Changing gain for Beta22

Guess I might as well tell you... Although you install an input volume pot, the configured gain is a matter of some components installed on each board, see Parts -> "gain" here: http://www.amb.org/audio/beta22/

No it is not possible.

A gain of 5 would be a nice middle ground.

What I did was to socket the gain resistors (2 resistors per channel) so that they could be changed quickly and easily with a small pair of needle nose pliers. I also did this to the S22 power supplies to make changing voltages easy. There are instances where you would want a gain setting that is much higher than normal. One would be some hard-to-drive planars. Another would be some Stax transformer boxes, or for ultra low level listening without using a transformer box(just driving the stators with the amp output). Low gain is good for other dynamics. It's nice to have the capability so you can listen and experiment for yourself.

You should also consider socketing the feedback caps as well, if you decide to do the resistors.

Did you also socket the compensation caps C2-C5 for stability vs bandwidth?

If the volume knob is not resting comfortably against its minimum attenuation position (maximum volume) you have more gain than you need. 

This dosnt account for the fact that different sources put out different nominal signal levels, and some recordings are quieter than others, and that some headphones are truly inefficient. Throw in a comfy 20db of headroom to account for that - if your volume knob is not pointing to 2 or 3 o'clock at your average listening level you have too much gain.

If you have VERY different source outputs, and headphones, and have never spun the volume knob past noon on a clock face your wasting ooh so very much.

Try voltage gain of 4 (which is still too much for most headphones, but mneh) . Its the same as having gain of 8 (only better) if you just spin the knob a few degrees/clicks on a stepper higher. 

IIRC, attempts to switch feedback resistors on the B22 have caused instability. The hardest thing about selecting gain for a headphone amp is admitting that you only need a very little bit, and often not even that much.

the B22 should have extremely low input noise - so a low impedance volume pot should not give noise probelms even at 20 dB attenuation - source distortion driving the pot impedance and with analog pots the poor matching at higher attenuation may be problems

if your source does good digital attenuation - 24 bit or properly dithered 16 bit you should use digital attenuation - "bit loss" need not be a probelm in good digital source

another wonderfully wasteful approach is a power output resistive divider "Hiss Buster" style - with the B22's high power you could have a 1 Ohm output Z divider

Quote:

Originally Posted by jcx 

the B22 should have extremely low input noise - so a low impedance volume pot should not give noise probelms even at 20 dB attenuation - source distortion driving the pot impedance and with analog pots the poor matching at higher attenuation may be problems

 

if your source does good digital attenuation - 24 bit or properly dithered 16 bit you should use digital attenuation - "bit loss" need not be a probelm in good digital source

 

another wonderfully wasteful approach is a power output resistive divider "Hiss Buster" style - with the B22's high power you could have a 1 Ohm output Z divider

agree 100%, I changed to digital last year (32bit and over 1mhz) and theres no turning back now, all of my sources use it and none of my amps have a pot. its not terribly audiophile approved though, I'm not sure why, as I cant imagine a more transparent solution for digital source than a well designed digital attenuator; preferably one that uses the dac registers like the sabre