Hi Gary,

There's also a slightly different approach you might want to consider, 
which is what I now think they did in the STS-1 and probably others.

Heat treat the Ni-Span-C for maximum strength (~1300 deg F/5 hrs) which 
gets the spec. yield strength up to 173 ksi.  That ends up with a 
relatively linear positive TCE of about +20ppm/degF.  Then you can parallel 
the Ni-Span spring with a relatively weaker steel spring which will 
contribute a negative TCE.   If the relative spring strengths are chosen 
properly you could end up with a fairly constant net zero TCE.

I admit that's going pretty far for ultimate performance.  That may be one 
of reasons why those highly respected instruments cost so much.  In 
practice, there are going to be circuit issues which will add significantly 
to the instrument temp. coeff. even if the spring is perfect.

Brett

At 10:36 AM 12/13/2008 -0800, Gary Lindgren wrote:
>Hi Brett,
>I won't have a daily temperature issue because I have the seismo in the wine
>cellar that is well insulated and in the basement. There are no daily
>temperature variation, it's only the seasonal I need to worry about. Yes
>indeed the Ni-Span-C 902 needs to be heat treated correctly. To achieve the
>zero change with temperature the material needs to be heat treated at 1000
>degrees F for 5 hours.
>Gary
>


__________________________________________________________

Public Seismic Network Mailing List (PSN-L)