[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Fw: [linrad] For general interest
Just wanted to say thank you one and all once again for some very
informed response to my question about where to place the phase
shift. It reminds me that termination of mixer ports has been the
subject of several letters in QEX in recent months where the use of
diplexers was recommended.
I had planned to buy Wes's new book now I have done so.
>----- Original Message -----
>From: "Pierre Vanhoucke" <pierre.vanhoucke@xxxxxxxxx>
>To: <josh-linrad@xxxxxxxxxxxx>
>Sent: Tuesday, August 19, 2003 1:07 AM
>Subject: Re: [linrad] For general interest
>
>
>> Hi Josh et all,
>>
>> I will try to summarize the article in a few lines:
>>
>> There is an argument against putting the phase shift in the RF path
>> because this path has to operate over a wide frequency range. In such
>> situation it is difficult to build a phase shift network that has an exact
>> 90° phase shift over whole range. This is not true for the local
>oscillator
>> if it is operating at a fixed frequency.
>>
>> On the other hand there is one reason that we might choose to use in-phase
>> LO and quadrature RF: The RF port of diode-ring mixers is often better
>> behaved than the LO ports: experimenters who builds their first phasing
>rig
>> are often amazed at how much difference an LO phase shift pair works when
>> connected to mixers than when it is observed with 50 ohms loads on an
>> oscilloscope. Often additional capacitors need to be tacked on the bottom
>of
>> the circuit board at one mixer LO port or the other.
>>
>> Conclusion: if the phasing receiver needs to operate at a single frequency
>> or over a very narrow band , the benefit of connecting the quadrature
>> network to the RF ports instead of the LO ports can outweight the bandwith
>> penalty.
>>
>> The article further describes different types of quadrature broadband
>> networks and the usage of digital circuits to provide accurate 90 °phase
>> shifts for the LO ports. If this digital approach for LO works better
>than
>> the quadrature networks , then the previous conclusion is of coarse
>reversed
>> . However at this moment there is apparently not enough well documented
>> information available about the succes of this digital approach, which
>> leeds the author to the conclusion that experiments in that area are
>highly
>> encouraged.
>>
>> Hope this was helpfull.
>>
>> I myself plan to carry out experiments in the future with both different
>> types of feeding the LO ports, and will keep you informed about the
>results.
>>
>> Best 73,
>>
>> Pierre /ON5GN
>>
>>
>> ----- Original Message -----
>> From: <josh-linrad@xxxxxxxxxxxx>
>> To: "Pierre Vanhoucke" <pierre.vanhoucke@xxxxxxxxx>
>> Sent: Monday, August 18, 2003 10:36 PM
>> Subject: Re: [linrad] For general interest
>>
>>
>> >
>> > Please post the answer to the list. I know I would like to know the
>> > answer, but do not have the book, and I probably will not get the book
>in
>> > the near future...
>> >
>> > 73, KD7HGL
>> >
>> > Later, JOSH
>> >
>> > On Mon, 18 Aug 2003, Pierre Vanhoucke wrote:
>> >
>> > > Hello Ron,
>> > >
>> > > An answer to your question can be found in the book 'Experimental
>> methods
>> > > in RF DESIGN' ( published by the ARRL ) in chapter 9.6 (page 9.24 to
>> 9.27).
>> > >
>> > > 73,
>> > > Pierre/ON5GN
>> > >
>> > > ----- Original Message -----
>> > > From: "W6WO" <ron-skelton@xxxxxxxxxxx>
>> > > To: <linrad@xxxxxxxxxxxxxxxxxxxxxx>
>> > > Sent: Monday, August 18, 2003 9:27 PM
>> > > Subject: [linrad] For general interest
>> > >
>> > >
>> > > > Here are a few sites that might be useful
>> > > >
>> > > > For comparing sound cards
>> > > > http://www.pcavtech.com/soundcards/compare/
>> > > > http://www.tracertek.com/comparisonchart.htm
>> > > >
>> > > > The Skyworks SKY73001-11 (http://www.skyworksinc.com/) is an
>> > > > interesting device designed specifically for direct conversion I-Q
>> > > > receivers. Brief specs are
>> > > > RF 2 -3500 MHz
>> > > > IF 0 -100 MHz
>> > > > IIP3 +26 dBm
>> > > > IIP2 +67 dBm
>> > > >
>> > > > I have been looking at the noise specifications of various DC
> > > > > regulators which could affect base-band noise levels, there is a
>wide
>> > > > range. In summary
>> > > >
>> > > > At the high end the LM 317 and 337 series have RMS noise levels of
>> > > > 0.003% of Vout. The spec states this is over 10 Hz to 10 KHz
>> > > >
>> > > > The best I could find from Motorola was the MC1723C that seems to
>> > > > have very low noise (2.5 uV) this is stated over the range 100 Hz to
>> > > > 10 kHz
>> > > >
>> > > > The MC78L/79L 100 m/A series have noise levels 5V/40 uVrms, 12V/75
>> > > > uVrms, 15V/90 uVrms stated to be in the 10-100 kHz range.
>> > > >
>> > > > It is reasonable to expect these levels could be reduced by Tantalum
>> > > > capacitors at the device output but it seems smart to select low
>> > > > noise devices to begin with.
>> > > >
>> > > > For further immunity the spec for the AD797 amplifier recommends 0.1
>> > > > uF and 4.7uF capacitors at the + and - supply pins with <5 mm leads
>> > > > to ground.
>> > > >
>> > > > Now a question.
>> > > > In most I-Q demod designs I note the phase shift is applied to the
>> > > > LO. In the simple 40 kHz 2x SBL-1 design that Leif published in QEX
>> > > > the shift is applied to the RF. What is the reason for this approach
>?
>> > > >
>> > > >
>> > > >
>> > > > --
>> > > > 73
>> > > > Ron
>> > > > W6WO
>> > > >
>> > >
>> > >
>> >
>> >
>>
--
73
Ron
W6WO
LINRADDARNIL