[time-nuts] High accuracy temp controller ckt

[Dr. David Kirkby]

On Fri, 12 Jul 2019 at 08:07, Bernd Neubig <BNeubig at t-online.de> wrote:

> AXTAL has a miniature oven for heating precision resistors or other
> components to a constant temperature. It is housed in a small DIP8 package.
> See attached AXR135 data sheet.
> Optionally this device can be offered unsealed, ready for inserting the
> component by yourself
>
> Regards
> Bernd


The resistors I am using are made by Caddock, are 1 M ohm & 10 M ohm, 0.01%
tolerance and 5 ppm / deg C. They are much too large to fit in a DIP8
package. A quick check with a ruler indicates that the 10 M ohm is 38 x 12
mm and the 1 M ohm is 19 x 10 mm.

A quick photograph shows this half finished project. Unfortunately, even
before I have finished it, I am wishing I had done a better job. However,
my original reason for doing this was to check the stability of a data
acquisition unit before sending it to Keysight. I think this will be good
enough for that, but I could certainly have done a job with some thought.

The thermocouple is just meant to be a temperature sensor that dissipates
no power. A thermistor or RTD buried in the polystyrene would have just got
hot. The resistors should not get hot, as the power dissipated will be
under 1 uW.

I also intend putting in a 100 M ohm resistor, but the specification of
that is much poorer (5%, 100 ppm/ deg C. )

I partially read the paper mentioned. I note that the authors used a thermo
electric cooler (TEC) as they wanted get low temperatures. I assume that
for crystals or resistors dissipating little heat, a TEC would be
unnecessary, and just a resistor acting as a heater would be fine. But will
read some of the references. Obviously for controlling a crystal
oscillator, or my resistor, stability is most important, whereas for the
authors of the paper, stability was not their major concern.

> --
Dr. David Kirkby,
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image_6546160.JPG
Type: image/jpg
Size: 1736054 bytes
Desc: not available
URL: <http://febo.com/pipermail/time-nuts_lists.febo.com/attachments/20190712/08652ee5/attachment.jpg>