Flood in the Internet
These days search engines throw tons of schematics of "Class A headphone amplifiers" on us. There we can find anything: from monsters capable of driving big loudspeakers to humble OpAmps heavily loaded by 30 Ohms without any hope for help. The big wave of so-called hybrid headphone amps is scary: they look very high-end'ish, and I am the first one to advocate valves in sound reproduction, but mind you - using 15V at anode of ECC88 is insane.
Probably I would never be writing about this headphone amp schematic. But recently I observed one good man torturing a very decent vacuum tube with unreasonable low anode voltage and grid currents in a hopeless attempt to build what supposed to become his reference headphone amplifier. I tried to give him few hints here and there, but had to give up on that once he told me the model of the headphones he was building that monster for: that was some computer-grade thingy, "hi-fi model headphone intended for use with MP3 players and computer systems".
Today I am striving to avoid negative feedback loops in my audio designs. But this article is about a humble all-common-sense OpAmp-based headphone amplifier, the one that served me well for many years already since the last Century. It's easy and inexpensive to build, virtually does not require any tuning and yet can serve very sophisticated ears 😉
The ideas that this amp is build on have been found on the Net, obviously still in the last Century 🙂 Should you know the originator of this or similar layout - please drop me a note - I'd be happy to share the source with my readers too.
Let's take a quality operational amplifier and force its output stage into the real class "A" operation mode. For achieving that goal we simply need to load its output by a current source. The idea most probably belongs to Walt Jung and was first used in Philips DAC960 in nineties. As you know except the output stage all other OpAmp's guts already work in pure class "A". The effect of this trick that shifts OpAmp into the linear mode is illustrated in the article about the Wien bridge oscillator. There we used a budget OpAmp thus the effect was clearly visible.
Still the little OpAmp is not the best choice for driving 30 Ohms headphones. It cries for a current amplifier. Here we'll use easily available IR MOSFET's. Provided the complementary source followers idling at some 100 mA it will stay in pure class "A" while working on 30 Ohms or any other load with higher impedance.
- OpAmp = OPA2134
- VD1 = red LED 1.7 V
- VT1 = MPS2907A
- VT2 = IRF610
- VT3 = IRF9610
- VD2, VD3 = 9...15 V Zeners
- C1 = 1uF (film)
- RP1 = 100 KOhm, log
- R1, R3 = 1 KOhm
- R2 = 100 KOhm
- R4 = 2 KOhm
- R5, R6 = 3 KOhm
- R7 = 360 Ohm (*)
- R8, R9 = 100 Ohm
- R10, R11 = 3.3 Ohm
- C2...C7 = 0.1 uF (film)
- C8, C9 >= 4700 uF x 35 V
- C10 = 47 uF x 16 V (aluminum)
The bias voltage on MOSFET's gates is developed over R5 with the same current that loads OpAmp and shifts it into the class "A". The only tuning that this layout might require would be setting the idle current through the source followers at around 100 mA. For tuning this idle current one may need to adjust value of R7.
A small film capacitor C11 of 100pF between OpAmp' output and its inverting input may proof useful in order to prevent oscillations when the amp is maxed out on output voltage (i.e. "clipping").