I’m adding a gain stage and buffer to a tube crossover.
Initially I put the gain and buffer stages after the filters but the degradation in SNR was significant. So, I moved the gain and buffer stage to before the filters.
While better, the SNR is still significantly degraded compared with the filters on their own.
Below are the schematic for the 125Hz to 14kHz section, amplitude vs frequency before and after and 1kHz real-time analysis before and after.
For the real-time analyses I’ve adjusted the input level to match the output levels.
Without gain and buffer stage: noise floor around -118dBFS, noise 0.11% and THD 0.024%.
With the gain and buffer stage: noise floor around -92dBFS, noise 1.11% and THD 0.2%.
The increase in THD I’d expect with the common cathode stage in there.
However, the noise has gone from 0.11% to 1.11%. The noise floor has gone up by 26dB.
I haven’t connected the crossover section to an amp yet to see if the increased level of noise is audible. Maybe noise 75dB below the fundamental is OK.
Is this the best I can expect or is there a better way of doing this? Different tubes? More supply filtering or a regulated supply?
Initially I put the gain and buffer stages after the filters but the degradation in SNR was significant. So, I moved the gain and buffer stage to before the filters.
While better, the SNR is still significantly degraded compared with the filters on their own.
Below are the schematic for the 125Hz to 14kHz section, amplitude vs frequency before and after and 1kHz real-time analysis before and after.
For the real-time analyses I’ve adjusted the input level to match the output levels.
Without gain and buffer stage: noise floor around -118dBFS, noise 0.11% and THD 0.024%.
With the gain and buffer stage: noise floor around -92dBFS, noise 1.11% and THD 0.2%.
The increase in THD I’d expect with the common cathode stage in there.
However, the noise has gone from 0.11% to 1.11%. The noise floor has gone up by 26dB.
I haven’t connected the crossover section to an amp yet to see if the increased level of noise is audible. Maybe noise 75dB below the fundamental is OK.
Is this the best I can expect or is there a better way of doing this? Different tubes? More supply filtering or a regulated supply?
