You can't afford it !!!

[harry]

Thermal inductance...  I don't think so (or can't think of a case)

Your thermal capacitance is actually a time lag in thermal response
yes... which you are thinking of as capacitance ???

So what you are asking for is something that conducts heat 'faster'
but faster than what.

I think this is really like a servo loop problem... we would like the
feedback
as fast as possible, then we can have higher gain without oscillation ???

So you want a package that is electrically isolated but an absolute
excellent
thermal conductor.

Answer (I have studied this a LOT lately...) Carbon...

Carbon has 4 bonds.. it will share 4 electrons by giving or receiving them.
If you
have a matrix of ALL carbon atoms, each bonded perfectly to all of its
neighbors,
you have a tetrahedron shaped crystal lattice of Carbon.  It has the highest
thermal
conductivity (Carbon is a metal in this case....) known, more than twice
that of copper,
and yet it is a total dielectric (Carbon is a non-metal at the same time...)
because the
matrix totally lacks "holes" for electrons to move into.

This crystalline matrix is gonna cost you... its Diamond !!!

BTW I told some people from EUPEC that I wanted this material for a water
base for their high power semiconductors (1200A / 1200V IGBT). They laughed
at me... and said did I have any idea what that would cost ???

I laughed back. No problem, I said... Do you know how SMALL I expect you
to make your die ????

H^) harry  (see... I don't only geek you guys... I do this professionally
too)

BTW lets keep this going... If someone has a thermal inductor we COULD
improve things a lot !!!

Magnus Danielson wrote:

> From: jhaible at t-online.de (jh.)
> Subject: Re: Minimoog old style VCO
> Date: Fri, 29 Dec 2000 03:52:17 +0100
>
> > > When Patrik said this I was really supprised, I did not know of this
> > > "misstake". I mean, today we speak alot about thermal coupling, is
> > > this maybe the classic misstake from which everyone learned from?
> >
> > I learned it the hard way. When I built the Synthi Clone I foolishly
> > thought I could have a single tempco for all 3 VCOs (by feeding
> > the V/Oct CV into the "other" transistor's base (which is not
> > connected to the matrix). "Electronically" this would work, but
> > thermally it was a big mess.
>
> Another misstake... OUCH!
>
> It would have kind of worked if you used transitors out of the same
> CA3046 or something. That is, if the transistors where thermically
> well connected with each other AND the tempco. Break any of those
> thermical links and the assumption is out the window.
>
> Humm... the thermoelectric conductance property behaves like
> electrical conductors/resistos and thermoelectrical capacitance works
> kind of like electrical capacitors. I just wounder if there is an
> inductance equalence... you could possibly do much better thermical
> filters if they existed.
>
> When you do owen oscillators will thermical filtering be of
> interest.
>
> Cheers,
> Magnus