First Audio Build (Need Math Help!)

[Mr.ac]

Ah, I see. Yeah that head unit would work out fine.
So any name brand amp should work. Just make sure the amp is 2ohm stable. Which 90% are.

[stugray]

Originally Posted by TOOSTUBBORN2FAIL

If you can prove what you are claiming, I will listen. I can objectively prove what I am stating. Tomorrow I will load up some varying music on the computer and post the spectrum analysis of each.

Think of it from a spectral density perspective.
An ideal speaker response curve might be to start at zero response at 0 HZ and ramp up to max response at say 10 Hz and remain flat until the response drops off again as the freq approaches 20kHz to zero response above 20 kHz.

But if I drive it with a broad spectrum noise source from 1 Hz to 1 MHz with the energy spread evenly over the whole bandwidth, then what happens?
In fact, I could drive it with a sinusoidal signal and sweep it from 1 Hz to 1 MHz.

If the speaker only responds above 10 Hz and below 20kHz, then where does all that energy go when the amp is driving from 0Hz to 10Hz and 20kHz up to 1 MHz?
Same thing as if I apply the energy at 0Hz, DC it goes straight to heat as the coil cannot move the cone at those frequencies.

Take the example mentioned above about driving with a square wave instead of a sinusoid.

If you take a FFT of a pure tone you get a discrete signal, a single carrier.



If you take a FFT of a square wave, you get a fundamental frequency (carrier) at the prime frequency and an infinite number of side lobes



The speaker cant move in a perfect square wave, so it cannot respond to the components of the FFT that are outside of the response curves.
If you perform a FFT of music some of the components found will be outside the response curve of the speaker.
If you integrate under the FFT outside the speaker response curve you will see there is power out there.
All of that power turns to heat in the speaker (transducer).

So if you drive a nice pure tone and keep it within the speakers response curve, all (most) of the energy is converted to sound and the coil doesnt get hot.
If you distort the pure tone, some of the FFT components will fall outside the response curve and get converted to heat.
The same theory applies to our vibration tables where we put entire spacecraft on the table and shake the shit out of them.
Sometimes we sweep nice sinusoidal signals and sometimes we ramp up to higher & higher levels with white noise.

Guess which signal the tables can output for longer before they overheat?

[Shankenstein]

Originally Posted by stugray

Think of it from a spectral density perspective.
An ideal speaker response curve might be to start at zero response at 0 HZ and ramp up to max response at say 10 Hz and remain flat until the response drops off again as the freq approaches 20kHz to zero response above 20 kHz.

But if I drive it with a broad spectrum noise source from 1 Hz to 1 MHz with the energy spread evenly over the whole bandwidth, then what happens?
In fact, I could drive it with a sinusoidal signal and sweep it from 1 Hz to 1 MHz.

Good post! From an audio perspective, there are plenty of "simulated program signals" that different groups have developed over the years (IEC, EIA, CCIR, etc). Most will focus on the vocal and guitar range... with some power elsewhere.



Anything that's below the speaker's output will generally get treated like DC... which heats the coil and doesn't contribute to cooling. At higher frequencies, the impedance is usually high enough that the power output is irrelevantly small.

For high-end OEM car audio speakers, with aluminum voice coils and formers... you might see 200+ Watts of DC power dissipation. High-end 6.5" door speakers might be ~150 Watts. High-end midranges might be ~50 Watts.

At high frequencies, you're likely to kill a speaker from power handling long before the suspension is wrecked. Even at max power on clipped program signal, most quality speakers can handle max power for 100 hours or more (requirements for German car companies).

The GT86 speakers are pretty robust (albeit inefficient). I would expect that they're just slightly under these targets.

[TOOSTUBBORN2FAIL]

If the peak to peak amplitude of the sine wave and square wave are the same, then the square wave will send more heat to the coil on average over time. As long as that heat level isn't past what the coil can handle, it won't do damage.

I personally over power all of my speakers and exercise volume control based on the mastering of the music. I also look for the most dynamic recordings I can find of the music I like.

[MidnightMonkey]

This has gotten really, really deep. Theory interests me a lot, but I haven't been in the mood to even remotely attempt to digest this stuff.

Hey stugray, I probably work for someone that indirectly works with your orginization. I'm a satellite network coordinator if that helps you figure out who I'm talking about.

Anyways. Do any of you guys know of somewhere I can get 'pink noise' tracks to do the EQ/ time alignment? Or have a better recommendation?

[stugray]

Originally Posted by MidnightMonkey

Hey stugray, I probably work for someone that indirectly works with your orginization. I'm a satellite network coordinator if that helps you figure out who I'm talking about.

If you just participated in the JPSS-1 JCT2 (all last week), then my callsign was BATC-TC.




Speaker porn:





The acoustic test cell at Lockheed that I helped build (in the background with the blue door):

[Shankenstein]

Originally Posted by TOOSTUBBORN2FAIL

I also look for the most dynamic recordings I can find of the music I like.

Word on that!

Dynamic range has been disappearing from recordings over the last few decades. It's good for radio... but not good for recreating the depth and realism of live audio.

Do you have any music recommendations?

Lately, I've been enjoying The Wood Brothers, Trombone Shorty, and classic Santana. On a lesser-quality side, I've been enjoying the evolution of trap music (rap samples in a EDM shell) from the classic Shawty Redd songs to more modern guys like Keys N Krates, Yellow Claw, Flosstradamus, etc.

[MidnightMonkey]

Originally Posted by stugray

If you just participated in the JPSS-1 JCT2 (all last week), then my...

Nah man I don't handle any launch, geosync, or station keeping. All I do is spacecraft telemetry and comms management.

Pretty bad ass shots. What platform are those birds?

[stugray]

Originally Posted by MidnightMonkey

Nah man I don't handle any launch, geosync, or station keeping. All I do is spacecraft telemetry and comms management.

Pretty bad ass shots. What platform are those birds?

I thought you might have been handling the data from the MOST during the Compatibility Test of JPSS.
The majority of my experience is with the Deep Space Network.
I have typically been one of the Command & Data Handling subsystem engineers for missions like Stardust, Genesis, Deep Impact & Kepler.

None of the pictures above are from my spacecraft. I just recognized the Maryland Sound speaker setups from a Google image search.
The third pic is the Launch Abort System (LAS) for the Orion capsule.

Sorry OP for the thread jack.

I tried to give the simple answer.
If the amp & speaker IMPEDANCEs dont match then all bets are off.
IF you drive a mismatched speaker with an amp of a different impedance, then you risk damaging the speaker, the amp or both.

At a minimum you will get distortion due to the mismatch of the impedances.
If the impedance is mismatched, then you get reflections which result in a noisy signal.

An Impedance is a complex value (literally not just figuratively) which includes components of Inductance and Resistance.
The inductance component varies with response to frequency.
A woofer will have a very large inductance (BIG coil moves a big cone) while a tweeter may have a small inductance (small coil or none at all - electrostatic or piezo).

And for the skeptics: I have a blown set of JBL-L90s in the basement with blown coils.
I used them for ten years and they could handle my amp at full volume with regular "Rock music".
My son blew them in one evening listening to Slipknot.
So my example is not just anecdotal.

[TOOSTUBBORN2FAIL]

Originally Posted by stugray

And for the skeptics: I have a blown set of JBL-L90s in the basement with blown coils.
I used them for ten years and they could handle my amp at full volume with regular "Rock music".
My son blew them in one evening listening to Slipknot.
So my example is not just anecdotal.

What regular rock music? Like I said before, Slipknot is very compressed. If by regular rock music you mean classic rock, it usually has 25% or less of the total average output that Slipknot has, over time. That could very easily explain things, which is what I've been trying to say. FWIW, I've killed one set of speakers, with some EDM. It was entirely my fault, from poor volume control. I say this, because my system was set up to handle any normal music at full volume without issue. But, EDM is not normal, and this particular song was brickwalled even further than Slipknot is. FWIW, I could play Slipknot on my stereo at full volume without damage.

Crest Ratio is the difference in decibels between the loudest peak in the sample to the lowest average sound that is above the noise floor.

For instance, this is Slipknot. Super compressed, to the point of being clipped. 6db crest ratio, this is like playing pink noise through the speakers. Pink noise is generally what thermal power ratings are created with. If your crossovers, gain, and eq aren't out of wack, speakers should be able to handle this at rated power without issues.



This is a new, remastered version of Hotel California. Just the first minute or so. This one was compressed compared to the original version, but much more dynamic than Slipknot. Less demanding on speakers, because playing this at the same volume setting as Slipknot will send less power to the speakers. 7db crest ratio.



This is a vinyl rip of Hotel California, much more dynamic than the remastered version, much less power going to the speakers at the same volume. 15db crest ratio.



For comparison, here is the Telarc version of the 1812 Overture, my favorite classic piece. Even more dynamic than Hotel California, even less average power going to the speakers. 18db crest ratio.




Here, for fun is Slayers Angel Of Death, the remastered cd version. 7db crest ratio.



This is the vinyl version of Angel Of Death. 13db crest ratio.

[stugray]

@TOOSTUBBORN2FAIL

Excellent post. Sounds like we are saying the same thing from different disciplines.

If you could take those exact sound wave scope traces (time base) and run them through a FFT (Spectrogram - frequency base), you would also see that the spectral energy of the Hotel California is likely in the response region of the speakers.
The same FFT spectrogram of the slipknot would likely have more harmonics outside the response region.

Here are a few songs run through Spek:














You can see how the first two slipknot songs are more spectrally dense and contains more energy.

Same thing comparing the pink noise to the white noise.

And then the Creed and Mangione are clearly less dense than the slipknot with the Evanescence in the middle.
That is why I say you will heat the speakers more with slipknot than with the Mangione.
(changed my comparison just to fit the pics).

[Mr.ac]

Originally Posted by MidnightMonkey

This has gotten really, really deep. Theory interests me a lot, but I haven't been in the mood to even remotely attempt to digest this stuff.

Hey stugray, I probably work for someone that indirectly works with your orginization. I'm a satellite network coordinator if that helps you figure out who I'm talking about.

Anyways. Do any of you guys know of somewhere I can get 'pink noise' tracks to do the EQ/ time alignment? Or have a better recommendation?

yep and meh. Still kind of point less (in my opinion)

If you want a great and one that I personally use test and tune cd look for Bass Mechanix Sonic Overload.
it has everything you need to blow shit up, or tune it just right.
from 1hz to 200hz Pink noise, shine waves, and just old fashion bass farts.

[MidnightMonkey]

Ok, guys. I'm wiring up the HU harness right now. Since I have to wire the remote wire to the power antenna wire, where then will I connect the amp remote wire to?

[TOOSTUBBORN2FAIL]

I would use the remote wire for both, but I would use heat shrink ed solder joints to connect them.