[time-nuts] Morion MV89A position

[Poul-Henning Kamp]

In message <20140129123719.bILaS7m8 at smtp6.mail.yandex.net>, Charles Steinmetz w
rites:

>Warren pointed out that the MV89 has a double oven and said that this 
>makes added thermal capacitance unnecessary.

It's more complex than that.

A regular singleoven OCXO usually has a pretty high heating current
so it can regulate both up and down and therefore handle rather
brutal changes in ambient temperature/air-flow etc.

Double oven OCXOs, in particular "high-end" models, are usually
much better thermally insulated and therefore draw a lot less heating
current.

That is not a problem when they are exposed to sudden cooling, they
can regulate heating up as fast as they need.

But when they are exposed to sudden heating, they cannot regulate
the heating current negative.

I have seen this assymetry with a number of double oven OCXOs.

The best way to mitigate it, is to make sure the temperature does
not rise rapidly.

Unfortunately, that is almost the most common failure case:
A/C or local fans failing.

Wrapping the OCXO in thermal insulation is an option, but not
a good one, since it will drive the heating current even further
down.  Good idea for battery power though.

What you want is to wrap your OCXO in a thermal impedance.

The best result I have managed so far, was by wrapping the OCXO in
domesticated geology, (bricks, concrete, cinderblocks etc), which
has high-ish thermal capacity but only moderate thermal conductance.

But for some reason people stare incredously at you, if you request
4U rackspace for a cinderblock.

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.