[time-nuts] Improving ocxo temp control

[Bill Hawkins]

My experience with industrial temperature control says there is always a
time lag between applying power to the heater and raising the
temperature at the thermistor. 

Because there is a lag, there is a gain beyond which the system
oscillates.

But I haven't read the non-referenced papers, so I could be wrong.

Bill Hawkins


-----Original Message-----
From: time-nuts [mailto:time-nuts-bounces at febo.com] On Behalf Of Bob
kb8tq
Sent: Friday, May 18, 2018 9:46 PM

Hi

There are a number of papers out and about about the limits on OCXO
performance.
The bottom line is that coming up with a high resolution control circuit
is the easy part of the task.

Simple answer to the question: 

Set up a thermistor bridge and feed the difference into an op amp. Crank
up the gain on the op amp to whatever you feel comfortable running. 

Some simple numbers: 

Thermistor changes 3% / C
Single thermistor bridge changes 1.5% / C Output of the circuit will
change the oven by 150C from power off to full on Neglecting the scale
factors for simplicity Put in a gain of 100 on the op amp

So, the bridge moves 1% and the controller goes from full off to full
on. 
Crank in more gain "as required". The op amp isn't bothered until you
get into the millions.  

With the simplified numbers above, the circuit has a thermal gain of
150. Getting much past 300 with a single oven is unusual. 

Bob

> On May 18, 2018, at 2:03 PM, Gilles Clement
<clemgill at club-internet.fr> wrote:
> 
> Hi,
> I am trying to improve performance of an OCXO.
> Could you point me at a good design of a high resolution oven
temperature controler please ? Preferably analog.
> Thx much,
> Gilles.