I think this test can be an indicator of the power produced by the amp.
Concerning the conclusions, I use two amps 50ASX2 (2 x 50W @ 4 Ohms; 2 x 25W @ 8 Ohms) and a Poppulse T150 (77 W atts RMS per channel into 4 ohm) .The first one has just enough power to meet my loudest listening level, the second is almost OK 🙂
Member
Joined 2009
Paid Member
What you have to remember is whether you are working with peak values or rms values. If you are measuring a peak voltage on a sine wave you need to divide it by sqrt(2) to get rms. Since the formula uses voltage-squared it's equivalent to dividing by 2.
Power (rms) = (peak voltage x peak voltage) / (2 x load resistance)
I work with pro sound engineers every day, and sometime am one myself.
This test is NOT for that purpose, was never designed or intended for that purpose. It's a whole other kettle of fish. Trust is not the issue. Those who do not understand that, do not understand the test.
Use the test for its intended purpose. Finding out how much voltage is going to the speakers in your home system. Is your amp close to clipping, or not? That's all there is to it.
I'll move it now. It's a great subject.
Moved here:
http://www.diyaudio.com/forums/equipment-tools/259833-power-soaks-dummy-loads.html
Moved here:
http://www.diyaudio.com/forums/equipment-tools/259833-power-soaks-dummy-loads.html
That sounds about right. 6.5 Watts RMS for that volume setting. I prefer to measure (or calculate) the voltage needed, but most people think in terms of power.
Most amps will give you plenty of headroom at the volume setting. Is your amp capable of 10-15 RMS watts clean?
Thats RMS= -3dB FS (Full Scale). If full scale is considered 0dB, then the highest possible level for an unclipped sine wave is -3dB RMS. Obviously the peaks of the wave will be at 0dB FS, or digital maximum.
The tones I supplied have an RMS value 12dB below Full Scale. And since a pure sine will never by higher than -3dB RMS, we can say that the test tones are 9dB below the maximum RMS permitted for a sine wave.
You have to remember that both types of level are valid and important. Peak value and RMS value.
Pictures may help
Below you will see the same 120Hz sine wave at its maximum value with peaks at 0dB
And also you see it at the level supplied at the top of this thread, with peaks at -9dB
Don't forget! The sine wave's RMS value is 3dB below peak. Thus the full scale waveform has an RMS value of -3dB FS and the second one an RMS value of -12dB FS.
Got it?
Since 12dB is approximately 4X, you can take the voltage you measure from the test tone and multiply it by 4. That will be the peak voltage your amp supplies at that volume setting. The maximum RMS voltage will be 9dB higher, or 2.82 x higher.
Below you will see the same 120Hz sine wave at its maximum value with peaks at 0dB
And also you see it at the level supplied at the top of this thread, with peaks at -9dB
Don't forget! The sine wave's RMS value is 3dB below peak. Thus the full scale waveform has an RMS value of -3dB FS and the second one an RMS value of -12dB FS.
Got it?
Since 12dB is approximately 4X, you can take the voltage you measure from the test tone and multiply it by 4. That will be the peak voltage your amp supplies at that volume setting. The maximum RMS voltage will be 9dB higher, or 2.82 x higher.
Attachments
Understood, but you missed my point that to ensure there's no clipping one needs to properly calculate what's required. Some [many?] amps today can be clipping from the get-go, yet do it rather benignly, ditto tubes, so how can they know?
Then there's the issue of wire sizing, quality of connections. After all, it's all about voltage drop over distance as you know and the amp can be pumping out more voltage than what's being measured at the speaker, at least in a fairly large room like mine where I need 20 ft [40 ft round trip] speaker runs.
Then there's losses through XOs and especially BSC, a voltage dividing network, so how is this being accounted for?
GM
Do you mean in a pro P.A. rig, or a home rig?
All the pro rigs I've used since 1997 have been self powered, so it's really a moot point for me. Guess I could open up the boxes and measure the voltages going to the drivers at typical SPLs, but have never bothered. If we ran amp racks, I might.
In a home system, it's usually less complicated. But I'm not clear on your question about BSC and XO and cable.
All the pro rigs I've used since 1997 have been self powered, so it's really a moot point for me. Guess I could open up the boxes and measure the voltages going to the drivers at typical SPLs, but have never bothered. If we ran amp racks, I might.
In a home system, it's usually less complicated. But I'm not clear on your question about BSC and XO and cable.
My loudspeakers are 6 Ohms nominal ,an my valve amp also has 6 Ohms nominal impedance matching, So if i do the voltage measurement, the results I get would not apply?
They would still apply as voltage, because voltage is what you are measuring.
If you need to figure out power, then yes, take the 6 ohms into account.
If you need to figure out power, then yes, take the 6 ohms into account.
The -12dB input reference in REW is exactly why it is suggested for the same reasons Pano and GM give for the finer points of a sine wave measurement. This is actually considered minimum. Setting the reference lower for more transient type waveforms is required. Biggest disadvantage in doing so is reduced S/N ratio.
Sometimes I think it's relevant to tie these details together for greater conceptual understanding. 🙂
Sometimes I think it's relevant to tie these details together for greater conceptual understanding. 🙂
Maybe I can, a little.
I chose -12dB as "hot enough, but not too hot" for measuring. As an average level it's a bit higher than most music, but no so high as to cause trouble. As Greebster said, a good level for S/N when measuring.
Dynamically mastered CDs will have an average level of -22dB or more below peak. The peaks will hit 0dB, but the average will be 22dB or more below those peaks. That can make a good reference level, but S/N may suffer in your measurements by using a signal that low.
-12dB RMS is a good compromise for running measurements. Loud enough to stay clean, but no so loud as to cause problems.
I chose -12dB as "hot enough, but not too hot" for measuring. As an average level it's a bit higher than most music, but no so high as to cause trouble. As Greebster said, a good level for S/N when measuring.
Dynamically mastered CDs will have an average level of -22dB or more below peak. The peaks will hit 0dB, but the average will be 22dB or more below those peaks. That can make a good reference level, but S/N may suffer in your measurements by using a signal that low.
-12dB RMS is a good compromise for running measurements. Loud enough to stay clean, but no so loud as to cause problems.