[Avro_Arrow]

I have two amps built the same except for the gain factor.
One has a gain of two and the other has a gain of five.
 
My question is: does the gain factor affect the dynamic range?
My gain of two amp sounds more flat, less dynamic than the
gain of five. What do you think? Any comment on this?

 

[gimble]

As far as I know, only if you have an input signal with Vpp high enough that the gain oversaturates the op-amp and causes clipping. There are probably other characteristics at work. "Dynamic range" is mainly relevant if we're talking about recording or compression.
 
If you're listening with the same headphones, I'd say a gain of five is simply more appropriate to their behavior as a load, i.e. a gain of two is underamplified for the headphones in question.
 
If you're listening on different headphones and on different amps simultaneously then comparing the result - stop that. You only want to modify one factor at a time if at all possible.

 

[Avro_Arrow]

I guess you didn't understand the question or maybe I didn't state it very well.
 
The question has nothing to do with headphones types or amps clipping...
 
I asked if the rate of gain affects the way an amp sound (all other things being equal).
We will assume that the amp is always operating in it's linear range.
 
Anyway, I did the math and I guess it's all in my head (or ears...).
If you adjust the volume so the output signal is the same on both
amps, the voltage range is the same no matter if the gain is two
or five.

 

[Beftus]

Talking 'bout opamps? If so, some opamps want to be run at higher gains only. Some aren't stable when run at low(ish) gains.

 

[Avro_Arrow]

Yes, the data sheet will tell you if a particular op amp is unity gain stable or only stable
down to a specific gain. By making your amp at least a gain of five you have a wider
choice of op amps for rolling...
 
Quote:

Talking 'bout opamps? If so, some opamps want to be run at higher gains only. Some aren't stable when run at low(ish) gains.

 

[gimble]

The question has everything to do with clipping. The only way what you said will happen is if you oversaturate the op-amp.
 
Otherwise it will reproduce the signal as best as it can given its slew rate and so forth - which is overspec for audio frequencies. Op-amp suitability for audio amplifiers goes on other factors, mainly how easy the IC is to work with, whether there are any peculiarities of the frequency response spectrum which make it undesirable, and what the noise behavior is like.
 
Most likely, the only reason that it sounds different is that you're generating a different signal. The gain is the cause but the gain is not the reason. You are altering the signal amplitude, and possibly the power response of the circuit.
 
Do you have access to bench equipment? Run a 2kHz tone into each amp then tune the volume divider so that the output signals match on an oscilloscope. Then listen to see what happens. They should be very similar. If not, you should analyze the circuit's power response in detail.
 
It's also possible that your op-amp (or other gain circuit) has different response at different gains...but you're comparing a small enough range that it shouldn't matter if it's stable at both. You can chart the frequency response out in LabVIEW fairly quickly at various gains to see, and it's probably in the data sheet, but I'm assuming that you specced parts which actually work at the set gains.

 

[digger945]

Does Class A or AB matter for your amp?
If you increase the voltage gain for a specific amount of power, will the current decrease appropriately, and would the impedance of the load play a part in choosing both?
Will the output waveform at a gain of 5 be more distorted than an output waveform at a gain of 2?
Does increased feedback decrease speed(slew rate)?

 

[Avro_Arrow]

Everyone is free to discuss this further if you wish, but I have answered
my own question to my satisfaction.

 

[cobaltmute]

It has been asserted by some that op-amps that have been compensated for unity gain (eg. OPA227), do not sound as good as their un-compensated brothers that have to run at gain 5+ (e.g OP228).  I personally have never tested.

 

[gimble]

Sounds like it would be moderately easy to test on a case-by-case basis since the main requirement is a gain 5 CMoy with an IC socket. A function generator & oscilloscope wouldn't hurt to confirm that the output is similar (noting that drive power is still a bit of an elephant in the room).
 
It does involve meaningfully different internals, so it could be true. Is it? No idea.

 

[cobaltmute]

It could prove a meaningful result for a specific model of op-amp.  It may not hold true for all unity gain op-amps as they may be designed differently.
 
The other test you could then do is use something like a NE5534 and use different compensation capacitors to see the affect of the compensation at the same gain level.
 
And of course, all these op-amps may sound different at different rail voltages, as has also been claimed.  So you need to test gain X at rail voltage Y1, Y2, Y3,etc.

 

[nikongod]

Did you volume match? Set 1 amp to your preferred level by ear, Play a 1Khz tone through both amps, and measure & match output voltages into a resistor or your headphones.
 
Something I personally find annoying with high gain amps is how difficult it is to actually set my desired level.