[hagtech]

Darn it.  I made a mistake.  When I originally measured the SNR performance of my phono stages, I used a custom-built A-weighting filter.  Well, turns out I built the thing wrong.  Today I rebuilt it to spec and double checked it against the chart.  Now the filter is right on the money and gives me correct readings.

The good news is    that my readings were about 8dB off.  I took out a Trumpet and measured residual noise at -74dB referenced to 5mV.  That is "A-weighted".  You can also say it's 1uV input referred noise.

Same thing with Cornet.  It is now specified at 75dB SNR.  That's darn good for a tubed stage.  

jh

[hagtech]

FYI - here's the circuit I am using.  Found it on the net, with credit given to an old Ampex tape deck manual. (I did not design this circuit, but would be proud if I had).



The modification I made was to the front end.  The input buffer applies gain so the output reads 20dB too high.  That is, it makes up for the insertion loss of the filter and adds an extra 20dB.  I do this since the signals to be measured are typically very small to begin with and I didn't want the low noise opamps to add their own noise to the mix.  They do, but not too much.  The noise floor of the circuit is pretty low.

jh