A Test. How much Voltage (power) do your speakers need?

[5th element]

Originally Posted by a.wayne

What 5th is saying is correct but it does not apply to what Tom and I are speaking of ....

And with all due respect, neither of you have understood this test.

Originally Posted by a.wayne

5th your are applying the fix digital signal as supplied by pano after using an variable analog signal to set gain with no reference .

There isn't supposed to be a reference that's the point. No one knows how loud it is in decibels because they do not need to.

The reference point in this case is set by the listener at their listening levels. This is the important part. It is done with music, not with a sine wave or with pink noise, but set realistically by using the most realistic thing you can get, music. One sets the volume level by setting it to the max that they would ever listen at.

Now how loud this is isn't important. Not to you, not to me, heck, not even to them. What's important is that their volume control is set to 5 and they know that they never listen to music any louder then 5 and that they only ever listen to music through their CD player.

You cannot get any louder, nor any larger in signal level, or larger dynamic peak, then is replayed at 0dB.

Ergo if you play a signal of 0dB though your system at volume setting 5, and the amplifier doesn't clip, then nothing you ever care to play through the system at volume 5 and below will make the amplifier clip (providing current limitations within the amplifier are ignored).

Tell me what is it about the above that you object to. What scenario can you invent where we are now going to cause the amplifier to clip with the volume set at 5? We've already shown that it wont clip with a 0dB signal, so what is it you think is going to come along that will make the amplifier magically start clipping?

[5th element]

Originally Posted by mr_push_pull

my thoughts exactly, I somehow have the feeling that there's a lot of confusion floating around but can't put my finger on it. maybe it's me who doesn't get it.

Oh no, you get it, you got it first time round.

[mr_push_pull]

Originally Posted by 5th element

Oh no, you get it, you got it first time round.

what do I win? an oscilloscope by chance?
anyway I didn't expect this to turn into such a puzzle.

[5th element]

Originally Posted by mr_push_pull

what do I win? an oscilloscope by chance?
anyway I didn't expect this to turn into such a puzzle.

I can highly recommend Tek scopes btw, I've got a second hand 2225 and its marvellous.

I didn't expect this to turn into such a puzzle either. My guess is that people are looking at this from the wrong direction and instead of realising exactly what it is we're trying to do here are vastly over complicating things and it's this that's causing the issue.

[Pano]

It puzzles me that it is such a puzzle. It's such a simple test I don't see how it can be so hard to understand.

Maybe folks are thinking there is some reference SPL in this test. There is not. It does not matter if "Loud Enough" for you is 70dB, 107dB or 45dB. All you are asked to do is find "Loud Enough" by playing a few good recordings. The absolute SPL does not enter into it.

Forget what books, papers, websites and reference material tell you should be your reference level, set it yourself with music. Although it may be interesting for you to know how that level compares to other sounds, that is not the point of this test. See post #291

[JRKO]

Originally Posted by Pano

It puzzles me that it is such a puzzle.

Your not kidding! the instructions in post no.2 are as clear and concise as can be. To be honest, reading through, I've found the confusion/OT nonsense very distracting. It made me think I had not understood the test!

The test is this: play a known sample at preferred max volume and measure = the max voltage your speakers will need - then post your results

Please please please tell me thats right

[AndrewT]

I propose to carry out a test for you all to consider.

I will turn the test back to front.
I normally listen sensibly with an average voltage to the speaker of 100mVac to 1Vac.

If I turn these speakers up louder I can hear the sound becoming "loud" due to the increase in distortion. I can turn it louder yet, but it is not nice to listen to. The average level is now 5Vac to 6Vac. This is more than 12dB below the maximum amplifier output, i.e. the amp can push more than 22Vac into these 89dB/W speakers.

That's the background.
The test.
What power of amplifier do you suggest I use for the test tone? 1W or 2W or 5W or 10W or 20W? I will build the amplifier to suit the majority view.

Then I will run the music at the level I think is nice but not distorted and give you both the average level and the test tone level. Then I can turn it up and measure the new average and test tone levels.



From there we can see what conclusions we can come to.

[MarcelvdG]

I ran the test twice, once with a recording I made myself of the 2006 concert of Rumours mixed chorus for light music, and once with a quite dynamic commercial recording of Ravel's Bolero (ZYX classic CLS 4327, recorded in 1996).

The Rumours-recording has absolutely no dynamic range compression. The loudest peaks in the recording are just below clipping: these are the applause, a slamming door and a section of a second or so where all ladies scream.

I put the Rumours recording somewhat louder than I'm used to. I normally set the level such that I can just hear the announcer, this time such that the announcer is at a normal speach level.

Rumours:
5.06 V left, 6.36 V right (peak voltages 20.22 V and 25.46 V)

Bolero:
2.26 V left, 2.55 V right (peak voltages 9.05 V and 10.18 V)

Loudspeakers: QUAD ESL-63, having a rated sensitivity of 86 dB at 2.83 V extrapolated to 1 metre. Distance between loudspeakers and listener: around 2.5 metres. Living room size: roughly 8.5 metres by 4.5 metres, height 3.5 metres.

[JRKO]

That sounds like a good reference AndrewT, somewhat out of my league in practical terms but very generous

As nearly 75% of responses are 5v and less I would think that range would be of most interest. Can Pano access the actual numbers to get a true mean/average? If not I would go for 2V

[Mooly]

Originally Posted by 5th element

I think Pano's choice of sine wave signal level could perhaps be what's creating the confusion. For want of a better explanation, because I cannot read his mind, Pano chose to use a signal at -12dB or whatever because this wouldn't put stress on your system and would be unlikely to cause your loudspeakers to be destroyed. The less complicated way to have done this would be for him to have used a sine wave played at 0dB but that could have been destructive.

Exactly, and testing at 0db is method I used here, with the proviso of disconnecting the speakers for the voltage test.
So how much power do you really need for domestic listening ?

some of the same arguments surfaced there too. Rereading that thread I came across this,

So how much power do you really need for domestic listening ?

Which may go some way to explaining the "higher than 0db" peaks I observed earlier in this thread,

A Test. How much Voltage (power) do your speakers need?

That was something that really puzzled me as it contradicts my "cast in stone" saying about the 0db level being an absolute, however... this current thread is a perfectly valid and, I believe accurate (to all practical intents and purposes) method of arriving at the amplifier power needed to reproduce the maximum music level set by you on listening. The brief tests and checks I did with "Cool-edit" showed that all the CD's I tried (where I know where the loud bits are peaked at around -1 to 0db. So the musics dynamic range below that level doesn't matter. If the recording is or is not very dynamic... it doesn't matter. Nothing on any of my music tracks seemed to peak over Odb in the way the white noise on the Philips player did. That anomaly really does seem player specific... but all that is another story and not for here.

Originally Posted by AndrewT

I propose to carry out a test for you all to consider. I will build the amplifier to suit the majority view....................................

Then I will run the music at the level I think is nice but not distorted and give you both the average level and the test tone level. Then I can turn it up and measure the new average and test tone levels.

From there we can see what conclusions we can come to.

That's a great idea. I would think any normal amp (or chip amp ?) on a variable PSU would be adequate. If you have a variac it's even easier. Maybe just check with a scope to ensure it clips where expected.

You'd be amazed just how little you can get away with. This little PSU with it's 15-0-15 / 18va transformer (yes 18va) and 6800uF caps is one of my most useful test rigs... inherently SC proof up to a point... it will drive a stereo amp far louder than you might think. Great for initial design and development.