Dear friends!
There is a headphone amplifier built on OPA2134 op-amp. Powered by a transformer power supply (-12V - 0V +12V).
Does the regulated voltage (LM317-LM337) have an advantage over the unregulated one for this amplifier?
It seems to me that the voltage reserve is quite large and voltage fluctuations even within 10-14V will not have a significant effect on the sound.
There is a headphone amplifier built on OPA2134 op-amp. Powered by a transformer power supply (-12V - 0V +12V).
Does the regulated voltage (LM317-LM337) have an advantage over the unregulated one for this amplifier?
It seems to me that the voltage reserve is quite large and voltage fluctuations even within 10-14V will not have a significant effect on the sound.
IIRC, PSRR falls off with increasing frequency. So, maybe practical PSRR depends on frequency of any unwanted noise components getting into the power. I will also say this, if you can change from one type of bypass cap to another and hear a real difference, then something is apparently going on at the power supply level. Maybe you should investigate further in that case. You can also take an FFT and look for spurs at harmonics of the AC power line frequency. Maybe consider that a 16-bit dac has a theoretical dynamic range extending down to somewhere around -93dBFS. A 24-bit dac can have considerably more dynamic range. Do you want your system DR to be limited by power quality?
Last edited:
Adding regulators (and their associated capacitors, heatsinks) increases circuit complexity and cost a bit. LM317/337 typically need ~2V dropout, so your transformer must provide enough voltage after rectification for regulation to maintain ±12V output. If your transformer is just ±12V AC, you might not have enough headroom.