The Black Hole......

Jump to Latest
#8,081
Jan,

When you have multiple transducers they affect each other and the impedance is not the same as using a single driver. It of course only affects the mechanical part of the impedance. So it is a small difference, but it is the important part of the impedance as it is the part that actually produces the sound.

Then there was the fellow in a court case who claimed to have used an impedance bridge to measure and verify loudspeaker impedances in a stadium system. That of course would not work. I suspect you can figure out why.
 
#8,083
Joe Rasmussen said:
We have an expression down here:

"Hans, you are as mad as a cut snake."

You can't weasel out of this one.

Go back and read:



Did I mention 3.18us and 50,048Hz?

Did I mention Neumann?

Did I not mention Hagerman's RIAA incorporated that 'F4' pole in suibsequent posts?

You know that I did.

You are calling me a liar and yet it is you that is lying.

Instead you should aplogise and you would have walked back.

Try something like this:

"Hi Joe, I think I might have goofed, I did realise what you were saying because I missed out on that time constant you mentioned and that would have meant the response was flat."

I would then have graciously replied:

"Don't worry Hans, we all goof sometimes, but that 17.000V was a good one. Let's just laugh it off as one of those things. Havagoodone!"

That's how real men handle things. Both end up with their self-esteem intact and respect the other person, even they goof.
Click to expand...

Joe ,
This would have been a proper answer:

Sorry Hans for not being more open when I tried to give a prove of my skills where even the axes on the image that I showed were secret.
And sorry that I overreacted when simple questions were asked.

More than anything else this postings shows that you can scream like a suckling pig, but other skills are still a mystery.

Hans
 
#8,084
simon7000 said:
I think you have got it, but the language being used around here reminds me of the hypothetical problem of if we meet an alien species, how do we communicate without common understandings?
Click to expand...

yes, well I wasnt sure pressure wave was appropriate, vs cycle, or if it should be something else to describe each 'push' in the positive/forward pistonic position. I'm not a scientist and i'm just trying to apply the terminology/descriptors I do have in the electrical domain to this acoustic domain, which as I mentioned is not a familiar area for me. I'm well aware i'm butchering the terminology, but i';m afraid its all i've got at the moment 😉
 
#8,085
simon7000 said:
Very basic, when you are at a point from two sound sources where they both contribute equal energy the level increases by 3 dB if the signals are not correlated. If they are correlated and in phase the increase is 6 dB.

However if they are in phase at your location there are other positions where they are out of phase and cancel. That is how a line array becomes directional.

The rule of thumb for drivers to work for a line array is that they must be close to each other, either a half or quarter wavelength depending on your expert. (The difference is really on what you consider to be the acoustic source/center.)
Click to expand...

AllenB said:
Thank you for making my point.
Click to expand...

Hans Polak said:
.
Hans
Click to expand...

I tried, I am human, I failed you and I am sorry.

Cheers!
 
#8,086
can we achieve the same thing with DSP? how much would it impact the result, if, for example, instead of building the baffle so that the acoustic centres are equidistant to the 'mic'/ear etc receiver, they can be on a flat plane, or some other arrangement, but via DSP, the trigger time for each device is altered so that the waves are coincident at a chosen position.
 
#8,087
Chris Hornbeck said:
"Think about this way: If 1000mW gets you 8mm of peak excursion, how much excursion would you get with 62.5mW. Maybe that might help."

An acoustic power ratio of 16 is an excursion ratio of 4. Keep getting the same answer. If we're really lucky, Joe D'Appolito won't see this and answer it with a brilliantly simple two sentence explanation.
Chris
Click to expand...

No, you figured it out. That ratio works out to 2mm peak, or -/+1mm.

Because there are three ways that I can see to figure that out, in power terms it works out since load does not change and we change the voltage, then figure out that halving the current also halves the excursion. Once you realise that, you just work back and you get 2mm.

That is when it dawns on you that you have done all this computational stuff when the simplest answer is displacement and that it needs to be 2mm each to match the 8mm of a single driver. So to get 90dB SPL, all that matters is that you get the same displacement from four drivers that you needed with a single driver. So 8mm becomes 2mm. Oh dear, the actual displacement did not change, ratio was 4, as you said.

Why do our brains so often look for complicated answers when the simplest is right there in front of us? People were bringing all kinds of things like 'acoustic' Watts and stuff.

But there is something else to take from this, that excursion-related distortion can be reduced by using multiple drivers and that only 1mm away from the cone resting place, that the driver is now in the sweet spot a lot more of the time.

The increase in sensitivity is also noted, in terms of the power required is a four-fold increase. Nothing to scoff at!

Of course, as Simon7000 would know, line sources in PA and large acoustic places and even directivity control. My late father was an acoustic engineer and recording engineer and also pioneered large space stadiums in Europe and Philips documented it, late fifties and early sixties. The technology was so basic then and look now? Whew!

There is also another big gain: Heat!

It is still a magic figure that 1000mW now only requires 62.5mW of heat dissipated in the voice coil (16:1). Typical drivers used in Hi-Fi and their relatively small VCs and they get unbelievably hot, and now the crossovers wander all over the place, it does not look nice. Enter waveguides for tweeters and use smaller series capacitors in the high-pass to flatten out the response, do it right, and distortion drops and power handling increases.

I have now done a number of designs based on four series-parallel driver arrays. The gain in sensitivity and increased headroom and clearly audibly lower distortion, along with effortlessness, is not hard to hear.

Now I hope to hear some considered and kind remarks, please?
 
Last edited:
#8,088
simon7000 said:
...where they both contribute equal energy the level increases by 3 dB if the signals are not correlated. If they are correlated and in phase the increase is 6 dB.... That is how a line array becomes directional.

The rule of thumb for drivers to work for a line array is that they must be close to each other, either a half or quarter wavelength...
Click to expand...

Muffed my reply above, just wanted to say thanks for that input.
 
#8,089
Joe Rasmussen said:
Sorry Jan, but you have not taken into adjusting for the 'target SPL' of 90dB.

May I please ask that you reread it?

At 96dB it will, as I have stated, be 1 Watt divided by four - that is fully understood. Each driver will see 250mW, absolutely no argument about that. But when we wind it back, to 90dB SPL from 96dB, then it will be 62.5mW per driver. Compare this to a single driver producing 90dB SPL.

So now we get the same 90dB SPL with 250mW with four drivers.

With a single driver, we needed a full 1 Watt to produce 90dB SPL.

So Jan, I would now ask the question of you:

This is about excursions. If the single driver produces an 8mm peak-to-peak excursion with 1 Watt input, how much excursion will there be at 62.5mW - that is essentially the question.

Could you please do the calculation, I would really appreciate it.
Click to expand...

What calculation? One driver gets 1W and x excursion. Four drivers each get 250mW and x/4 excursion (assuming for the discussion a linear driver).
No power doubling.

This points to a higher principle - the 2nd law.
Think about it. If it would work as you think, you could repeat it with 16 drivers, 64 drivers, 256 drivers, etc. until you have a gazillion drivers with 300dB/W/1m sensitivity. Really?

Jan
 
#8,091
Even easier: construct a device that converts acoustical power into electrical power. Details are not important; say it converts 1W acoustical into 1mW electrical. Use it to convert the driver output into electrical power.

Now use Joe's trick to increase the sensitivity to the point that the electrical output exceeds the 1W electrical input from the amplifier.
Congrats - you constructed a Perpetuum Mobile and several Nobels will await you! You will also become the world's richest person because you will be able to power the planet from your 1W single ended tube amp.

Jan
 
#8,095
So what is the problem?

Have you noted that when crossovers meet up at the crossover frequency, that they are both -6dB, both the midrange and the tweeter... and yet the sum flat?

Please explain that.

It's the same thing that I am pointing at. Does it make sense now?

Edit: Note that both drivers have to be in phase at the crossover frequency.
 
Last edited:
#8,096
You miss my point. You cannot increase efficiency with any number or combination of same-efficiency drivers.
You can, of course increase power handling and whatnot, but NOT efficiency which is output power/input power.
If you could, you could, in the limit, go over unity which is physically impossible.
No math, no equations, just some conceptual thinking.

Jan
 
Last edited:
#8,099
I don't see any paradox here, as long as one accepts this won't scale to provide a yield equal or over 1. In fact, it is the same with coaxial electric motors, or fans, or combustion engines, and the practical limitation (or the "dissipative force" in physics language) is keeping them in sync (that is, practically, they will always attempt to brake each other). The efficiency gain will be lower than 6dB and the more motors you add, the lower the efficiency gain (if at all, beyond certain limits).

You could think at best of the efficiency asymptotically approaching unity, and in case of speakers, any practical implementation will be far from that, anyway.

As a general physics foundation, remember that energy conservation does not apply to dissipative systems, which is why efficiency can't exceed unity.
 
You must log in or register to reply here.
Top Bottom