[time-nuts] VRefs - I'm a bigger Voltnut than a timenut..

[Randy Leifer]

Bigger Volt-nut here.
...albeit, less knowledgeable than most of you.
I have been in the chase for the perfect volt.
I have 2 JRL SCO-105 saturated cell references running 24/7.
But checking it's accuracy....I must rely on my cal'd HP3457 meter.
Although, those cells (6 total, 3 in each cabinet) are pretty stable, and good for determining drift of the Vrefs I build.
I have a Geller 5v, and I've built a few of my own refs using some 
National LM399AH-50 along with a chopper buffer. 
(these LM399's come with a 1000 hr burn-in and a drift plot)
I am getting a 10uV drift with these circuits, but that is over a +/- 15* F ambient temp range (as is the voltmeter)...and they're only in plastic cases in open air--not shielded yet.
http://i5.photobucket.com/albums/y177/Midiot/my_lm399_ref.gif

I am currently building/testing/tweaking a temp stable insulated enclosure,
using a Peltier, a 5 watt controller, and 0.1% glass thermistors.....to house my Vrefs and some standard resistors. 
I am shooting for 25* C, but the outside air still has an slight effect on the internal air (+/- 0.5* C on the inside).....not bad but I think I can do better. I put a low-ohm thermistor outside the box, in series with the main sensing thermistor inside, and this helps. But this may be a whole new topic.
I'm up for a Vref project. How about you?

=Randy=