[volt-nuts] Some thoughts about 3458A stability

[WarrenS]

I have taken JUST the opposite approach and achieved good results with cheap devices using the following rules.

1) Run them in at a combination of heat and time

2) the most important step is then Select out the good ones.

3) add a simple adjustable TC circuit that does a first order temp correction to zero the TC over a limited lab range
The error due to TC that is possible is better than  0.1 PPM Total over a lab temp environment.
(see attached)

4) After that DO NOT heat them up at all, Keep them at room temp.

5) Do not turn them on except when needed for short times
 
With no high temp and no power they pretty much just sit there and don't change.
Then you can use them as a reference should be use, as a transfer device, Not as a working reference. 

ws

****************

n message <4A78695C.5050301 at freenet.de>, "Dr. Frank Stellmach" writes:

>This can be overcome by a permanent heating ( as with the 732s), [...]

I actually thought about doing that, has anybody tried it ?

-- 
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.