[hpsdr] New RF Digitizer

[Bob McGwier]

Ramakrishnan:

They are definitely getting processing gain. However,  if you want to 
talk about processing gain,  check out Mercury again.

It is a 16 bit part with 135 MHz sampling fed to Cyclone II for the 
digital downconversion.  SWEET!    So when thinking about 
Atlas+Ozy+Mercury remember that the signal path from A/D to USB 2.0 will 
have two cascaded Cyclone II's.  This will provide quite a bit of 
processing power.   The 16 bit part from Linear Technologies has a large 
real dynamic range based on both their own spec sheets and measurements 
and calculations Phil has done using the evaluation board.  If you look 
at the Mercury Wiki at Phil's set up for these tests,  it is easy to see 
that there was all sorts of potential for problems with the breadboard 
construction and still the Ozy/LTC2208 evaluation board performed 
brilliantly.     look forward to playing with my Mercury/Ozy/Atlas hardware!

While I am supportive of the RFSpace people and efforts,  they will not 
compete with the performance of the (not yet delivered) HPSDR hardware.  
What they do SPECTACULARLY well on is their software suite.  I wish it 
were open source ;-).   Moe and Peter and friends are really capable 
people and I hope they do really well with this stuff.

Bob




Ramakrishnan Muthukrishnan wrote:
> ***** High Performance Software Defined Radio Discussion List *****
>
> On 11/14/06, Ahti Aintila <oh2rz.sdr at gmail.com> wrote:
>   
>> Not bad and very compact! I hope that our heavy-weight champion
>> Atlas+Ozy+Mercury combination can put the dynamic range at least 20 dB
>> better. Am I asking too much?
>>     
>
> RFSpace guys may be compensating the 14-bit ADC dynamic range with
> over sampling which  is probably as good as gaining afew bits. They
> are using a 65 MSps part (AD6620 from the picture).
>
>   


-- 
Robert W. McGwier, Ph.D.
Center for Communications Research
805 Bunn Drive
Princeton, NJ 08540
(609)-924-4600
(sig required by employer)



 1163524391.0