has anyone tried putting a bandpass filter on the output of a headphone amp?
my reasoning is that since digital audio and human hearing is only 20hz-20khz that anything above or below (noise) would be filtered out, would also have the added benefit of removing dc from the an opamp output.
highpass set at 20hz and lowpass set at 20khz, since highpass would be first in line the capacitor sits first on the main output which should theoretically remove the dc (if any) from the output so protecting headphones.
or am i being daft.
my reasoning is that since digital audio and human hearing is only 20hz-20khz that anything above or below (noise) would be filtered out, would also have the added benefit of removing dc from the an opamp output.
highpass set at 20hz and lowpass set at 20khz, since highpass would be first in line the capacitor sits first on the main output which should theoretically remove the dc (if any) from the output so protecting headphones.
or am i being daft.
Typically, it's done on the input stage. Taking a basic CMOY style headphone amp, putting a 100pF across the feedback resistor reduces the gain around 50 Khz to 160 KHz.
As an example, using RJM's Sapphire, the feedback resistor is 20K. Putting 100pF across that resistor reduces gain above 79.5 KHz. Better would be 50 pF, which gives 159 KHz. As most DACS run above 190 KHz, the noise they generate is not amplified with the audio signal.
The input capacitor on most headphone amps is designed to have no effect at 20 Hz. Again, taking the Sapphire, the 0.47 uFd has a -3 dB of 7 Hz. To move that up, just use a smaller capacitor. But you might not like what the effect is on the lows.
Before I get called out on this, yes the cap across the feedback resistor is to stabilize the opamp. However, it also reduces the gain of the circuit from the typical 12 dB or 20 dB to 0 dB above the frequency the cap is set to. Technically, it's not a filter, though it does not amplify the noise leaking from the DAC.
As an example, using RJM's Sapphire, the feedback resistor is 20K. Putting 100pF across that resistor reduces gain above 79.5 KHz. Better would be 50 pF, which gives 159 KHz. As most DACS run above 190 KHz, the noise they generate is not amplified with the audio signal.
The input capacitor on most headphone amps is designed to have no effect at 20 Hz. Again, taking the Sapphire, the 0.47 uFd has a -3 dB of 7 Hz. To move that up, just use a smaller capacitor. But you might not like what the effect is on the lows.
Before I get called out on this, yes the cap across the feedback resistor is to stabilize the opamp. However, it also reduces the gain of the circuit from the typical 12 dB or 20 dB to 0 dB above the frequency the cap is set to. Technically, it's not a filter, though it does not amplify the noise leaking from the DAC.
Last edited:
True. However, no gain is still 20 dB of so below the signal. Auditory masking serves for some effect.
The other issue with a bandpass filter is it needs to be isolated - a pot in front of it would change the frequency response. The system would need to be something like - active buffer (known low impedance drive) - bandpass filter - active buffer (or gain stage) - pot - power buffer - headphones.
The active buffer/gain stage/pot could be combined into one as Doug Self does on his preamps. But, it's inverting, so an inverting stage before or after that would be necessary.
While there are ways to do all this, it will involve more active stages. Headphone amps are just amps. Messing with the signal is what preamps with tone controls and filters are designed for.
The other issue with a bandpass filter is it needs to be isolated - a pot in front of it would change the frequency response. The system would need to be something like - active buffer (known low impedance drive) - bandpass filter - active buffer (or gain stage) - pot - power buffer - headphones.
The active buffer/gain stage/pot could be combined into one as Doug Self does on his preamps. But, it's inverting, so an inverting stage before or after that would be necessary.
While there are ways to do all this, it will involve more active stages. Headphone amps are just amps. Messing with the signal is what preamps with tone controls and filters are designed for.
ok, i can rule out a pot as there is a digital one on the RPi DAC+ which feeds the headphone amp. i'm going to test out removing input DC blocking on set of another headphone amp boards for giggles as i cant see any DC from the RPi outs and all the blurb i have read seems to state that there should be none. once i have done that then i'll try DC blocking the output to the headphones.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.