[time-nuts] Designing and building an OCXO and GPSDO

[Bill Hawkins]

Phil,

Good on ya. When I have the time, I'll try going back to first
principles.

I've spent 40 years working with PID controllers and their sensors and
actuators.

Rule of thumb: If the transport delay (dead time between changing the
heater and sensing the change) is 10% of the system time constant, you
have to reduce the PID gain by about half to maintain stability (of the
phase margin in a feedback loop).

In your case, do NOT mount the sensor on the crystal. Mount it halfway
between your heaters and expect the copper plate to come to a
steady-state temperature inside the insulated cavity. Do not mount it
with any form of insulating adhesive which will increase the delay.
Maybe hold it down with a copper strip, like the crystal.

As Bruce Griffiths suggests, use an analog sensor and a resistance
bridge. Then the null signal can be digitized with low accuracy
requirements, compared to digitizing the temperature to 10E-9. Mount the
analog stuff in the controlled enclosure. Use fixed resistors to balance
the bridge at 50C so there are no adjustments in the oven. Plan on
putting the oscillator in there, too.

Never mind being dead on 50C because where you want to be is on a flat
spot on the temperature curve, so maybe you don't have to hold 10E-9.
Your digital PID is looking at the digitized bridge error as its
measurement. A setpoint can vary that error to find the flat spot. Don't
forget that components and solder joints age at 50C, so plan on making
adjustments once a week until the frequency stabilizes.

The accuracy of the sensor is not important. What you want is precision
and stability. You want the highest gain you can get in the PID, which
means doing the stuff for delay as well as determining the time constant
of your oven at 50C so you can set the integral term.

Bill Hawkins


-----Original Message-----
From: Philip Pemberton
Sent: Sunday, August 10, 2008 4:45 PM

   What I'm planning to do is mount the crystal on a copper plate with
two power transistors, using heatsink compound between the copper and
transistors/crystal case, and fit a temperature sensor to the top side
of the crystal case. I'm planning to use a copper bracket to hold the
sensor onto the crystal, and in turn mount the crystal to the copper
base.

   As far as temperature regulation goes, I'm going to use a PIC
microcontroller (one of the 8-pin chips with an A/D converter) to
monitor the temperature of the crystal, and use a PID loop to control
the two power transistors to maintain a temperature of 50C +/- 2 Celsius
(the accuracy spec of the temperature sensor). I also have other
higher-accuracy sensors (Dallas DS18S20 and DS18B20) that I can
calibrate with; these are accurate to around half a degree Celsius with
a resolution of 0.5C.