[hpsdr] 100D Noise Figure?

[Helmut, DC6NY]

Hi Clyde and Phil,

Phil, of course your formula is correct, but I think you can also accept the
well known formula for cascaded stages and the contribution of noise
depending on the specific NF and gain/loss of the single stage:

F = F1 + F2 -1/G1 + 
..

According to this formula the contribution of noise by the ADC would be so
low or negligible for the whole system, if the supplier data for the ATT,
filter and preamp are correct. That is a contradiction to that what the
supplier of ADCs tell us. They recommend to calculate with noise figures of
30 to 40 dB.

I think my both setups , one with Mercuries and the other with Hermes are
well calibrated under fullscale condition of the ADC. Under your mentioned
conditions( 28 MHz, 2.9 kHz, Signal avg) I can read minimum noise levels in
the range of -127 dBm to -125 dBm (dither/random off and RX terminated
directly at the socket on the board). This is already a difference of at
least 3 dB to your figure.
I’m not really sure that we can trust the readings of the noise level
absolutely, when only a few bits of the ADC are applied, while calibration
was done with 16 bit. When my noise source gets repaired -hopefully soon -
I’ll make some measurements to get perhaps further insight.

73, Helmut, DC6NY


-----Ursprüngliche Nachricht-----
Von: hpsdr-bounces at lists.openhpsdr.org
[mailto:hpsdr-bounces at lists.openhpsdr.org] Im Auftrag von Phil Harman
Gesendet: Samstag, 1. Juni 2013 08:13
An: Clyde Washburn; hpsdr at lists.openhpsdr.org
Betreff: Re: [hpsdr] 100D Noise Figure?

***** High Performance Software Defined Radio Discussion List *****