[hpsdr] Odyssey Siren Design

Wed Sep 27 20:28:29 PDT 2006

Hello Ahti,

Thank you very much.  You've given me more to think about.

1. In some more research today, I think the only other fully differential
opamp that comes close to the noise figure of the OPA1632 is the THS413x.
However, it costs almost $2.00 more than the OPA1632.  The OPA1632 can go
lower in supply voltage (+/-2.5V) whereas the THS413x is 5V (advantage
OPA1632).

2. and 3. Now I get it.  To get the sampling pulses you need, there is no
other part.  Not even a combined part.  All the combined swithes that I
found the clocking is common to all the switches, and that won't work for
your circuit.

Finally, I'll try to use the SN74LVC2G53.  But I could also simply run the
SN74AUC2G53 at 2.7V.  It is only another regulator.  Your thoughts?

I think I have enough information now, I'll continue drawing the schematic
and post it to the list soon.

73,

- Steve, N7HPR
 (n7hpr at tapr.org)

> -----Original Message-----
> From: Ahti Aintila [mailto:oh2rz.sdr at gmail.com]
> Sent: Tuesday, September 26, 2006 9:43 PM
> To: Steven Bible
> Cc: hpsdr at hpsdr.org
> Subject: Re: [hpsdr] Odyssey Siren Design
>
>
> Hello Steve,
>
> and thank you for your valuable comment.
>
> 1. In some TI application note OPA1632 was suggested as a buffer
> before ADCs. According to my limited search the speed and noise
> specifications look really good, the 1/f noise is lower than most of
> the THS4XXX devices. Maybe, I missed some.
>
> 2. I wanted to use SPDT switch, as also suggested by Phil C, in order
> to use direct I and Q clocking without 1-to-4 decoder as used in the
> present SDR-1000 TRX board. Wider sampling pulses with proper
> impedance match, according to my understanding, yield lower losses in
> the integrating circuit. I admit that my supply voltage (3.3V)  may be
> really above the recommended 3.6V maximum with strong signals. By all
> means, use SN74LVC2G53 instead. That seems to be a good choice. For
> improved dynamic range I suggest 5V supply and ±5V (or more) for
> OPA1632.
>
> 3. Your "SN74CB3Q33253" must be a typo. Obviously you mean
> SN74CB3Q3253? I don't like 90° sampling pulses as mentioned above.
> With the 180° pulsing scheme the clock skew should not be a problem
> within I and Q channels individually and any skew errors between I and
> Q may be software corrected in the calibration process. Anyhow, if you
> can find a DPDT switch that would be great.
>
> 73, Ahti OH2RZ
>
>
> On 27/09/06, Steven Bible <n7hpr at tapr.org> wrote:
> > ***** High Performance Software Defined Radio Discussion List *****
> >
> > Hello Ahti,
> >
> > Joe and I have been studying your circuit for Project Odyssey.  I have a
> > couple of questions:
> >
> > 1. What criteria did you use in selecting the OPA1632?  I've
> been reading
> > the app notes that you recommended, and out of curiousity,
> searched the TI
> > web site for full differential op amps to see what else is
> available.  I was
> > surprised that TI's parametric search did not mention the
> OPA1632.  There
> > were many THS4xxx devices that popped up.  I am sure that low noise is a
> > major criteria.  I did not see any of the THS4xxx devices have
> any better
> > noise than the OPA1632.  I was a little concerned that the
> OPA1632 was an
> > expensive part (Digi-Key $3.94 qty 1), but when I saw the prices of the
> > THS4xxx devices, it suddenly because less expensive :-)
> >
> > 2. I was studying the data sheet for the SN74AUC2G53.  I noted
> that the Vcc
> > is to be 2.7V, but your schematic shows you are powering it at
> 3.3V? (the
> > absolute maximum is 3.6V).  2.7V is not a problem I think for
> this circuit,
> > but wanted to know your comments.  I also found the SN74LVC2G53
> can work at
> > higher voltage if needed.  I have not compared the data sheets
> yet.  I also
> > found that sourcing the SN74AUC2G53 to be a little more
> difficult than the
> > SN74LVC2G53.
> >
> > 3. I am a little concern with using two separate muxes and the possible
> > clock skew between them.  I am sure that using two separate
> devices makes
> > the PCB layout easier.  But I was thinking of using the
> SN74CB3Q33253 that
> > was mentioned on the list (or was it the SoftRock40 list) to minimize my
> > thinking of clock skew, but I wanted to get your thoughts.
> >
> > Thank you very much for sharing your circuit.  We hope to have
> it built in a
> > couple of weeks for testing for Project Odyssey.
> >
> > 73,
> >
> > - Steve, N7HPR
> >  (n7hpr at tapr.org)
> >
> >
> >
> >
> > > -----Original Message-----
> > > From: hpsdr-bounces at hpsdr.org
[mailto:hpsdr-bounces at hpsdr.org]On Behalf
> > Of Ahti Aintila
> > Sent: Thursday, September 21, 2006 1:30 AM
> > To: Joseph Julicher
> > Cc: hpsdr at hpsdr.org
> > Subject: Re: [hpsdr] Odyssey Siren Design
> >
> >
> > ***** High Performance Software Defined Radio Discussion List *****
> >
> > Joe,
> >
> > The bias arrangement of the opamp OPA1632 may not be the best
> > regarding the power consumption. Due to the difference between the
> > reference voltages of the signal switches and the opamps there will be
> > some extra current drawn, if Vcom is forced to 2.5V. For safety, I
> > recommend the common reference voltage 1.65V as corrected in the
> > version "b":
<http://kotisivu.dnainternet.net/ahti/sdr-1000/2rzisd-b.pdf>
> >
> > 73, Ahti OH2RZ
> >
> > On 20/09/06, Joseph Julicher <n9wxu at mac.com> wrote:
> > > very nice.
> > > That makes complete sense and answers all the issues.  I will use
> > > that in Siren.
> > >
> > > >> _______________________________________________
> > > >> 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/
>
> _______________________________________________
> 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/
>

 1159414109.0