Amplifier Sensitivity, Decibels, and You!

240 Volts

Strange, yet oddly normal
Time for one of those long, boring semi-technical posts that no one here reads... :boring:


There have been various posts from people who have just acquired a "new" vintage amplifier, have connected it up to their speakers and fed it with a nice clean signal from a (relatively) modern CD player, and have commented along the lines of "I only turned the volume up to 10 o'clock and the whole house was shaking - boy that amp is powerful".
They never seem to consider that their amplifier might well be producing near to full output power even though the volume control is nowhere near maximum - a consequence of the sensitivity mis-match between vintage and modern equipment.

It seemed to me that it would be useful to go over a few basics regarding the decibel (dB) scale and how it relates to the sensitivity of the inputs on vintage amplifiers.


First a few basics about the dB scale (and a little bit of math - but nothing too difficult :) ) :

1. The decibel is a relative rather than an absolute measurement, i.e. it is used to measure the ratio of one signal to another. I am sure everyone is most familiar with it's use in representing the signal to noise ratio of equipment.

2. Positive dB values mean that a signal is greater than the reference value (ratio greater than 1), negative dB values that it is less than the reference (ratio less than 1).

3. If a signal passes through a number of amplification (+dB) or attenuation (-dB) stages, then the overall gain is found by simply adding up the dB values of each stage. For example, assume a signal passes through components with the following typical gain : phono amplifier +60dB, pre amp -20dB, power amp +30dB : The overall gain is therefore +70dB.

4. To convert the ratio between two voltages, V1 & V2, to decibels we use the formula : dB = 20 * log(V1 / V2)
A doubling of voltage = +6dB.
Conversely, to convert a dB value to the ratio between two voltages use the formula : Voltage Ratio = 10^(dB value / 20)
(The symbol "^" means "to the power of").

5. To convert the ratio between two powers, P1 & P2 to decibels we use the formula : dB = 10 * log(P1 / P2)
A doubling of power = +3dB.
Conversely, to convert a dB value to the ratio between two powers use the formula : Power Ratio = 10^(dB value / 10)


So back to the original question : Just how far do we have to turn up the volume to get maximum output power from our amplifier?

Look at your amplifier's manual and find the sensitivity value for the input you are using. The value it gives is the input voltage required in order to produce maximum rated power when the volume control is set to maximum. If the input signal you are feeding to the amplifier is greater than the sensitivity value, then maximum output power will be produced before full volume on the control.

Let's work through a real example to make things easier to follow (using data for my Pioneer SX-1250 receiver and Marantz CD-65SE CD player - typical of many vintage amp / modern CD player combinations) :

Input sensitivity of SX-1250 "Aux" (and "Tape") inputs : 150mV
Output voltage from CD-65SE at maximum signal level : 2000mV (2V)
So the CD player output signal is 20 * log(2000 / 150) = +22dB higher than that required for full output power from the SX-1250.

What this means is that the receiver's full power of 160 watts will be produced with the volume control set to -22dB (referenced to maximum = 0dB). Fortunately, in common with much high-end equipment, the SX-1250 has a volume control marked in -dB, making it easy to see where this occurs.

Perhaps the most "surprising" thing to come out of this calculation is just how little the volume needs to be turned up to get full output : -22dB on the SX-1250 corresponds to somewhere between the 11 and 12 o'clock positions.


So the next time someone tells you how loud their system is even at low "volume" settings (and implying that it would be ten times louder if turned up fully), just pause for a minute before you get too impressed and instead consider if they may have an amplifier/source sensitivity "mis-match". :scratch2:


Finally, we can apply the same calculations to the use of graphic equalizers and bass / treble / loudness controls. If +3dB of boost is applied at some frequency, then the power required at that frequency is doubled. If +10dB of boost is applied, then the power required increases by a factor of 10. So applying high levels of equalization (for example to compensate for the falling bass response of a speaker) massively increases both the power requirements of the amplifier and the handling capacity of the speakers.

- Richard B.
 
Great info and easy to follow. That should have occurred to me- but didn't.
Thanks very much.
 
mrbruce said:
Great info and easy to follow. That should have occurred to me- but didn't.
Thanks very much.

What also should have occurred to me is that volume level using the tuner input should be an accurate gauge of performance. Right?
 
Richard,
A very simple and reasonably easy explanation for explaining the issue. Thanks for taking the time. Guy's like you have taught me so much here at AK!
Thanks again!
Jim
 
mrbruce said:
What also should have occurred to me is that volume level using the tuner input should be an accurate gauge of performance. Right?
Yes, if you have a receiver then hopefully the manufacturer has got the output of the tuner matched to the sensitivity of the amplifier section. :yes: :yes:

- Richard B.
 
The general tendency of the output voltage of sources creeping up over time is interesting. I think its partially related to the greater stability of a higher voltage signal (in regards to noise pickup), but also the fact that since louder sounds better, at least in a momentary comparison, newer devices sound 'better' right after being plugged in, because of the slight boost in volume. Accident, or intentional? I don't know!

Thanks for the dB overview, it was well written. I used to give classes to new front of house engineers, and acoustic measurement not an easy thing to teach!

Peace,
Sam
 
Good thread, you've managed to answer some questions that have been bugging me for ages! I've got a dB-marked volume control, so this is good stuff to know - *and* I can put it to the test.

revsven
 
dB marked volume controls are the only way to go when you want to look like a serious audiophile. ;) I don't actually find them that useful, because of the crazy sensitivity issue (and the variable output levels of most sources), but they sure are cool to have. Reminds me of FOH mixing. ;) I do like explaining to people why 0 is louder than '22' though.

peace,
sam
 
Thanks 240 Volts. Have long wondered why my DVD/CD and tuner (9/10 o'clock)seemed to require less power for more output than my turntable (12 o'clock). Wasn't sure if it was a source signal thing or what (have to remember to turn the volume down when switching from turntable to CD/tuner or I blow out my eardrums). Thanks for explaining this enigma.

bjarmson
 
Does anyone think that the same increase in output levels has occured in FM broadcasts also? When listening to FM I can't get past 10-11 o'clock on the receivers I own without pegging the power meters. Since these would already have the tuner output level matched to itself, it seems pretty loud at low volume knob settings.
 
Great explanation, 240Volts.

Now for a follow-up, you should explain to the folks how 'loudness', as heard by your ears, is a function of speaker efficiency - as well as the gain of the amplification system. For example, your volume control may be at -22 dB for speakers "A" but needs to be at -10 dB for speakers "B" to achieve the same 'loudness' as judged by your ears.

A good second lesson, methinks. :yes:

Paul
 
Morden2004 said:
Great explanation, 240Volts.

Now for a follow-up, you should explain to the folks how 'loudness', as heard by your ears, is a function of speaker efficiency - as well as the gain of the amplification system. For example, your volume control may be at -22 dB for speakers "A" but needs to be at -10 dB for speakers "B" to achieve the same 'loudness' as judged by your ears.

A good second lesson, methinks. :yes:

Paul

Very good point. Another often misunderstood variable is that the 'audio taper' (logrithmic response) of many variable volume controls is niether accurate or consistant through it's range. A truly made 'step attenuator' multi-position volume switch can be made that is accurate in fixed DB steps but of course they are pricey and take up a lot of space.


Lefty
 
technut said:
Does anyone think that the same increase in output levels has occured in FM broadcasts also? When listening to FM I can't get past 10-11 o'clock on the receivers I own without pegging the power meters. Since these would already have the tuner output level matched to itself, it seems pretty loud at low volume knob settings.
AFAIK the maximum audio signal levels for FM broadcasts are very strictly controlled (at least here in the UK, but I guess the same applies in other countries as well). This is because the audio signal modulates (increases and decreases) the carrier frequency - if it changed too much then it could interfere with broadcasts on nearby frequencies.

What has happened with some radio stations in the UK is that the average level has increased even though the maximum level is unchanged. Subjectively the broadcast sounds louder. This is achieved by compressing the audio. It may sound better on portable and car radios, but on a decent hi-fi system it sounds terrible. :thumbsdn:

- Richard B.
 
Well that confirms my suspicion that FM used to sound better than it does now, unless my hearing is just going to hell... :scratch2:
 
"What has happened with some radio stations in the UK is that the average level has increased even though the maximum level is unchanged. Subjectively the broadcast sounds louder. This is achieved by compressing the audio. It may sound better on portable and car radios, but on a decent hi-fi system it sounds terrible. "
...
yup...and i dread to think how much worse it will be when diabolical audio broadcasting ( DAB ) takes over completely in the uk :thumbsdn:
..
..
.
great explanation btw...a thoroughly enjoyable demystification of the volume control for the technically challenged like myself... :thmbsp:
 
Good information, thanks. But I do have a question. If the receiver is maxed out at lets say ten o'clock, then what would happen if you put it up to twelve 'oclock?
 
I still have a few questions. The imput of the tape & auxilary are 150mv sensitivity and the main amp in is 1000mv sensitive. you say the pre amp = +40db and the main amp is another +30 db. if i put my volume in the 12 o clock position (-20db) it has the same input and output voltage. So the pre amp can only give +20db more than the input signal. if you give a +20db boost you have a signal of 15 volt? (then i don't need a main-amp) If i put my source to the main amp in there should be no troubles with a lower volume, and what about the 600mv threshold of the diodes ?
 
Last edited:
Back
Top Bottom