Message-ID: <00a001c401bb$0601a620$687a37c0@w2ksn>
From: "Stewart Nelson" <sn@scgroup.com>
To: rsgb_lf_group@blacksheep.org
References: <002601c3f93e$478b35a0$6507a8c0@Main> <001f01c3fb72$f5156750$687a37c0@w2ksn> <001901c3fb8e$2f2b4ac0$6507a8c0@Main>
Date: Wed, 3 Mar 2004 23:33:43 -0800
MIME-Version: 1.0
Subject: LF: Re: Re: Re: WOLF development
Content-Type: text/plain; charset=iso-8859-1; format=flowed
Content-Transfer-Encoding: 8bit
Sender: <owner-rsgb_lf_group@blacksheep.org>
Precedence: bulk
Reply-To: rsgb_lf_group@blacksheep.org

Hi Alan,

First, my apologies for taking so long to answer.

Second, my apologies for not understanding what you meant by
'keyed oscillator'.

I think that the LC oscillator has to be quite stable, unless
there is a good way to see and compensate the drift in software.
Suppose that the oscillator is gated on for 100 milliseconds,
and the software just averages the beat phase over that interval.
Then a 5 Hz oscillator offset would result in a 90 degree error
in the reference.  If the gate is on for only a few milliseconds,
that effect is minimized, but then the beep isn't much better
than what you get by feeding the 1 pps directly to the antenna input.

I once modified an RX for external locking, by using a PLL to
generate the same frequency as its reference crystal, and
injection locking via a gimmick wire glued to the RX PCB.
The good news is that the receiver would then still work with
the external box disconnected.  The bad news is that not all
LO frequencies came from this reference, so the stability was
improved, but still quite imperfect.

73,

Stewart


----- Original Message ----- 
From: "Alan Melia" <Alan.Melia@btinternet.com>
To: <rsgb_lf_group@blacksheep.org>
Sent: Wednesday, February 25, 2004 2:41 AM
Subject: LF: Re: Re: WOLF development


> Hi Stewart,.......answer below...sent direct
> ----- Original Message -----
> From: "Stewart Nelson" <sn@scgroup.com>
> To: <rsgb_lf_group@blacksheep.org>
> Sent: 25 February 2004 07:42
> Subject: LF: Re: WOLF development
> 
> 
> > Your keyed oscillator idea is interesting, but it appears to
> > introduce another unknown.  For example, if the 136 kHz oscillator
> > drifted up 0.01 Hz, it would affect the audio in the same way
> > as the receiver LO drifting down 0.01 Hz.  One way to resolve
> > this ambiguity would be to digitally divide the 136 kHz into
> > the audio range, and feed that into the sound card, too.
> > Do you have a simpler solution?
> >
> I suspect you have "forgotten" the properties of a keyed oscillator in this
> application. If the 136khz oscillator is stopped and re-started by the 1pps
> signal it will lock and not drift. I agree that if just the oscillator
> output is keyed  (so the whole comb drifts as well) then the osc drift will
> be a variable, but the keyed oscillator becomes a locked comb-generator.
> (Tube technology revisited !!) Thus it should be possible to construct this
> as an LC oscillator. I have not done the maths but I suspect that the
> smallscale sawtooth FSK created by the drift during the interval between
> locking pulses will "spread" the line a little, but averaging should be able
> to remove this satisfactorily.
> 
> My LF RXs (AOR 7030 s ) whilst stable (TCXO), have no facility to feed in an
> external standard. The TCXO is a "funny frequency" too.
> 
> Fascinating stuff.
> Cheers de Alan G3NYK
> alan.melia@btinternet.com
> 
> 
> 
>