We should all really know this, but when people worry about their volume pot being past 12 O’clock, I’ve often struggled to convey just how little volume 12 O’clock is and how none linear the pots are so I just checked the Pot on the amp I’m currently working on.
This is 0 volume on the dial.
And this is where the volume pot is when it’s passing 25% (1/4) of the input
At half way it’s passing about 1/40th if the input signal, the less input signal your passing in the more relative noise your amplifying.
Now this is a Stepped attenuator and a conventional audio pot is a bit less aggressive, so you might be passing as much a 1/10 of the input signal through at 12 O’clock, and 1/4 would be closer to 3 O’clock
Isn’t this very specific to the pot and implementation? I remember with AIC10 for instance if the pot is at 12 o’clock the amp is at unity gain (which I believe is the term for passing 100% of the input?)
No the log taper will be much the same across all audio pots, stepped attenuators tend to be more aggressive to give finer control in the usable portion of the volume.
The reason they are logarithmic is because that matches how we perceive loudness, since it isn’t linear with amplitude.
Unity gain is where the input level = the output level, on the AIC that wil mean that it’s probably ~1/10th-1/20th of the input being amplified.
Almost all headphone/pre amps are fixed gain devices with the pot before any amplification, so if the amplifier has a gain of say 20, the you have to scale the input down by 20 to get that output level. There is nothing magical about unity gain.
The Extra resistors on the headphone out of the AIC, just scale the output down reducing overall gain.
Depends, you have 2 volume pots, so your just trading off one setting with the other.
I’d start with max out the volume on the second amp, and ensure the input is low enough it doesn’t clip (since that basically bypasses on of the pots), but that may result in the preamp seeing a VERY low amplitude signal and a low amplitude signal on the interconnects that could result in excess “noise”.
Unity Gain on one or the other is probably a safe bet.
How loud the input level is on the preamp (where it’s volume is) will change it’s distortion profile, and potentially how the sound is “colored”, so it may just be a question of playing with it until you like it. Generally lower volumes on the pre will result in a sound less colored by the pre.
The simple answer is: get a low impedance pot. From the factory they’re usually around 50K. So you’re throwing away a lot of signal and having to amplify a tiny signal (adding more noise). With a 10K pot, you’re throwing a lot less of the signal so a lot less has to amplified (less noise).
The not so simple answer is: it depends on gain devices. Some are more linear at higher gain. With higher gain, you need a smaller signal or you’ll blow your ears at 9:01 o’clock on the dial.
Also, some preamps and headamps have higher than needed impedance (throwing away a lot of signal) so they don’t blow the outputs of old or “unique” tube sources.
The one thing you absolutely shouldn’t worry about is what o’clock your volume is at. Turn it up until you have satisfactory levels. If it sounds compressed or distorted, it’s not powerful enough to drive your cans.
This is not correct there are advantages to low impedance pots, most notably they are less prone to RF noise, but you input impedance is determined by the pot so you have to be careful. And picking an impedance is a compromise. If low impedance pots were always better, manufacturers would use them.
Volume pots though just act as dividers regardless of there rated impedance, so the logarithmic curve is identical regardless of impedance.
Very different, most guitar pots are linear not log and much smaller values.
Impedance in an amplifier pot is primarily what sets input impedance. It can’t increase the impedance of a circuit, but often input circuits have VERY high impedance, and the total winding of the pot is in parallel with that, resulting in the input impedance being more or less the impedance of the pot. It’s a bit more complicated that that because circuit impedance varies with frequency, and you really want to pot that gives a consistent input impedance through the audible range.
There are secondary considerations, given in most cases the pot is at a low % of it’s total resistance the remainder of that is in series with the Miller Capacitance of the input device and forms a low pass filter, that’s often in the megaHertz region, lower values result in a lower pass frequency, so they tend to be less prone to picking up stray AM. That may not be the only reason they pick up AM transmissions.
It’s a really easy property to demonstrate, any higher impedance pot with a length of wire out side the case and you will hear when you touch it. Stick a smallish resistor in parallel with it and the hum goes away.
So you have this trade off of less noise for low impedance pots vs lower input impedance resulting in lower compatibility with sources. The pacific for example requires an input impedance of at least 20K Ohms.
So the common values are 100K, 50K and very occasionally 20K or even 10K, but if your amp is in the latter Half of that list it’ll be picky about the source.
Let’s define signal.
Voltage wise which is the amplitude of the signal a divider, well divides, it doesn’t matter of you divide 10K/100K or 1K/10K.
This only behaves this way because the input impedance of the actual input stage is usually VERY high, so you can ignore it in the divider calculation.
Current wise, the lower impedance POT requires MORE current I=V/R from the source.
Power wise well given voltage is constant that’s’ just proportional to Current.
In most amplifiers, with a pot, pretty much all the current at the input flows through the pot.
So I guess I’m just missing your point.
If the pot were just in line with the signal, or the input device impedance were low, it would be true because the divider would be the pot against the impedance of the input device. BUT because the input impedance of the device is usually large relative to the pot, you can actually more or less ignore it in the divider calculation. Input impedance for a Tube is in Megaohms, Mosfets it’s a function of the capacitance of the gate, and again VERY high, some transistor stages aren’t, but generally you design an amp for it to be as high as possible.
Edit I think I understand the confusion. The input of MOST headphone amps looks like this. R2/R3 make up the Pot and R1 is the impedance of the actual input section.
All that matters is Vin.
So for a 1V input with a 10K pot (as depicted)Vin would be ~0.1V (actually it’s 0.0999V because of the 1Meg resistance) if I swap that out for a 100K pot at the same 10% it would still be ~0.1V (actually 0.0991V)so were talking less than 1% difference.
Even if there was a significant difference simply tweaking the pot up slightly would result in the same Vin.
Ok, explain something to me:
For every single hp amp (and a few speaker amps where I’ve just put a pot between source and amp) that I’ve swapped a high impedance pot for a low impedance one, I get better transients/impact. Why is this?
It’s hard to guess but, as I said there are good reasons to use smaller pots, if your source can drive it. Most notably less noise.
If I had to guess it’s probably the lower impedance resulting in one of the implicit filters moving the cut off point down. All amplification devices have some capacitive coupling between the pins, and that gets “amplified” by the gain of the stage, and can become none trivial. So you end up with all sorts of “filters” implicit in a circuit that limit bandwidth, and cause phase shifts.
Or possibly it could be just the reduced noise floor.
Additionally the impedance of the pot can sometimes influence the sound. For example a 100k pot can cause some higher frequencies to be attenuated more and so your amp could lean darker than if it had a 10k pot. This can be a desirable effect depending on the architecture of the amplifier. It might also only be audible to some people depending on their hearing range.
Yes it’s a function of the capacitance of the input device, the gain of the input stage and the impedance of the pot, the upper part of the Pot is part of a low pass filter formed with the “miller capacitance” of the input device.
I believe the Aegis build instructions tell you not to use a larger impedance Pot for this reason, since the 6SL7 has a gain of about 70, and paralleling the 2 half’s basically doubles the input capacitance so you’d be looking at almost ~400pF of capacitance on the input. And any large impedance pot would cause high end roll off.
