[hpsdr] HORTON

[Alex]

Here's an off the wall question. Why not merge the MERCURY and HORTON designs? 
It's just a software matter to decide how much data is sent onwards to the host computer.
Another mitigating factor might be how much data can SASQUATCH handle?

Just trying to get my mind around the details of the difference between the two receiver cards.

Alex, N3NP

----- Original Message ----
From: Robert McGwier <rwmcgwier at comcast.net>
To: Philip Covington <p.covington at gmail.com>
Cc: Alex <harvilchuck at yahoo.com>; hpsdr at hpsdr.org
Sent: Monday, May 22, 2006 5:21:14 PM
Subject: Re: [hpsdr] HORTON

I second this.  There is some hope that we can get very good performance 
with the mercury board through the processing gain associated with 
filtering and decimation from 135 e+6 samples per second down to (say) 8 
e 3 samples per second.  That is a very large theoretical (emphasis on 
theory) processing gain and must be accompanied by actual filters in 
front of the A/D's.  Note: blocking dynamic range will be determined by 
the instantaneous dynamic range of the A/D's so it is a little out of 
whack with "normal receivers" if anything we have been doing in the last 
few years with all of this is "normal".  Never mind,  it will be a very 
nice receiver indeed to have  larger than 100 dB dynamic range while 
processing the entirety of 20 meters!  With some care on parts selection 
to get parts that exceed the performance of the QSD in the SDR-1000, we 
can certainly exceed is performance with Horton.  Most of the 
degradation we have seen in the SDR-1000 QSD with increasing frequency, 
I believe, is due to aperture jitter induced by parts and layout and not 
by theoretical shortcomings of the  QSD.  This was overcome in the 
SDR-1000 by using a preamp in front of the QSD on the "RFE" board with 
sufficient gain for it to determine the system noise figure as some cost 
in IP3 and IMD DR3.

My brain is still a little fuzzy from getting home at 3:30 AM! and 
getting up for work at 7 AM.  I know it is fuzzy because I think all of 
that work at Dayton was fun.

Bob
N4HY




Philip Covington wrote:
> ***** High Performance Software Defined Radio Discussion List *****
>
> On 5/19/06, Alex <harvilchuck at yahoo.com> wrote:
>   
>> ***** High Performance Software Defined Radio Discussion List *****
>>
>> I see on the wiki that there's some consideration for use of the AD7760 and the ADG901.
>> I thought the LTC2208 was the current thought?
>>
>> Plus why not use Analog Devices' Demodulators (same family as the ones I'm planning on for CASMIR)?
>>
>> Alex, N3NP
>>     
>
> For narrow bandwidth applications the QSD will be very hard to beat
> performance wise.   The Analog Devices Demodulators will come nowhere
> close to a properly designed QSD. Mercury will be LTC2208/2209 based.
>
> 73 de Phil N8VB
>   


-- 
AMSAT VP Engineering. Member: ARRL, AMSAT-DL, TAPR, Packrats,
NJQRP/AMQRP, QRP ARCI, QCWA, FRC. ARRL SDR Wrk Grp Chairman
Laziness is the number one inspiration for ingenuity.  Guilty as charged!





 1148348500.0