23-02-2017, 02:01 PM
(23-02-2017, 05:55 AM)Diabolical Artificer Wrote: What I don't get Mark in your post above is how the volume should be set. To get full power ish your amplifier would need to be at 3/4 volume say, so that would work if your listening to a record with a large dynamic range, a bit of classical say where you have quiet passages then a crescendo at the end. For rock music which is full on from start to finish though it would be uncomfortable. A better set up would be with a separate pre-amp and power amp. Preamp set low, power amp high. Hope you get the gist.
Gain and level are two concepts that ought to be trivially simple, but are not. Even in the relatively "sane" world of professional audio, these are still massive areas of confusion. Things are much worse in hifi...
First, there is no correlation between the physical position of the volume control and the output level of the amplifier. Just because the volume control has been turned to half way, it does not mean the amplifier is at half-power, half-loudness or half-anything. Manufacturers play tricks here to give you the impression that masses of power is available to you - in reality, many amplifiers clip with the volume set to maybe only a third of the way around.
So let's break it down.
- You have input level. CD players should produce 2V RMS when replaying a sine wave peaking at 0dBFS.
- Then you have amplifier gain.
- Then you have the clipping point.
Let's pick 100 watts as the amplifier power. This requires a voltage swing at the output of 28.3V RMS, assuming 8 ohm loudspeakers.
If the CD player outputs 2V RMS, then surely the required gain is only x14, right?
Well, this makes far too many assumptions about the source equipment and the recordings. Firstly, 2V RMS represents the highest level you'll get from a recording. Actually, it's more accurate to say 2.8V peak, but even that is debatable. Putting that aside, however, let's take the 2 or 2.8V as the peak level. So as described earlier, a decent recording might have an average level that is 20dB less than that - call it 200mV or thereabouts.
But actual values of PMR (peak to mean ratio) are hard to find, and vary from recording to recording. So there will be times when a piece is quieter so you'll need to turn the volume up a little, and there will be noisier recordings that you'll want to turn down.
So is a total gain of x14 the right answer? No, because recordings vary, it's nice to have a bit of spare gain. In the pro world, we call that "gain in hand".
But what about the source? CD players might well be 2V nominally (they do vary a bit), but other sources are quieter. A hi-fi tuner might only give 400mV for a fully modulated broadcast. In fact, that's not a bad thing - nearly all FM broadcasts are very heavily compressed, so have a much smaller PMR, meaning that if you match the peak levels, most CDs will sound quieter. I said it was complicated!
All that considers the input applied to an amplifier, and as you can see, it's a bit of a mess. To make an amplifier that works well with a range of equipment and recordings, you must provide more gain that is strictly necessary, otherwise "quiet" sources and recordings won't reach full output power. But the flip-side of that is simple: it is very easy to drive an amplifier into clipping. We do it all the time.
If we assume that the lowest signal level will be 400mV, then to reach 28.3V at the output, we need an overall gain of x70. Rather more than x14. x70 is 37dB - not difficult by any means.
But in order to impress the customer, the designer will usually build in some more. Values of 40dB or more are easy to find. It makes for a very "touchy" volume control when using CDs or similar.
This is like throttle response. My wife's 1.4l Fiesta feels a lot quicker than my 1.8l Focus, but that's totally illusory because virtually all of the engine output is controlled in the first half-inch of the peddle. If you're accelerating, and decide to put the foot down the remaining couple of inches, nothing changes! It's the opposite for the Focus, which is much more linear and progressive. It makes for a much more relaxing, smoother and economic drive, but the power is there when needed...
All of this explains why "clipping" LEDs would be a good idea, but because we are so used to mild clipping, manufacturers don't fit them because they don't want to be constantly explaining the above to confused non-technical consumers.
Regarding pre and power amps, in the hi-fi world, power amps don't have gain controls, so there's no material difference between pre/power and integrated.
If you turn down the gain of a power amp, there is a risk you could drive the preamp into clipping! If you do the opposite, then you're amplifying pre-amp noise needlessly.
All of the above applies at all stages of an amplfier. Amplifiers are simple - think about a sound desk! Take it from me - deciding what the "nominal" levels should be at each stage of a complex chain - so that you don't clip, but don't get needlessly close to the noise floor - is what keeps designers awake at night. On Tuesday night I was agonising about this for a new design which I can't really expand on here. Suffice to say, people sometimes get it wrong.
Overall, when thinking about level at different points of an audio chain, we say "gain structure". For a hi-fi amp, it's simple. There's only 1 volume control in most cases. Perhaps tone and balance controls - do you put those before the volume control (where clipping might take place), or after the volume control (where their noise would be present at all settings of the volume control, and be potentially quite annoying in a domestic setting).
If you think all that is a mess, it's worth saying that professionals can't agree on how to measure audio level either. There are many competing standards, and all exist for good reason, but all have shortcomings. Even the very newest ones like R128 and BS 1770. Broadcasters receive far more complaints about sound quality than picture quality. It's difficult...
