2 ms·
Recently I became interested in DiY audio gear and I spent a lot of spare time on Rod Elliott's ESP audio pages. One project is https://sound-au.com/project07.h
by MisterTea 7y ago
Recently I became interested in DiY audio gear and I spent a lot of spare time on Rod Elliott's ESP audio pages.
One project is https://sound-au.com/project07.htm https://sound-au.com/project07.htm, a discrete op amp project. It's been awhile since uni days so I bought a bunch of pnp and npn bjt transistors from digikey and spent a few days at my bench tinkering with the circuit. different gains, miller caps to set stability and bandwidth, how input and output impedance affect gains, etc. Very satisfying seeing your little mess of transistors come to life and amplify a signal as designed.
One thing it taught me was how the different sub circuits come together to make a power or operational amplifier. Long tail pair aka differential amplifier, current mirrors, current sources, Vbe multiplier, class A stages, Class AB stages, etc. Another great article which builds on it is https://sound-au.com/amp_design.htm https://sound-au.com/amp_design.htm. This is where you get to see how these building blocks come together. This basic topology is present in most amps be they little op amp chips or big beefy audio power amps: input -> diff amp -> class A gain -> class AB buffer -> output.
Interesting to learn that there isn't much difference between a little TL071 and my Rod Elliott P3A.
- bcaa7f3a8bbc 7y agoWe'll never have the luxury of laser-trimmed resistor or matched transistor, an integrated opamp usually outperforms discrete designs, at least for low-power applications. But yeah, it's much fun to build an integrated component from scratch and see how it works. BTW, NwAvGuy, the audio guru who have vanished, has some interesting criticisms on the dismissal of opamp in the audiophile community. https://nwavguy.blogspot.com/2011/08/op-amps-myths-facts.html https://nwavguy.blogspot.com/2011/08/op-amps-myths-facts.htm...
- Kirby64 7y agoFor laser trimmed resistors, there's always hand sorting resistors. Even today, that's how I've seen it done in commercial op-amp characterization. Tests like CMRR (Common mode rejection ratio) require precisely matched input resistances or the result is thrown off by mismatches.