[time-nuts] time-nuts Digest, Vol 127, Issue 5

Wed Feb 4 20:23:18 UTC 2015

Hi Bill --

I've been quietly observing this thread because I don't have the 
knowledge to talk about the doubler circuit generally.

But the crystal filter is, in my opinion, a really bad idea.  Any 
resonator is ultimately a thermometer, and a high-Q device on the output 
of the circuit is going to make it suffer badly from temperature-related 
phase shift.  After all, that's why we put crystal oscillators in 
(sometimes double) ovens.

When I designed the TADD-1 distribution amp, I allowed for an optional 
LC bandpass filter to be fitted, and I've regretted that ever since. 
Even a relatively low-Q BPF made the tempco of the board orders of 
magnitude worse.

Notch filters for the problem frequencies are a much better idea.

73,
John
----

On 2/4/2015 2:45 PM, Fuqua, Bill L wrote:
>   OOPS. what was I thinking,
>    I was trying to minimize the number of transformers and
> slipped up. The inputs should be in opposite phase and
> output in parallel with a bifiler wound toroid transformer to
> provide balanced outputs, one to the crystal and one to the
> neutralizing (phasing) capacitor. The advantage of the crystal filter
> is that it provides much higher Q and to some degree, depending
> on it's selectivity reduces sideband noise. The crystals are cheap
> and if you like you can add more stages. No tuning required except for
> the phasing capacitor. If you happen to have a source of precisely
> cut series resonant 10MHz crystals you could easily go to 100Hz
> bandpass or even less by using a small value  loading resistor.
>    A ladder filter could be used  but there is still coupling far off
> resonance thru a capacitive ladder network consisting of the
> crystal holders' capacitances and discrete capacitors.
>    An complex LC filter could be constructed but it requires
> a number of stages along with careful selection of components.
> 73
> Bill wa4lav
>
>
> Message: 4
> Date: Wed, 04 Feb 2015 06:57:48 -0500
> From: Charles Steinmetz <csteinmetz at yandex.com>
> To: Discussion of precise time and frequency measurement
>          <time-nuts at febo.com>
> Subject: Re: [time-nuts] 5>10 doubler
> Message-ID: <20150204145756.vsAmCHQL at smtp2m.mail.yandex.net>
> Content-Type: text/plain; charset="utf-8"; Format="flowed"
>
> Bill wrote:
>
>> Push-Push Jfet amplifier with parallel inputs and a Toroid output
>> transformer, no secondary along with a simple filter using a 10 MHz
>> series resonate crystal connected to one drain and an adjustable
>> capacitor connected to the other would work fine. You connect the
>> other ends of the two together and a loading resistor to
>> ground.  The capacitor is used to neutralize or null out the shunt
>> capacitance of the crystal so that a capacitive path for the other
>> frequencies , 5, 15, 20, etc is eliminated.
>
> I concur with what Bruce said regarding crystal filters (or any
> narrow bandpass filter) at the output frequency.
>
> More fundamentally, I'm not sure I understand your description of the
> circuit.  You say it is a pair of FETs with parallel input and a
> transformer (autoformer) output.  To me, that suggests the circuit
> pictured below (one feeds the sources in parallel, the other feeds
> the gates in parallel -- it doesn't make any difference in how the
> circuit operates).
>
> The usual push-push doubler feeds the FETs differentially, and takes
> the common-mode output.  The diagrammed circuit reverses this -- it
> feeds the FETs in parallel (common-mode) and takes the 10MHz output
> differentially.  As drawn, the circuit would have essentially no
> output at the input frequency or any of its harmonics (only that due
> to the mismatch between the FETs).  The only signals it would amplify
> are uncorrelated signals -- i.e., the FETs' input noise voltages.  A
> quick simulation confirmed no significant output at the input
> frequency or its harmonics for matched FETs.  Simulating mismatched
> FETs produced a 5MHz signal rich in harmonics, but at a very low
> level and with no suppression of the 5MHz and its odd harmonics.
>
> I assume I misinterpreted your description and that you had a
> different circuit in mind, or that if you did have this circuit in
> mind I'm missing something about its operation.  Can you please
> describe again what you had in mind, and how it generates 10MHz?
>
> Best regards,
>
> Charles
>
> -------------- next part --------------
> A non-text attachment was scrubbed...
> Name: Fuqua_doubler_sm.gif
> Type: image/gif
> Size: 17911 bytes
> Desc: not available
> URL: <http://www.febo.com/pipermail/time-nuts/attachments/20150204/ed4fa496/attachment.gif>
>
> ------------------------------
>
> Subject: Digest Footer
>
> _______________________________________________
> time-nuts mailing list
> time-nuts at febo.com
> https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
>
> ------------------------------
>
> End of time-nuts Digest, Vol 127, Issue 5
> *****************************************
> _______________________________________________
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.
>