On the relationship between preamps dBFS scale, pre-out V RMS and balanced vs unbalanced

[flyingflatfour]

Could someone please explain to me if it is possible for any preamp to calculate for any given position on the dBFS scale a pre-out output based on a known dBFS/preout value?

I.E. a given preamp measures 1.4 V RMS on it’s preouts when the volume is set at -0 dBFS. From there, can we know where to set the volume if we want a 1 V RMS output, or the other way around?

Now if a preamp sends for instance 1 V RMS out from an unbalanced preout to a balanced amplifier input which specifies a 1 V RMS input sensitivity, what will the amp “see”? The convention is to use a conventional two-core + screen balanced cable and to terminate the cold/return signal with the screen on the RCA side.

Actually, I am not even sure of my usage of the term dBFS to describe a volume scale?

Thank your for your help.

 

[PaulD]

A few things...
dB is a relative measurement, an expression of one value against another. Decibels are a logarithmic measurement, useful in audio because our ears work logarithmically. You can express any change in dB. They are useful for expressing audio system changes in a way that is related to our perception.

dBFS (dB Full Scale) is usually used in digital systems, 0dB being the maximum output or signal level of the system (all bits on).

For your example, you want 1.4V reduced to 1V, that is a reduction of 2.92dB, or -2.92dB change. [20 x Log10 (new value / orig value)]. We would usually round that to -3dB for convenience, unless the precision is useful to you (e.g. for precision measurements). Balanced systems typically have 6dB more signal when both signal lines are driven anti-phase (some balanced systems are impedance balanced with only one conductor driven).

dB(SPL) is used for sound pressure level (SPL) measurements, the reference level is 20uPa which is the threshold of hearing at 1kHz. There is also dBV, dBU, and so on, which have their own definitions and reference levels.

Here are some links that may be useful:
https://animations.physics.unsw.edu.au/jw/dB.htm
http://www.sengpielaudio.com/calculator-db-volt.htm
https://en.wikipedia.org/wiki/Decibel
https://www.edn.com/audio-levels-dbu-dbv-and-the-gang-what-you-need-to-know/
https://www.music-production-guide.com/decibel-scale.html

 

[Pluto]

Your description is none too clear I'm afraid.

When you say

a given preamp measures 1.4 V RMS on it’s preouts when the volume is set at -0 dBFS

By 0dBFS do you mean, the volume knob fully up? As has already been pointed out, the term 0dBFS usually refers to a digital signal, the absolute maximum audio level that can be accommodated. It's worth appreciating that translating knob rotation into an accurate number of dB is a thankless and largely pointless task. People pay $00s for pots (i.e. knobs) that do this with reasonable and predictable accuracy.

Reading your original posting, I cannot help but feel that you are treating some notional specification a bit too seriously. The difference between 1.4V and 1V is about 3dB, a number barely worth thinking, let alone getting anxious, about.

 

[flyingflatfour]

A few things...
dB is a relative measurement, an expression of one value against another. Decibels are a logarithmic measurement, useful in audio because our ears work logarithmically. You can express any change in dB. They are useful for expressing audio system changes in a way that is related to our perception.

dBFS (dB Full Scale) is usually used in digital systems, 0dB being the maximum output or signal level of the system (all bits on).

For your example, you want 1.4V reduced to 1V, that is a reduction of 2.92dB, or -2.92dB change. [20 x Log10 (new value / orig value)]. We would usually round that to -3dB for convenience, unless the precision is useful to you (e.g. for precision measurements). Balanced systems typically have 6dB more signal when both signal lines are driven anti-phase (some balanced systems are impedance balanced with only one conductor driven).

dB(SPL) is used for sound pressure level (SPL) measurements, the reference level is 20uPa which is the threshold of hearing at 1kHz. There is also dBV, dBU, and so on, which have their own definitions and reference levels.

Here are some links that may be useful:
https://animations.physics.unsw.edu.au/jw/dB.htm
http://www.sengpielaudio.com/calculator-db-volt.htm
https://en.wikipedia.org/wiki/Decibel
https://www.edn.com/audio-levels-dbu-dbv-and-the-gang-what-you-need-to-know/
https://www.music-production-guide.com/decibel-scale.html

Hi Paul,

Thank you very much for your reply!

Very good explanations and a lot to read in the links you provided.

If I understand correctly, the volume setting on a preamp is measuring a difference of power relative to a reference? Hence the 20 (2 x 10) log (V/V’) formula because P=V^2/R? Should one use a suffix or just dB? On an AVR very often the autocalibration will trim speaker levels so as to reach the 85 dB(C) reference level with the volume at 0 dB. Indeed I realise that using dBFS is wrong in this context!

Still, I do not understand yet what a balanced amp would « see » from an unbalanced source sending a 1 V RMS signal?

 

[flyingflatfour]

Your description is none too clear I'm afraid.

When you say

By 0dBFS do you mean, the volume knob fully up? As has already been pointed out, the term 0dBFS usually refers to a digital signal, the absolute maximum audio level that can be accommodated. It's worth appreciating that translating knob rotation into an accurate number of dB is a thankless and largely pointless task. People pay $00s for pots (i.e. knobs) that do this with reasonable and predictable accuracy.

Reading your original posting, I cannot help but feel that you are treating some notional specification a bit too seriously. The difference between 1.4V and 1V is about 3dB, a number barely worth thinking, let alone getting anxious, about.

Hi Pluto,

I realise I haven’t made this obvious: I absolutely don’t pretend to understand much of what we are discussing here. My intention is just to understand, for the sake of understanding, so it doesn’t matter if it’s really important in the end .

To answer your question, yes I meant the volume knob reading 0 dB on an AVR for instance. It’s not necessarily fully up then, since on my Yamaha for instance you can go to +15 dB. I realise that dBFS was plain wrong and misleading.

 

[flyingflatfour]

We are not talking power (current x voltage), we are only talking signal voltage level as the input impedance is high and the output impedance is low. Very little current is flowing through this connection. Thus the 20 x Log10(new-value / orig-value) formula. See chapter 1 of Doug Self's book Audio Engineering for a complete explanation

Thanks for clarifying. Could you by any chance point me to an online explanation of where the doubling factor comes from in this case? If I understand logs correctly it derives from an expression where V was to the power of 2?

I assume you mean dB(SPL) C-weighted

You assumed right. Is there an abbreviated convention for this?

So for your example, the balanced amplifier receiving an unbalanced 1V signal "sees" 1V... Not sure how it could be different.

Right, so this simply means that as long as a preout can match an amp's specified input sensitivity, then it doesn't matter if the preout is unbalanced and the amp is balanced, there is no loss of -6 dB somewhere?

On the other hand a balanced preout to an unbalanced amp input would effectively have the amp only see 1/2 or -6 dB of the preout voltage?

What bugs me then is this input sensitivity figure. Some amps have both a balanced and unbalanced input but only one input sensitivity. Shouldn't it be double for the balanced input?

 

[Blumlein 88]

Thanks for clarifying. Could you by any chance point me to an online explanation of where the doubling factor comes from in this case? If I understand logs correctly it derives from an expression where V was to the power of 2?



You assumed right. Is there an abbreviated convention for this?



Right, so this simply means that as long as a preout can match an amp's specified input sensitivity, then it doesn't matter if the preout is unbalanced and the amp is balanced, there is no loss of -6 dB somewhere?

On the other hand a balanced preout to an unbalanced amp input would effectively have the amp only see 1/2 or -6 dB of the preout voltage?

What bugs me then is this input sensitivity figure. Some amps have both a balanced and unbalanced input but only one input sensitivity. Shouldn't it be double for the balanced input?

Let us think about the connection. Unbalanced will be ground and the signal. If you have 3 volts it is ground (0 volts) and the signal connection is 3 volts above that.

Now in a balanced connection you will have ground, +signal and -signal. If you have 3 volts it would be ground (0 volts), +1.5 volts on +signal, and -1.5 volts on -signal. For a total difference of 3 volts. Yes I'm playing a little loose with the connections for those who know, but for this purpose let it slide.

So imagine I have an unbalanced source feeding a balanced amp. Ground will connect to ground. Signal will connect to +signal, and -signal will be connected to ground. The balanced input will see 0 volts for ground. +1.5 volts for +signal and 0 volts for -signal. A difference between the plus and minus signal of only 1.5 volts. Were it a balanced signal being fed to the amp it would see a 3 volt difference.

Does that sort of answer your question?

 

[solderdude]

What bugs me then is this input sensitivity figure. Some amps have both a balanced and unbalanced input but only one input sensitivity. Shouldn't it be double for the balanced input?

That can be input circuit dependent.
RCA inputs usually are used directly.
XLR inputs can either be used directly (balanced all the way inside) or could be made SE by using an input transformer (galvanically separated) or by use of a summing circuit. In both cases the resultant signal can be made any voltage the designer uses.

This means that a 2V S.E. signal can be 2V inside and a balanced signal also can be 2V inside.
In that case regardless of the used method (S.E. / balanced) and the actual voltages used in the transmission cable the incoming signal will have the same level regardless of the used cables and can even have the same overload characteristics (headroom)

 

[Atanasi]

If I understand correctly, the volume setting on a preamp is measuring a difference of power relative to a reference? Hence the 20 (2 x 10) log (V/V’) formula because P=V^2/R?

Sound levels are typically measured with respect to the power of the sound wave. The power of the sound wave is related to the square of the amplitude, just like the power of an electrical wave is related to the square of the voltage. The electrical voltage and the amplitude of the sound wave are analogous but distinct.
By definition, 10 dB increase means increasing the respective amount by a factor of ten. In the case of sound, it means multiplying the power by ten. Correspondingly, the amplitude is multiplied by the square root of ten, or on the dB scale, increased by 5 dB.

Electrical signals represent amplitude through voltage. So, 20 dB increase in power is a tenfold increase in voltage. That's why audio applications typically use this factor of 20 in dB formulas when computing from voltage relations, even though the definition of decibels uses a factor of 10.

 

[flyingflatfour]

Thanks to all of you.

When I was talking about about power in the preamp I made things confusing but it was just saying that theoretically as soon as you have a voltage and a resistance that is not infinite, you have current (I = U/R) and ultimately power (P=V^2/R) even if it’s a very small amount of power. From that perspective I believe it makes sense to say that the volume level in dB of the preamp is a measure of relative power (a function of 10 log P/P’) or of 2x relative voltage (a function of 20 log (V/V’) ? I know it has no practical purpose, it’s just for the sake of understanding.

Please forgive me as my math and physics are very rusty . I realise this is very basic for some of you, but I believe it’s absolutely essential to get the basics right from the beginning if one wants to make any real progress in learning anything.

About balanced vs unbalanced, I understand that the doubling of V comes from the extra conductor carrying the same signal than the + conductor but inverted so their voltages add up. So far so good.

Now let’s imagine a balanced amp with a 1 V input sensitivity. I understand this means if it is fed 1 V from it’s input it will reach it’s maximum rated power. A balanced preamp will have to produce a +0.5V and a mirrored -0.5V signal in order for the amp to see a total of 1V. On the other hand an unbalanced preamp will have to produce a single +1V signal to achieve the same result. Is this correct?

If it is, the next question that follows is how come it is even possible to connect an unbalanced preamp to a balanced amp? I thought the idea behind balanced circuitry was to compare the + and - signals to eliminate pickup noise/interference. But if there is no - signal anymore how can it still work at all?

 

[Atanasi]

If it is, the next question that follows is how come it is even possible to connect an unbalanced preamp to a balanced amp? I thought the idea behind balanced circuitry was to compare the + and - signals to eliminate pickup noise/interference. But if there is no - signal anymore how can it still work at all?

There are two types of balanced outputs: Some are floating in the sense that one wire can be shorted to ground and the other wire would carry the right signal, though maybe with a reduced amplitude. Some are strict in that shorting a wire to ground is not allowed. Typically, audio interfaces for content production have floating outputs, in order to support interfacing with unbalanced devices, while headphone amps with balanced outputs are strict.

Balanced inputs, on the other hand, are based on the difference between the wires, and they should work alright even if one wire is shorted to ground.

 

[flyingflatfour]

There are two types of balanced outputs: Some are floating in the sense that one wire can be shorted to ground and the other wire would carry the right signal, though maybe with a reduced amplitude. Some are strict in that shorting a wire to ground is not allowed. Typically, audio interfaces for content production have floating outputs, in order to support interfacing with unbalanced devices, while headphone amps with balanced outputs are strict.

Balanced inputs, on the other hand, are based on the difference between the wires, and they should work alright even if one wire is shorted to ground.

Thank you. I will read some more about this.

 

[Trdat]

This thread is not too old so hopefully a few chime in. Ill be honest, I still don't understand anything so perhaps a some basics to get going.

My biggest confusion is the DB on the left hand side of the intermodulation distortion graph. Is the DB literally SPL DB? Let's take the Marantz 7705 measurements as an example. Amir mentions to keep the volume under 82 and you should be alright. Is the 82 something I look for on the Marantz? Or do I measure the volume with an SPL meter? How do I figure out what 82 is?

Then if we take Blumlien 88 measurements on the 7701 he mentions with 0dbfs input and the volume at 0(Now I am presuming this is the knob at 0db?) its about 3 volts. Okay what input? Are we talking about the digital input from my computer into the Marantz? Then he mentions if we send lower level digital signal and increase the volume up to 4.5 volts your okay. So what do we mean by digital signal? Again is this the volume control from the computer?

Basically, I would like to know how I find the volts on my Marantz to know at which point distortion is kicking in? Is there a simple way to do this or can we only guess?

And lastly is there a way for Amir to just tell us on that device at what volume on the knob is the maximum volts that distortion kicks in so we can just simply know and tinker with it ourselves? I am guessing he might have done this but I have not understood it properly...