[time-nuts] 10811 EFC control
Charles P. Steinmetz
charles_steinmetz at lavabit.com
Thu Oct 28 17:27:32 UTC 2010
A while back, someone asked how to use the EFC on a 10811 to allow
fine frequency adjustment without spoiling the inherent stability of
the oscillator. I finally found a copy of the schematic for the
circuit I have used with success, which I'm pasting below in the hope
it will make it through the list server.
Pin 6 of a 10811A accepts an EFC voltage of +/- 5 V relative to
oscillator ground (pins 2, 4, and 5), for a total frequency
adjustment range of around 1 Hz. I believe the D version works the
same. Since I can generally get within a few mHz using the existing
trimmer and some patience, I find no need for anywhere near this much
range in the EFC. Accordingly, I shoot for +/- 0.2 to 0.5 V across
the trimmer, so that the tempco of the circuit has that much less
effect on stability and the adjustment has more resolution.
As drawn, the EFC supply starts out with a bridge so you can steal
just about any center-tapped AC in the instrument. I generally use
4700 uF filter capacitors. If you get to the point where the voltage
you stole is so high that dissipation in the 5 V regulators is a
problem, add some dropping resistors between the bridge and the
filter caps. Some instruments will have +/- DC supplies you can
steal. Not shown are the regulator output capacitors -- I use a
tantalum and a multilayer ceramic to ground on each regulator output.
The LEDs are standard red or UV parts, and the diodes in series with
the adjustment pot are 1N4148 or similar (any silicon signal diode,
really). Together, these provide first-order temperature
compensation for the EFC voltage. The LEDs and diodes should be
reasonably isothermal, but there is no need to go nuts here. I
typically choose the resistors at the regulator outputs for 10-15 mA
through the LEDs, and the potentiometer string for around 1 mA with
about +/- 0.2 to 0.5 V across the pot. The output filter should have
as low a frequency corner as you can arrange. I usually use 100 k
and a 5 uF film capacitor for a corner around 0.35 Hz. If you offset
the frequency with the internal trimmer, you will know which polarity
of EFC you need and therefore can use a larger, polar cap for a lower
noise corner -- but I have never done that. A 10-turn pot has more
than enough resolution, but I usually use a 20 or 25-turn anyway.
Note that pin 6 is grounded external to the 10811 in instruments that
do not use the EFC, generally on the instrument's motherboard (or an
intermediate daughter board, depending on the particular
instrument). You will have to determine how it is done in your
instrument and undo it for the EFC to work.
Finally, install the EFC fine frequency control where you can adjust
it from the top, so you don't have to turn the instrument on its side
or top to make the adjustment.
-------------- next part --------------
A non-text attachment was scrubbed...
Size: 48318 bytes
Desc: not available
More information about the time-nuts