[hpsdr] Status of Hermes Gigabit

[Kok Chen]

On May 2, 2015, at 11:43 AM, Chuck Hutton wrote:

> You really didn't address the main point - would you agree that 2 MSPS 24 bit IQ data is in need of GigE?

I was simply pointing out that with today's firmware, the I/Q component is already 24 bits wide.   There is no need to increase the width of a sample any further, even when a better ADC appears.  And as Helmut mentioned, there is already plenty of dynamic range from the current ADC.   In fact, for RTTY and PSK31 signals, the dynamic range at the demodulator is better than 135 dB; the more conservative 125 dB number that we use for Hermes is for the 500 Hz "ARRL bandwidth."

(Indeed, the only time you need to increase the resolution of an I/Q sample to maintain a constant ARRL dynamic range, is when the sampling rate is *slower*!)

Since you asked, I write my own Mac OS X desktop program to process the I/Q data from Hermes/Angelia, and the current 100 Mb/s suits me just fine.  For my needs (investigating statistical demodulators for digital modes), I have no need for more than 192 ks/s rate from the hardware.  I use 96 ks/s almost 100% of the time. 

But that is just me.  If you are writing your own desktop program that requires a 2 Ms/s I/Q signal, then you will need to modify the Hermes/Angelia firmware yourself too.

73
Chen, W7AY


 1430599283.0