I've always considered 10dB as the 'nominal' value for line stage gain in a preamplifier. Does that sound about right to you?
The trick to having high dynamics is in the gain of the preamp and the amplifier. The old krell ksa50 has a gain of x40. Maybe that's also why it seems to play so loud. However, you can start from the maximum signal of a cdplayer with DAC chips powered by 5V; then divide by 2, and from there go up the chain.
In the end, the more gain there is, the louder it seems to play, but this clashes with the acceptance of the various stages before clipping
In the end, the more gain there is, the louder it seems to play, but this clashes with the acceptance of the various stages before clipping
During what time period?
When turntables were a primary source I think it was likely more. Say 20dB. After the takeover of CD, it might be less as you suggest, 10dB. But even then, that might be too much or unnecessary.
When turntables were a primary source I think it was likely more. Say 20dB. After the takeover of CD, it might be less as you suggest, 10dB. But even then, that might be too much or unnecessary.
I believe the 2Vrms output for CD players and other digital devices is at 0dBFS. All other analog sources have a reference level at some standard for the medium that is well below full "undistorted" output, which was actually 3%THD. Example: LP standard is 3.54cm/sec at 1KHz, but there's headroom above that, and the EQ curve, and tape reference is 200nW/m, which depending on tape speed and track format might be 10-12dB below 3%. Or less, if it's a cassette.
I don't recall there actually being a level difference between the CD and other sousrces when it came out, because back then the 0VU reference level was -20dBFS or a bit higher (puts the average CD output voltage in line with 200-400mv), but if the CD was mastered in the 2000s or 2010s, it's probably mastered hot, so it plays louder. You can't fix that with any sort of level match because it's not consistent across all CDs. In the early days the CD mastering standards called for the highest peak to be just below 0dBFS, and the recommendation was little or no limiting. Those days are long over.
The gain of a preamp is whatever is required to take an average input voltage high enough to drive a power amp and speakers to an average level, and still be able to "turn it up" to impress your friends by clipping your amp. Start with the maximum input the amp can take for full power before clipping, and work backwards from there.
I don't recall there actually being a level difference between the CD and other sousrces when it came out, because back then the 0VU reference level was -20dBFS or a bit higher (puts the average CD output voltage in line with 200-400mv), but if the CD was mastered in the 2000s or 2010s, it's probably mastered hot, so it plays louder. You can't fix that with any sort of level match because it's not consistent across all CDs. In the early days the CD mastering standards called for the highest peak to be just below 0dBFS, and the recommendation was little or no limiting. Those days are long over.
The gain of a preamp is whatever is required to take an average input voltage high enough to drive a power amp and speakers to an average level, and still be able to "turn it up" to impress your friends by clipping your amp. Start with the maximum input the amp can take for full power before clipping, and work backwards from there.
With my mixing and matching of various line stage designs with various power amp designs, this is something I want to dial in much better. I'm using
stepped attenuators with all builds and prefer the attenuator to reach near full rotation at a chosen max volume so all of the steps of expensive attenuators can be utilized..
A few variables here. Hmm, more to work on!
stepped attenuators with all builds and prefer the attenuator to reach near full rotation at a chosen max volume so all of the steps of expensive attenuators can be utilized..
A few variables here. Hmm, more to work on!
I go for 0dB - 6dB which puts the volume potentiometer in a good position for power amp with around 28dB gain. Nothing worse than a pot with very loud volume at 9 o'clock and little fine adjustment.
With each preamp build, I am finding that I must attenuate the input by as much as 1db. Then again, it does matter which input I am using. I do not know the actual gain of my tube based phono preamp, but it is lower than the other sources. Just trying to get the right amount of useable volume control at the knob.
Every time I've put an output/buffer board in an RIAA preamp, I've liked it less. I don't usually need much gain, so I typically go passive, with no board at all. As close as I can measure it, the gain of my output wire is 0 dB. 😉
'Normal' or not, it always ends up to pairing one setup as best you can with the rest of the system. When it works there is no substitute for the satisfaction.
Building a preamp which allows to connect old equipment can be realized using a 2 stage gain concept. Design your different input stages with fixed select-able gain e.g. -12/+15dB and you are on the safe side. To be able to play silent source material very loud let the preamp do a extra +12 to +18 db gain (volume controlled).
With this trick you can play old tape or radio sources (~ 200mV rms) up to modern CD or sound card levels (~ 2V rms ) tuned at their sweet spot.
Have a look at my SA2022 preamp concept. Maybe you find some ideas there.
BR, Toni
With this trick you can play old tape or radio sources (~ 200mV rms) up to modern CD or sound card levels (~ 2V rms ) tuned at their sweet spot.
Have a look at my SA2022 preamp concept. Maybe you find some ideas there.
BR, Toni
Good idea to be mindful about is that ~ 2.0 volts input will drive your power-amp to the rails.
If your pre-amp is selecting among 2 volt inputs, perhaps the best gain option is a buffer with a gain of 1.
Lately I have bumping my head against the pre-amp max output voltage will be ~ 10 volts. There will likely be a op-amp with 15 volts + / - rails in chain. Either the input or output will clip at or near 10 volts.
Beware of too much gain.
If your pre-amp is selecting among 2 volt inputs, perhaps the best gain option is a buffer with a gain of 1.
Lately I have bumping my head against the pre-amp max output voltage will be ~ 10 volts. There will likely be a op-amp with 15 volts + / - rails in chain. Either the input or output will clip at or near 10 volts.
Beware of too much gain.
That is true! E.g.: SA2022 preamp has a default gain of 0dB , but volume control can switch the opamp buffer output stages gain from 0 to +6, +9.5, +12dB when needed to get the maximum THD+N out of the signal.
Last edited:
🙄Poweramp gain switching is difficult, if you want maximum stability, slew rates and THD+N.
And of course dynamically - controlled from preamp volume control ... click and pop free ... 😎
And of course dynamically - controlled from preamp volume control ... click and pop free ... 😎
My DAC (Denafrips Terminator Plus) has a maximum output of 4V RMS on the balanced outputs, so even a 1x gain linestage has more than enough output to clip my amps, even with digital recordings from the 80s. Best to check your sources and figure out how much gain you really need.
I guess one thing you could do, is to make yourself a list of twenty or forty "typical preamplifiers" sold by reputable high-end audio companies. Then dig through their marketing material to find out what is the line stage gain, and enter that information into a table.
Once you have the data on all of these high end preamplifiers, you can plot it, or make histograms, or calculate mean and standard deviation, or any other gyrations that seem like good ideas to you. And then you'll know what high end manufacturers think is a sensible gain for a preamp line stage. Naturally you can always say, but, but, but, that's JUST THEIR OPINION. It frees you to use whichever preamp line stage gain that YOU like, and to hell with reputable high-end audio companies' opinions.
Once you have the data on all of these high end preamplifiers, you can plot it, or make histograms, or calculate mean and standard deviation, or any other gyrations that seem like good ideas to you. And then you'll know what high end manufacturers think is a sensible gain for a preamp line stage. Naturally you can always say, but, but, but, that's JUST THEIR OPINION. It frees you to use whichever preamp line stage gain that YOU like, and to hell with reputable high-end audio companies' opinions.