> > My question is, how does one design an enclosure to prevent this > mistake? Or is it trial and error. Perhaps put a thermistor on the OCXO > and if the case temperature rises beyond what is normal case temperature > in free air, then the enclosure has too much insulation?
I suggest a thorough reading of Schaum's summary book on heat transfer, or get a secondhand copy of Incropera & DeWitt (?sp). It's pretty simple to build models for the planned enclosures. For example, the enclosures with insulation are mostly conduction transfer problems, the clever enclosure mentioned a couple of days back, with nylon standoffs and a sealed dewar (?) is a combined convection/radiation balance problem, harder because of the convection cells that will set up inside the dewar. Forget about thermal resistance, thermal capacitance, etc, and look at the basics of heat transfer! Also think about this: What's the REAL problem? Enough rant. > > With all this talk about thermal capacitance and resistance, perhaps > what we need is a thermal diode; it lets heat out but prevents thermal > fluctuations from getting in. A thermal Gore-Tex layer. There's always Maxwell's demons, if you can catch one :-) -- "The power of accurate observation is commonly called cynicism by those who have not got it." -George Bernard Shaw Dr. Don Latham AJ7LL Six Mile Systems LLC 17850 Six Mile Road POB 134 Huson, MT, 59846 VOX 406-626-4304 Skype: buffler2 www.lightningforensics.com www.sixmilesystems.com _______________________________________________ time-nuts mailing list -- [email protected] To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
