Gain "staging" or gain "structure"

[Gys Florian]

Hi,

I would like to talk about "Gain staging", or "Gain structure" as I'm not sure to understand it fully :
I would like to understand if there is room for improvement playing with attenuator nobs on an amplifier versus going the "plug, play, forget" way of life ;-)

Please note: I'm using this article as my main source for maths.

Setup
Préamplifier Yamaha CX-A5100:
Max Output Level : 4V unbalanced, 8V Balanced at 0.06%THD

Amplifier Parasound Halo A23:
Continuous Power Output : 125W RMS x 2, 20Hz-20KHz, 8Ω, 0.06%THD both channels driven
Input Sensivity : Balanced 1V for 28.28V output into 8Ω, Unbalanced 1V per leg for 28.28V output into 8Ω
Gain : 29dB with attenuator nobs on "THX" position (attenuators defeated)

First, it seems Parasound A23 sensivity is not given for full Power, as 28.28V output into 8Ω translates to 99.97W (Parasound A23 is 125W RMS into 8Ω)
If A23 is 125W into 8Ω, then max amplifier voltage output is 31.62V. From there we can conclude with a gain of 29dB that we need to send 1.12V from preamp to drive A23 to full power.
And I guess for Unbalanced it does translates to 2.24V, given unbalanced sensivity is given "per leg", which correspond to Gene's comment : "THX standard gain level is 29dB; utilizing balanced inputs decreases this to 23dB, though naturally the output of the preamp is boosted by 6dB under this scenario (i.e. voltage output of the preamp is doubled)"

So, as a conclusion : My preamp needs to send 2.24V to drive my amp at full power using balanced inputs which will be easy considering it is able to output as much as 8V.

Now my question... :
If, on the amplifier, I'm starting using attenuator nobs : for simplicity let's say I will set them to -6dB.
Then... this means my preamp will now need to send 4.48V to drive the amp to full power, right?
In that case: is it better than not using attenuators? Because, isn't sending the signal from preamp as hot as we can (without any distortion of course) the perfect way to increase Signal to Noise ratio?

 

[TLS Guy]

Hi,

I would like to talk about "Gain staging", or "Gain structure" as I'm not sure to understand it fully :
I would like to understand if there is room for improvement playing with attenuator nobs on an amplifier versus going the "plug, play, forget" way of life ;-)

Please note: I'm using this article as my main source for maths.

Setup
Préamplifier Yamaha CX-A5100:
Max Output Level : 4V unbalanced, 8V Balanced at 0.06%THD

Amplifier Parasound Halo A23:
Continuous Power Output : 125W RMS x 2, 20Hz-20KHz, 8Ω, 0.06%THD both channels driven
Input Sensivity : Balanced 1V for 28.28V output into 8Ω, Unbalanced 1V per leg for 28.28V output into 8Ω
Gain : 29dB with attenuator nobs on "THX" position (attenuators defeated)

First, it seems Parasound A23 sensivity is not given for full Power, as 28.28V output into 8Ω translates to 99.97W (Parasound A23 is 125W RMS into 8Ω)
If A23 is 125W into 8Ω, then max amplifier voltage output is 31.62V. From there we can conclude with a gain of 29dB that we need to send 1.12V from preamp to drive A23 to full power.
And I guess for Unbalanced it does translates to 2.24V, given unbalanced sensivity is given "per leg", which correspond to Gene's comment : "THX standard gain level is 29dB; utilizing balanced inputs decreases this to 23dB, though naturally the output of the preamp is boosted by 6dB under this scenario (i.e. voltage output of the preamp is doubled)"

So, as a conclusion : My preamp needs to send 2.24V to drive my amp at full power using balanced inputs which will be easy considering it is able to output as much as 8V.

Now my question... :
If, on the amplifier, I'm starting using attenuator nobs : for simplicity let's say I will set them to -6dB.
Then... this means my preamp will now need to send 4.48V to drive the amp to full power, right?
In that case: is it better than not using attenuators? Because, isn't sending the signal from preamp as hot as we can (without any distortion of course) the perfect way to increase Signal to Noise ratio?

There is no simple answer to your question in a way and in a way there is.

The first thing to remember is that music is, or should be, highly dynamic, with wide dynamic range. I know in the pop industry they don't know or understand this, but I don't cater or engineer for those idiots.

So in well written, orchestrated and recorded music there is going to be a vast range of voltage at inputs and outputs throughout playback.

So what is this all about?

Simply it is optimizing signal to noise ratio and avoiding clipping of any input output during playback of any program. Sounds simple, but not so easy to do. Without instruments, which incidentally few invest in, but should, you can only do this by trial and error.

So if you hear no hiss or noise at your listening position with volume or trim pots set in a sensible range you are OK, as long as you are not getting distortion in the loud passages.

Now this is where the difficult part starts. Unless you have a circuit then you don't know how the trim input pots are designed.

That may be pots on the input. This will prevent clipping, but degrade S/N. If they are after the first stage, then they can optimize S/N but not prevent overload clipping of the input stage.

If on the other hand the trim is in the feedback loop of an input IC opamp, then the trim can not only prevent clipping but also improve S/N at the same time. In my experience that although the latter has been known and implemented for nearly 50 years now, the cheap easy fix is what you frequently get even on high priced gear.

 

[WaynePflughaupt]

Responding to this specifically:

Now my question... :
If, on the amplifier, I'm starting using attenuator nobs : for simplicity let's say I will set them to -6dB.
Then... this means my preamp will now need to send 4.48V to drive the amp to full power, right?
In that case: is it better than not using attenuators? Because, isn't sending the signal from preamp as hot as we can (without any distortion of course) the perfect way to increase Signal to Noise ratio?

Really, you’re fretting over nothing. I’m confident that there is nothing you can do to such stellar components, gain-setting wise, to achieve any audible noise. Just reduce the amplifier’s gain to that you get a reasonable “travel” from the Yamaha’s volume control (i.e. with the amp’s gains all the way open you’d have a “hair-trigger” from the Yamaha – not what you want). As TLS said, as long as you don’t get any audible noise or distortion from loud passages, you’re fine.

Regards,
Wayne A. Pflughaupt

 

[Irvrobinson]

The first thing to remember is that music is, or should be, highly dynamic, with wide dynamic range. I know in the pop industry they don't know or understand this, but I don't cater or engineer for those idiots.

Come now, Mark, the pop music industry understands their target markets, and the people in those markets mostly listen in vehicles and using earbuds while doing something else (exercising, flying, etc). Wide dynamic range is actually annoying in those environments. Anyway, you don't strike me as being in the target markets; somehow I can't envision you being into Cardi B...

 

[Gys Florian]

In that case I believe it is better to leave them wide open and move on (THX standard gain).

I always thought it was better to optimize level between preamp and amp, so at the end you obtain something like this :
With a source signal at 0dB level(max recordable level on a CD), and with my Yamaha preamp volume set to 0dB (no attenuation, no amplification), then Yamaha outputs should be at max level without any distortion (which is 8VRMS), and Parasound amplifier should be at max level without any distortion (which is 31.62V into 8Ω).

Thank you for answering.

 

[Gys Florian]

Come now, Mark, the pop music industry understands their target markets, and the people in those markets mostly listen in vehicles and using earbuds while doing something else (exercising, flying, etc). Wide dynamic range is actually annoying in those environments. Anyway, you don't strike me as being in the target markets; somehow I can't envision you being into Cardi B...

That is right, but... wouldn't be better to make good records with full dynamic range, and then to compress it on the fly on the player when needed?
For example, In my Yamaha I can set it to "max dynamic" which is my prefered setting, but I can also set it to "Minimum", and then all movies on BluRay source sounds as compressed as bad music records!

I believe it is very easy to compress on the fly on any player, but with bad records you will never be able to recover the loss of dynamic.

 

[Irvrobinson]

That is right, but... wouldn't be better to make good records with full dynamic range, and then to compress it on the fly on the player when needed?
For example, In my Yamaha I can set it to "max dynamic" which is my prefered setting, but I can also set it to "Minimum", and then all movies on BluRay source sounds as compressed as bad music records!

I believe it is very easy to compress on the fly on any player, but with bad records you will never be able to recover the loss of dynamic.

Are you actually thinking that every vehicle and every mobile device, most of which are legacy by nature, will know when to compress or not and by how much? Are you also thinking that studio compression is a simple on or off process? It isn't.

 

[Gys Florian]

Responding to this specifically:

Just reduce the amplifier’s gain to that you get a reasonable “travel” from the Yamaha’s volume control (i.e. with the amp’s gains all the way open you’d have a “hair-trigger” from the Yamaha – not what you want).

I disagree on this part (already tried with previous power amp, a Yamaha P2500S) :
first test : attenuators wide opened (0dB position), then Yamaha CX-A5100 calibration
Second test : attenuators on -6dB position (wich was reducing idle noise in tweaters), then Yamaha CX-A5100 calibration

In both case, you end up because of calibration to exactly same output volume level for each Yamaha volume steps

 

[Gys Florian]

Are you actually thinking that every vehicle and every mobile device, most of which are legacy by nature, will know when to compress or not and by how much? Are you also thinking that studio compression is a simple on or off process? It isn't.

I don't know how they make it in studio, and I don't know if it requires lot of work.
But what I do know, is that uncompressed music on a bluray commpressed on the fly with my Yamaha sounds to me exactly like a compressed CD, certainly never worse.

 

[TLS Guy]

In that case I believe it is better to leave them wide open and move on (THX standard gain).

I always thought it was better to optimize level between preamp and amp, so at the end you obtain something like this :
With a source signal at 0dB level(max recordable level on a CD), and with my Yamaha preamp volume set to 0dB (no attenuation, no amplification), then Yamaha outputs should be at max level without any distortion (which is 8VRMS), and Parasound amplifier should be at max level without any distortion (which is 31.62V into 8Ω).

Thank you for answering.

I think you will be short on headroom. I would do this play a CD or BD you know well and set your Yamaha at -10 db, and then set your power amp volumes to where the sound is plenty loud and then leave everything alone on the power amp, and just use the Yamaha volume control. I would do that for starters.