In article , "Tam/WB2TT"
writes:

I don't know what you mean by ordinary crystal, but even.001% is not very
good. You may want to look at high accuracy temperature compensated
oscillators. Depending on where you live, you may actually be able to phase
lock your oscillator to the WWB carrier. You would have to use a phase
detector that does not pull you off to one side if you lose the signal.
Also, the ~3.58MHz color subcarrier on some TV stations is locked to a
cesium standard.

"Ordinary" quartz crystals are available off-the-shelf with tolerances
ranging from 20 ppm to 200 ppm. That equates to 0.002 to 0.020%.

TCXOs (Temperature Compensated Xtal Oscillators) will have much
better tolerance over temperature -but- the oscillator circuit itself must
be trimmed for the nominal temperature frequency.

If the operating environment temperature has a "room temperature"
range, then it might be better to use a VCXO (Voltage-Controlled
X O) enclosed in an insulated container. The voltage control (from a
very hum-free and very stable source) allows trimming down to better
than 1 ppm with the least disturbance of the temperature-insulating
properties of the enclosure...no tuning trimming shafts nor access
holes for intruding alignment tweakers needed.

There are some TCVXOs available stock at specific frequencies but
those cost a bit more than the average crystal oscillator unit. ECS
made a few of those which were dealt through DigiKey, but those
aren't in their current industrial on-line catalog.

In the NTSC TV standards, the color sub-carrier reference frequency
of 3.579454545454545... MHz is related to the 5 MHz NIST standard
frequency by: (88 x Cs) / 63 = 5 MHz, were Cs is the sub-carrier f.
Integer dividers and multipliers can create one from the other.

The NTSC aural carrier frequency offset is exactly 4.5 MHz and that
can be found from the color sub-carrier frequency by: (44 Cs)/35 =
4.5 MHz.

To compare one crystal source against another, a color sub-carrier
oscillator can be divided down by 63 to achieve 56.1818181818.. KHz
and a 5 MHz oscillator divided by 88 to get the same frequency. The
divided frequencies can be input to a phase detector whose output
can be used for direct phase comparison or for a PLL that locks one
of the oscillators to the other. Integer flip-flop dividers can be:

88 = 2 x 2 x 2 x 11 63 = 7 x 9

Len Anderson
retired (from regular hours) electronic engineer person

In article ,
(Avery Fineman) writes, adding to a previous post:

In article , "Tam/WB2TT"
writes:

I don't know what you mean by ordinary crystal, but even.001% is not very
good. You may want to look at high accuracy temperature compensated
oscillators. Depending on where you live, you may actually be able to phase
lock your oscillator to the WWB carrier. You would have to use a phase
detector that does not pull you off to one side if you lose the signal.
Also, the ~3.58MHz color subcarrier on some TV stations is locked to a
cesium standard.

In the NTSC TV standards, the color sub-carrier reference frequency
of 3.579454545454545... MHz is related to the 5 MHz NIST standard
frequency by: (88 x Cs) / 63 = 5 MHz, were Cs is the sub-carrier f.
Integer dividers and multipliers can create one from the other.

The NTSC aural carrier frequency offset is exactly 4.5 MHz and that
can be found from the color sub-carrier frequency by: (44 Cs)/35 =
4.5 MHz.

Using a "3.58 MHz" color TV crystal, it's possible to use 17 flip-flops
in a set and reset depending on states of stages 4, 16, and 17 to
achieve a 60, 50, or 100 Hz precise output. That's done in the old
National Semiconductor MM5369 divider, a little 8-pin CMOS DIP
that runs on 5 to 15 VDC single supply, crystal oscillator active
circuitry on-board. If one can still find them, the following suffixes
will produce the following outputs with a colorburst crystal:

MM5369AA - 60 Hz
MM5369EYR - 100 Hz
MM5369EST - 50 Hz

Couldn't find the 20+ year old notes on the division presetting so all
I've got for those is a 1980 National MOS Databook (which doesn't
say squat about the divider chain's details). It's an interesting game
on paper with a calculator to figure out the exact division value. :-)

That little oscillator-divider was used in several Russian "Woodpecker"
(over the horizon radar) blanker add-ons for receiver two decades ago..
Some additional decade dividers coming off the output would get down
to the Woodpecker's impulse rate.

Len Anderson
retired (from regular hours) electronic engineer person

In article ,
(Avery Fineman) writes, adding to a previous post:

In article , "Tam/WB2TT"
writes:

I don't know what you mean by ordinary crystal, but even.001% is not very
good. You may want to look at high accuracy temperature compensated
oscillators. Depending on where you live, you may actually be able to phase
lock your oscillator to the WWB carrier. You would have to use a phase
detector that does not pull you off to one side if you lose the signal.
Also, the ~3.58MHz color subcarrier on some TV stations is locked to a
cesium standard.

In the NTSC TV standards, the color sub-carrier reference frequency
of 3.579454545454545... MHz is related to the 5 MHz NIST standard
frequency by: (88 x Cs) / 63 = 5 MHz, were Cs is the sub-carrier f.
Integer dividers and multipliers can create one from the other.

The NTSC aural carrier frequency offset is exactly 4.5 MHz and that
can be found from the color sub-carrier frequency by: (44 Cs)/35 =
4.5 MHz.

Using a "3.58 MHz" color TV crystal, it's possible to use 17 flip-flops
in a set and reset depending on states of stages 4, 16, and 17 to
achieve a 60, 50, or 100 Hz precise output. That's done in the old
National Semiconductor MM5369 divider, a little 8-pin CMOS DIP
that runs on 5 to 15 VDC single supply, crystal oscillator active
circuitry on-board. If one can still find them, the following suffixes
will produce the following outputs with a colorburst crystal:

MM5369AA - 60 Hz
MM5369EYR - 100 Hz
MM5369EST - 50 Hz

Couldn't find the 20+ year old notes on the division presetting so all
I've got for those is a 1980 National MOS Databook (which doesn't
say squat about the divider chain's details). It's an interesting game
on paper with a calculator to figure out the exact division value. :-)

That little oscillator-divider was used in several Russian "Woodpecker"
(over the horizon radar) blanker add-ons for receiver two decades ago..
Some additional decade dividers coming off the output would get down
to the Woodpecker's impulse rate.

Len Anderson
retired (from regular hours) electronic engineer person