[ARC5] BC-221 - drift and thermal stability. (also in receivers)

[Leslie Smith]

Hello List,
                      About thermal stability.
I don't have the experience Brian reports, but I can say, with great
certainty, that he is correct when he writes, "Thermal engineering can
be tricky."  I built several VFO's and plotted the result of freq. vs
temperature, and freq vs time (diurnal).  I did this over several days,
and left the window of my workshop open during winter, so I could
measure the temperature at 6AM - at the coldest time of day.  After that
I began stacking "cold bricks" around my VFO's and - after that -
recorded the drift vs temperature as the bricks melted.

I got "weird" shaped graphs at the extremes - where a temperature
reversal took place.
It took some time before I understood the cause for clover-leaf shapes
in my graphs.  Here is my understanding:  Each component has a
temperature coefficient, and the temperature gradient across each
component in a box is different.  This is why builders who report that
packing white "beans" into a VFO box does wonders for short term
stability.  

When Brian writes "there are all sorts of [thermal] pathways from hot
places ... " he is certainly correct.
This is what I measured in gear I built.  If you take Brian's words "can
be tricky" to mean "weird" - then he's correct!

As a corollary to this:  When some-one writes they built a V.F.O. that
was so stable it "stayed zero beat for hours" when left in an open box
near a window - I know they are 'messing with my mind'.  When some-one
writes that the Colpitts circuit or the Vackar circuit is very stable,
then I know they were lucky in choosing (at random) a set of components
that had some degree of corresponding thermal matching.  In other words,
they were lucky.  (Or - alternately - the fellow had no idea of the
meaning of stability, or how to measure it.)

My results - measurement in the drift of the BC-221 - may be reliable
(or not reliable).  I measured the frequency with a Fluke 1912-A
counter.  It was some decades old and not calibrated (to my knowledge.) 
I have no idea whether I measured drift in a Fluke 1912-A counter or in
the BC-221.   

At the same time I got some idea of the performance of respected gear
with respect to frequency.
I thought the result was worth the effort.  Maybe I'll build one or two
crystal oscillators, pack them in a box with beads and see what the
Fluke tells me about them.


-- 
73 de Les Smith
  vk2bcu at operamail.com

On Tue, Dec 1, 2015, at 13:06, Brian wrote:
> Hello Bob,
> 
> I applaud your efforts and your generosity in sharing your results.
> 
> Thermal engineering can be quite tricky. The thermal mass of these radios 
> you're testing is not just a single number. There are all sorts of
> pathways from hot places - sources - to sinks: via radiation, conduction and 
> convection. So, I would expect many differently shaped frequency vs 
> temperature over time curves during the course of a day. As you have
> already noticed, the relationship with time is not linear. I suspect you are
> using time as a substitute variable for temperature.
> 
> When I was designing oscillators, I attached RTDs all over the place and 
> tracked frequency vs temperature over very long time periods - like
> several 
> days. I came to the conclusion that there is no such thing as a uniform 
> change in frequency with temperature, nor a simple rise in temperature to
> a plateau. When I came to test ovenized frequency standards (and
> generators) -  doesn't that tell my age? I found some of the best, in terms of long-term 
> frequency stability vs ambient temperature, were in HP counters and
> Collins  synthesized sets, eg, PRC-47. At one stage I had a Marconi signal
> generator (2000 series?) that I set up one morning to zero-beat against WWV; this
> was in anticipation of a sale that evening. I left the generator on till
> later in the afternoon. After an hour, there had been no drift. I came back
> some 6 hours later; I listened for a zero beat. I could hear nothing. Initially,
> I thought it had drifted a long way off frequency. Then I turned the
> frequency dial a little and there was the tell-tale whoop - it had stayed on zero
> beat the whole time. I was sad to see it go. But it was a big bugger for which
> I > needed to do my knee-bend exercises for a week before lifting it.
> 
> You have chosen a particular order for your tests. What happens when you 
> revisit, say, the 80 m and 40 m measurements after the 10 m measurements? 
> What happens if you start with the 10 m measurements?
> 
> On Tuesday, December 01, 2015 6:35 AM, you said:
> 
> <snip>
> 
> Thank you kindly for the comments though and it's fun doing this stuff
> and allows me to putter aorund with my test gear.
> 
> 73  Bob, KE6F 
> 
> ______________________________________________________________
> ARC5 mailing list
> Home: http://mailman.qth.net/mailman/listinfo/arc5
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:ARC5 at mailman.qth.net
> 
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html

-- 
http://www.fastmail.com - Send your email first class