Hal Murray wrote: >> A good (and well-aged) double-oven OCXO with aging compensation should >> be able to provide that type of performance. The problem is: how >> much physical movement will the unit experience? What are the >> temperature extremes that are expected, and how fast does the >> ambient move between the temp extremes? Can it be GPS locked during >> that time? >> > > Assuming you don't GPS lock it... > > How well will two units from the same manufacturer/batch track? > > I'm assuming they will see roughly the same temperature profile and that you > start with them running at the same frequency. > > > Has anybody made a triple oven setup? I'm thinking of a big box with a few > holes and a temperature controlled fan. The idea is to let the oven (and > electronics) keep itself warm when the ambient temperature gets very cold. > > I'm picturing an open hill top at night with clear skies so there would be > lots of radiation cooling. Of course, with an open hill top, GPS recption > would be easy. > > > > Hal,
A triple oven is a particularly bad idea, its just too difficult to achieve a wide operating temperature range without losing temperature control or cooking everything in the innermost oven unless one uses a watercooled peltier device to control the outer oven temperature. A well designed single oven with feedforward (or the outer oven of a dual oven system if you cant redesign/rebuild the inner oven) can easily achieve a thermal gain of more than 1E5 see: http://tf.nist.gov/timefreq/general/pdf/595.pdf A bootstrapped oven design like that used by Wenzel (not as new as he'd have you believe - its been in use in one form or another for the better part of a century) will also help in improving performance. The optimum solution is of course to combine the 2 techniques. http://www.wenzel.com/documents/Sub-pico%20Multiplier.pdf With a thermal gain of 1E5 for the outer oven temperature fluctuations at the outer shell of the inner oven will only be a 1 millikelvin or so (assuming the ambient temperature range is less than 100K). Bruce _______________________________________________ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.