beauties (and warts).
I think the set will be mostly good...at least for at the one measurement spot Lol. I call doing stuff like that vanity measurements Lol

 

[gnarly]

Ok, fair point, thank you. On REW scope, CH1 is a measurement channel, where my mic happens to be. In my setup CH2 (reference channel) contains nothing but noise. Here's "less pristine" version with all channels, including A+B math.

Cool. Glad i could help.
What do you use the A+B math for?

 

[audiofooled]

Cool. Glad i could help.
What do you use the A+B math for?

Well, in this case, its measurement + noise. I just thought I turn on all the channels in case people think I'm pulling their leg. But as you said, to heck with square waves.

 

[Mnyb]

Is this a hoax I tried to google this “company” to find some pictures of thier products and specs ?

 

[fpitas]

Is this a hoax I tried to google this “company” to find some pictures of thier products and specs ?

Well, if they were living in a cave on Mars all these years and were not aware of the extensive history, and that a lot of us have used phase flattening, you had best avoid them anyways.