Re: A Low Noise Balanced Input Moving Coil Preamp Using the ZTX851
Posted: Wed Jul 03, 2019 7:56 am
Thanks Hans for checking the sims and the FFTs of motor on/off.
I re-checked some measurements yesterday using the Phono Transfer System RIAA EQ board with the ZTX851 MC preamp.
I did this to allow me to switch between flat and RIAA EQ and trim out the insertion loss from A/D loading of the final output.
With an Rsource of 10Ω and an overall gain of 68 dB I'm still measuring an Ein of almost exactly -141 dBu.
That works out to a total Rnv of 14.5Ω.
Rsource plus Rgain is 11Ω leaving 3.5Ω for both transistors or approximately 1.75Ω rbb' each.
-141 dBu Ein over a 20 kHz BW works out to be about 0.48nV√Hz.
These numbers are consistent with what I measured earlier.
I'm not sure how AudioTester handles the bandwidth of its level calculations.
If I apply a second-order LPF to the input (in AudioTester) it's not reflected in the level measurements only the plotted FFT.
The bandwidth setting in the Analyze dialog only affects noise calculations for THD measurement, not level.
The Weighting filters can be reflected in the level measurements but there isn't one defined for noise measurement. (i.e. a 1,2 or 3rd order LPF.)
I need to figure out how to define a "User" filter.
So I can't tell you what the bandwidth is for level measurements but it must be pretty close to 20 kHz or the sims and measurements wouldn't be within the 0.5 to 1.5 dB that they are.
At this point its so much quieter that the DL-103's 40Ω||100Ω termination that if I'm off by a dB I'm way past caring.
Trying to hear the noise or hum floor at these levels while playing a record is risking one's hearing, monitors and may be death defying.
I re-checked some measurements yesterday using the Phono Transfer System RIAA EQ board with the ZTX851 MC preamp.
I did this to allow me to switch between flat and RIAA EQ and trim out the insertion loss from A/D loading of the final output.
With an Rsource of 10Ω and an overall gain of 68 dB I'm still measuring an Ein of almost exactly -141 dBu.
That works out to a total Rnv of 14.5Ω.
Rsource plus Rgain is 11Ω leaving 3.5Ω for both transistors or approximately 1.75Ω rbb' each.
-141 dBu Ein over a 20 kHz BW works out to be about 0.48nV√Hz.
These numbers are consistent with what I measured earlier.
I'm not sure how AudioTester handles the bandwidth of its level calculations.
If I apply a second-order LPF to the input (in AudioTester) it's not reflected in the level measurements only the plotted FFT.
The bandwidth setting in the Analyze dialog only affects noise calculations for THD measurement, not level.
The Weighting filters can be reflected in the level measurements but there isn't one defined for noise measurement. (i.e. a 1,2 or 3rd order LPF.)
I need to figure out how to define a "User" filter.
So I can't tell you what the bandwidth is for level measurements but it must be pretty close to 20 kHz or the sims and measurements wouldn't be within the 0.5 to 1.5 dB that they are.
At this point its so much quieter that the DL-103's 40Ω||100Ω termination that if I'm off by a dB I'm way past caring.
Trying to hear the noise or hum floor at these levels while playing a record is risking one's hearing, monitors and may be death defying.