Hi Andrea I have here a Cirkit  2nd edition Toko catalogue dated 1993 (a
firm local to me I dealt with quite a lot) The cat is in good condition but
it is 128 pages and a glued spine so scanning risks breaking it up. However
the 10k range occupies just one page and if the part adjacent to the spine
does not copy well all you will lose is the 100off price column (and I dont
think you can buy them for that now 20 years on :-))  )  Give me a little
time and I will do you a scan of the page, and mail it direct as a PDF.

I might actually have some new Toko coils in the component drawers but they
may not be 10K. However I also have a lot of 5MHz OCXOs including Racal an
Toyocom so I might have something useful if you dont find a source nearer to
you.

Best Wishes
Alan
G3NYK

----- Original Message -----
From: "Andrea Baldoni" erm1eaae7@ermione.com
To: time-nuts@febo.com
Sent: Tuesday, January 27, 2015 3:53 PM
Subject: [time-nuts] 5>10 doubler and old Toko RF catalogue (Cirkit 2nd
ed.1994)

Andrea wrote:

One thing to watch for is the 5MHz leakage component.  If you are
going to use the 10MHz standard for time-nuts experiments, the 5MHz
component needs to be WAY down (< -80dBc) or you will get funny
periodic ripples in stability plots.  Despite having two 5MHz traps,
one recently published design suppresses the 5MHz component only
about 52dB below the 10MHz output, and the 20MHz and 30MHz components
are also only -50 to -55dB.

For this reason (and some others, see discussions over the last
several months in the archives) I prefer a doubler built with a
quadrature hybrid coupler and a balanced mixer.  There is a write-up here:

http://www.ko4bb.com/manuals/download.php?file=02_GPS_Timing/4_App_Notes_and_Articles/Frequency_doubler_quadrature_DBM.pdf

I recently revived an old, stalled project to develop a JFET
push-push doubler for use at 5MHz (see schematic below).

FETs with very high transconductance and very small pinchoff voltage
(what a tube designer would call a "sharp cutoff" characteristic)
(e.g., 2SK369, BF862, etc.) are attractive on first look because they
can operate with lower conversion loss or even some conversion
gain.  However, they are not well suited for doubler duty for two
reasons: (i) their characteristics have a very short range of
2nd-order curvature, so in order to keep noise down they must be
driven into regions of higher-order distortion and therefore generate
lots of spurious energy; and (ii) they are devilishly hard to match
well enough to suppress the input frequency feedthrough.  Note that
you also need to put enough voltage on the FET drains to get them
well into the saturation region -- a Vcc of 5v is not enough.  Again,
the penalty is lots of spurious energy.  So, the lower conversion
loss of sharp-cutoff FETs is not the benefit it might at first appear
to be -- it is much easier to add gain after the doubler than to
remove unwanted spurious mixing products.

The design below uses medium-cutoff FETs and a Vcc of 15v (I found
that J111 and J310 work best and can be matched sufficiently with a
one-point match; 2N4416 and others also work, but are fussier and
would benefit from a 2- or 3-point match).  At an input of 500mVrms,
their long 2nd-order characteristic is used efficiently to generate
10MHz with relatively little spurious energy.

I had no problem finding one or more FET pairs matched to within 1mV,
given 20 devices from the same lot (YMMV).  With properly adjusted
traps at 5, 20, and 30MHz, all spurious responses were below
-80dBc.  The inductors can be commercial RF parts with Q of 200 or so
(I used some high-quality through-hole RF inductors I had on hand --
I doubt any SMD inductors will work).  The trap capacitors should be
C0G/NP0 ceramics for the bulk of the capacitance, plus very small
trimmers (I used 27pF, 27pF, and 100pF plus 0.2--6pF glass piston
trimmers).  I wound the two transformers on Mix-61 toroid cores (each
winding is 20 turns on a FT37-61 core -- the inductance is a little
lower than called out).  Mini-Circuits parts (or equivalents) may also work.

Best regards,

Charles

Using the square law characteristic will inevitably increase the phase noise floor particularly in the flicker region with respect to just using the switching characteristic of a JFET, diode or BJT  (non saturated).The only viable solution is to use better filtering of the output of a switching multiplier.If you intend to use a diode ring based mixer configuration diode connected (collector shorted to base) npns such as 2N222's are significantly quieter (as shown by NIST) than schottky diodes for frequencies below 40MHz or so.
Bruce.

 On Wednesday, 28 January 2015 8:25 AM, Charles Steinmetz <csteinmetz@yandex.com> wrote:

Andrea wrote:

One thing to watch for is the 5MHz leakage component.  If you are
going to use the 10MHz standard for time-nuts experiments, the 5MHz
component needs to be WAY down (< -80dBc) or you will get funny
periodic ripples in stability plots.  Despite having two 5MHz traps,
one recently published design suppresses the 5MHz component only
about 52dB below the 10MHz output, and the 20MHz and 30MHz components
are also only -50 to -55dB.

For this reason (and some others, see discussions over the last
several months in the archives) I prefer a doubler built with a
quadrature hybrid coupler and a balanced mixer.  There is a write-up here:

http://www.ko4bb.com/manuals/download.php?file=02_GPS_Timing/4_App_Notes_and_Articles/Frequency_doubler_quadrature_DBM.pdf

I recently revived an old, stalled project to develop a JFET
push-push doubler for use at 5MHz (see schematic below).

FETs with very high transconductance and very small pinchoff voltage
(what a tube designer would call a "sharp cutoff" characteristic)
(e.g., 2SK369, BF862, etc.) are attractive on first look because they
can operate with lower conversion loss or even some conversion
gain.  However, they are not well suited for doubler duty for two
reasons: (i) their characteristics have a very short range of
2nd-order curvature, so in order to keep noise down they must be
driven into regions of higher-order distortion and therefore generate
lots of spurious energy; and (ii) they are devilishly hard to match
well enough to suppress the input frequency feedthrough.  Note that
you also need to put enough voltage on the FET drains to get them
well into the saturation region -- a Vcc of 5v is not enough.  Again,
the penalty is lots of spurious energy.  So, the lower conversion
loss of sharp-cutoff FETs is not the benefit it might at first appear
to be -- it is much easier to add gain after the doubler than to
remove unwanted spurious mixing products.

The design below uses medium-cutoff FETs and a Vcc of 15v (I found
that J111 and J310 work best and can be matched sufficiently with a
one-point match; 2N4416 and others also work, but are fussier and
would benefit from a 2- or 3-point match).  At an input of 500mVrms,
their long 2nd-order characteristic is used efficiently to generate
10MHz with relatively little spurious energy.

I had no problem finding one or more FET pairs matched to within 1mV,
given 20 devices from the same lot (YMMV).  With properly adjusted
traps at 5, 20, and 30MHz, all spurious responses were below
-80dBc.  The inductors can be commercial RF parts with Q of 200 or so
(I used some high-quality through-hole RF inductors I had on hand --
I doubt any SMD inductors will work).  The trap capacitors should be
C0G/NP0 ceramics for the bulk of the capacitance, plus very small
trimmers (I used 27pF, 27pF, and 100pF plus 0.2--6pF glass piston
trimmers).  I wound the two transformers on Mix-61 toroid cores (each
winding is 20 turns on a FT37-61 core -- the inductance is a little
lower than called out).  Mini-Circuits parts (or equivalents) may also work.

Best regards,

Charles

time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.

On 1/27/2015 11:57 PM, Bruce Griffiths wrote:

What about filtering the doubler output with a 10 MHz xtal ?

73  Alberto  I2PHD

This email has been checked for viruses by Avast antivirus software.
http://www.avast.com

Am 28.01.2015 um 00:59 schrieb Alberto di Bene:

In a correctly designed oscillator, the close-in phase noise
is determined by the crystal only and it's generally accepted
that there are 20 dB differences from a single batch. Our
data support that. Only 10% are really good.

If you filter with a crystal that is in resonance it will impress
its own phase noise on the signal-to-be-filtered, and the odds
are that you lose.

Also, crystal dissipation should be only in the 100s of uW max.
Using it to filter a 10 dBm signal probably will do harm.

It is a different game when you want to notch away sub/harmonics.
The power levels of the harmonics are much lower and the
crystal does not resonate on the resonator frequency. If you
mis-tune a trap, it won't deliver its notch but it will leave all other
frequencies alone. A LC series circuit that produces a crater of
15 % center frequency could have more global effects if mis-aligned.

When playing with the notch filters for my doubler I got the idea
that one could use them to filter away sideband noise even at the
16 dBm level. One would have to be careful not to tune a notch
over the carrier, but stay away 100 Hz to some KHz. That might
bring an oscillator from "quite good" to "impressive" over a limited
offset range. I have made a filter board for a dozen notch crystals.
That could be populated with the 90% losers from oscillator production.

There is also an offset generator from JPL IIRC where they divide
100 MHz down to 10 KHz, subtract that from the 100 MHz with a
single sideband mixer and use the 99.99 MHz for a DMTD system.
Here one could use the notch filter array to get rid of the
residual carrier & wrong sideband.  DDS-free.

But I have no time to follow that in the moment. :-(

73, Gerhard, DK4XP

Gerhard wrote:

One problem with using crystals as traps (notch filters) is that the
series resistance of a crystal is several orders of magnitude higher
than that of a good series-resonant LC -- generally in the 50-100 ohm
range.  So, although the notch is very narrow, it will not be very
deep unless it is in a high-impedance circuit.  For example, in a 50
ohm circuit (50 ohms looking each way, so 25 ohms at the node) you
will be lucky to get 3dB of attenuation.  To get 40dB of suppression,
the nodal impedance would need to be at least 5k ohms, perhaps
even >10k ohms -- and the high impedance adds noise, which means
there is a phase noise penalty.  Another problem is that the narrow
notches are prone to sliding off frequency with small temperature changes.

Also, while a 5MHz trap crystal will almost certainly be a
fundamental-mode resonator, that will probably not be true at, say,
30MHz -- so a 30MHz trap would most likely have a notch at or near
the desired output frequency.

Made with good, high-Q RF inductors (forget SM parts), an LC trap is
generally preferable to a crystal trap.  There is still some
temperature sensitivity, but the greater width is much more
forgiving.  At the same time, the Q is high enough that you don't
have to worry about effects 5MHz away when you are trapping
frequencies of 30MHz and below.

Best regards,

Charles

On 1/28/2015 11:28 AM, Charles Steinmetz wrote:

It is very straightforward to use LC networks to transform the
impedance of the crystal to a much lower value and get around this
problem.

Rick Karlquist N6RK

And the narrow notch for the harmonic is not required anyway, since the
fundamental is fare enough, therefore a high Q  LC trap will work
better, also with the setting of the biasing af the active devices the
spures could be reduced to [ just observe the output with a spectrum
analyzer and set the bias of one site to minimum harmonics, there will
be no common optimum for all harmonics, but a good compromise could be
achieved ]
73
Alex KJ6UHN
On 1/29/2015 2:16 PM, Richard (Rick) Karlquist wrote:

On 1/29/2015 5:41 PM, Alexander Pummer wrote:

When I designed the 5071A RF chain, I used five cascaded frequency
doublers to go from 10 MHz to 320 MHz.  I definitely used traps
to reduce the 10, 30, and 40 MHz spurs (using 10->20 MHz as an
example).  It was no easy thing because I could only use coils of
moderate Q (less than 50) and I needed at least 80 dB suppression.  You
might wonder why I needed to reduce 40 MHz spurs in the 20 MHz
output.  It turns out (little known fact) that the if I drove
the 20->40 MHz doubler with 20 MHz contaminated with 40 MHz
harmonics, it would degrade the spectral purity of the 40 MHz
output.  Strange but true.

The 5071 filters are basically cascaded notch filters, as
opposed to band pass filters.  Doing this allowed me to
have zero adjustments.  Previous atomic clocks used narrow
high-Q filters that had to be tuned up, and were temperature
sensitive.  The production engineers had to constantly stay
on top of these filters because they were so temperamental.
OTOH, the 5071 filters just work.  There was never even a
production change to them AFAIK.  The key to getting the notch
filters to work was to use 2% components, and use two coils
and or two capacitors together to get around the fact that
the standard values are quantized to 10%.  Additionally, I
measured each tank circuit in situ on the PC board and tweaked
it to take into account the de facto board parasitics.

Rick Karlquist N6RK