[hpsdr] Odyssey Siren Design

[Bob McGwier]

Ahti:

The final piece of the puzzle that fit this together for Phil C, you, 
etc.  and led to my laplace transform analysis of your and Phil C's 
original proposals was Dan Tayloe suggesting that the QSD should be 
viewed as an integrator and that the capacitors in his unbalanced design 
and in Gerald's balanced QSD design were only rough approximations to 
it.   The dominant terms in the Laplace transform analysis (the perfect 
tool for both transient and longer term response) turned out to be  
........  AN INTEGRATOR with a twist.  Your circuit has the integrator 
with the twist.    Dan's comment was the first time I went "AHA!" and 
was the last piece of the mental puzzle for me personally.   It was the 
first time I had a Gestalt of the actual underlying theory of operation 
instead of just plugging things into some simulator and fiddling around 
until things looked better.    If necessary,  I can dig out the very old 
note to the softrock group where Dan makes this statement.

I am certain in my own mind that you and Phil have gotten this to the 
point where this can proceed to alpha test this and get circuit boards 
out for full scale analysis.   Horton, Phoenix, whatever,  needs to have 
this circuit closely coupled, gain/impedance matched to the follow on 
A/D for noise immunity and to get the promise this circuit holds for 
dynamic receiver performance.  I suggest that this is the next step.

Congrats,
Bob
N4HY



Ahti Aintila wrote:

>***** High Performance Software Defined Radio Discussion List *****
>
>Hi Joseph and Lyle,
>
>Excellent ideas! Anyhow, please note that LMH6624 has a rather high
>1/f-noise corner (10 kHz).
>
>You possibly may find also my modification of the Active Integrating
>Quadrature Sampling Detector concept interesting (ISD for short). This
>schematic is modified for +3.3V and +5V supply voltages:
><http://kotisivu.dnainternet.net/ahti/sdr-1000/2rzisd.pdf> To increase
>the blocking range with high level signals, you may need to adjust R13
>for minimal commom mode DC-output. I have not tested this circuit, but
>here are my measurements with another circuit:
><http://kotisivu.dnainternet.net/ahti/sdr-1000/AIQSDb.pdf>
>
>See also Ray Anderson's write-up on Phoenix pages.
>
>73, Ahti OH2RZ
>
>On 20/09/06, Lyle Johnson <kk7p at wavecable.com> wrote:
>  
>
>>***** High Performance Software Defined Radio Discussion List *****
>>
>>    
>>
>>>I am about 80% finished in the schematic entry on Siren and ran into an
>>>interesting design idea that may help simplify the power supply for use
>>>in portable (or extraterrestrial) designs.
>>>
>>>We are using the TI TLV320AIC31 Codec which has differential inputs and
>>>outputs.
>>>The I/Q demodulator is the standard QSD circuit.
>>>To improve the S/N ratio a good instrumentation amplifier is typically
>>>used after the QSD commutator and before the CODEC.
>>>However, there are two problems:
>>>1) the INAMP requires a bipolar power supply, and
>>>2) the TI CODEC has a differential input and the INAMP will convert the
>>>differential QSD signal to a single ended signal.
>>>
>>>I am proposing the following solution:
>>>Put a low noise 10x buffer on each line beween the codec and the qsd.
>>>The LMH6624 is a low noise opamp from National.  At 1MHz it has
>>>0.92nv/sqrt(hz) noise.  That is quite low.
>>>http://www.national.com/pf/LM/LMH6624.html
>>>
>>>The + and - signals are directly fed to the CODEC.  The circuit shown
>>>provides the following transfer function.
>>>
>>>Vout = (Vin - 1.65)10 + 1.65
>>>      
>>>
>>It will only have an output swing of 2V (+/- 1V about the center) with a
>>single 5V supply, and each amplifier consumes 12 to 18 mA.
>>
>>A Linear Tech LT6231 is 1.1 nV-rt-Hz (at 10 kHz), stable at unity gain,
>>draws 3.5 mA/amplifier, rail-to-rail output.
>>
>>Anyway, I like the idea of a low-noise buffer to reduce the equivalent
>>noise figure of the ADC input!
>>
>>73,
>>
>>Lyle KK7P
>>
>>_______________________________________________
>>HPSDR Discussion List
>>To post msg: hpsdr at hpsdr.org
>>Subscription help: http://lists.hpsdr.org/listinfo.cgi/hpsdr-hpsdr.org
>>HPSDR web page: http://hpsdr.org
>>Archives: http://lists.hpsdr.org/pipermail/hpsdr-hpsdr.org/
>>
>>    
>>
>_______________________________________________
>HPSDR Discussion List
>To post msg: hpsdr at hpsdr.org
>Subscription help: http://lists.hpsdr.org/listinfo.cgi/hpsdr-hpsdr.org
>HPSDR web page: http://hpsdr.org
>Archives: http://lists.hpsdr.org/pipermail/hpsdr-hpsdr.org/
>
>  
>


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



 1158768023.0