[Amps] Power-off tube cooling

[Mike Hyder --N4NT--]

Oh, and the question is not "will the 0º end reach a dangerous temperature?"
The question is: "will the rising temperature of that end coupled with the
falling temperature of the glass enclosure create trouble?"  It might.
Glass conducts heat much less well than metal does, so the glass which
contacts the metal rod will become hotter while the rest of the envelope
cools.

We used to cut Coke bottles in half by heating a ring around them with the
friction of a string, then dipping them into water.  This was no doubt
caused by the temperature difference we caused.  If you can cut a Coke
bottle in this way, then temperature differences in the glass around the
seal could be equally catastrophic, I would think.

Mike N4NT

----- Original Message ----- 
From: "Bill Turner" <wrt at dslextreme.com>
To: "Mike Hyder --N4NT--" <N4NT_Mike.Hyder at charter.net>
Cc: <RFlabnotes at aol.com>; <amps at contesting.com>
Sent: Saturday, January 17, 2004 11:19 AM
Subject: Re: [Amps] Power-off tube cooling


> On Sat, 17 Jan 2004 09:52:42 -0500, Mike Hyder  --N4NT-- wrote:
>
> >The ball is still
> >2000º so the gradient along the supporting rod now changes as the
supported
> >end's temperature rises until the temperature in the rod reaches
> >equilibrium: 2000º at both ends.
>
> _________________________________________________________
>
> Yes, but when power is removed (the original question) the 2000º
> immediately starts dropping too.  The question is will it drop faster
> than the 0º end rises?  Or more specifically, will it drop fast enough
> to keep the 0º end from reaching a dangerous temperature?  It depends on
> the thermal conductivity of the wire, the thermal masses at both ends
> and the amount of radiation.
>
> In other words, there is no universal answer.  In a case like this, I
> would assume the tube manufacturer has done their homework and go with
> their recommendations.  If there is no mention of a cool-down period I
> would not be concerned, but if there is, one would be foolish to ignore
> it.
>
> --
> Bill, W6WRT
>