Thanks, I'll have to give that a look.
building an active balanced ground? - Page 9
Hmm, I just measured the DC offset on my INA134/137 resistorless circuit. I've got about 1.3mV off each opamp to input/reference ground. When measuring from output left/right to output ground, I get aroun 0.1mV. Not bad I suppose. My multimeter can't go any more precise than that.
Argh, so close yet so far. I finished building the s22 and 3x a20 boards intending to hook them up to my active balanced ground circuit... except one of my a20 boards is blown (the other two are just fine though). Grr, and no spare parts on hand either. Sigh, I don't suppose anyone has a spare MJE243 and MJE253? (Q5/Q6 from here)
Whooo, it works! Behold the monstrosity...
(not shown: the Fuze providing music)
- Sansa Fuze for music
- s22 providing power (+/-18V) for the ABG circuit and the three a20 boards
- the ABG circuit feeds OL, OR, OG each into its own a20 board
- connect output from each a20 board to the headphone
- no resistors on the current ABG circuit, using INA134/137 which are differential receiver chips that have built in resistors
At the moment I've only tested it with cheap airline headphones so I don't really know how good it sounds. The a20 boards are extremely susceptible to temperature variations and produce voltage offset due to that, so I'm going to rig up a temporary cardboard box to protect it from drafts (like from my arm movement or breathing) before using my good headphones.
Whoa, I missed this. The project in the pic above has been on the shelf. One of the a20 boards blew and I don't have a replacement. I do have some b22 boards though...
I did build a strictly opamp version for a member though, and I thought it sounded quite good though ridiculously susceptible to EMI. It used three opamps: one as input buffer and control for a bass boost switch, one for the left and right channels, and one for the ground. I have plans for another one using some proper differential opamps and BUF634's (so 6 chips in total) and no resistors in the circuit... but who knows if I'll ever get around to it.
I've gone as far as proof of concept that the implementation works, but I lack the time/resources to take it beyond the "reconfigured cmoy" stage.