Hello Time-Nuts, I recently characterized a lot of 29 used (surplus telecom) Symmetricom SA.22C oscillators. These units never left the U.S. and I removed them from the original boards. I have attached a JPG of the plotted results. I also have a PDF scan, but it is too big to attach. Basically, it looks like there is a bias for the oscillators being a bit fast. I think this correlates very well with the data that has been published regarding helium permeation of the Rb gas cell (http://freqelec.com/space/ieee_perfrmnc_qz_rb_space.pdf). Most of the oscillators are 2011 manufacture and have probably reached equilibrium, it would be interesting to plot the data versus date code for a range of oscillators less than five years old. One question is - what about those slow units? - where did they start out, and are there other things going on? I think Corby has done some great work with his HP5065. And recently has shown his efforts to enclose the Rb cell and purge it with dry nitrogen (is it He free?). My thinking is that it should be pulled to about 10 torr vacuum (about the same as the Rb cell pressure) to create a He equilibrium to minimize aging. I'm sure there are lots of thoughts on both sides. My reasoning is that for an oscillator that ages towards fast at atmospheric pressure, and ages towards slow in space - there should be some pressure that aging would be minimized. Whether this is the case for the 5065 is also a good question. The good news regarding the SA.22c lot is that the DDS can be set to put the oscillators back on frequency such that temperature, voltage, and retrace dominate the frequency error. Enjoy! Skip Withrow

Hi Cool data !!! There are a *lot* of things that can impact an Rb cell over the long haul. I suspect they play the same sort of games that crystal guys do to eliminate any effect that they can identify. How they do this or that is always going to be a “company secret” sort of thing. You are left with some drift that they simply haven’t figured out yet. The data you have looks a *lot* like the long term aging distribution on a group of OCXO’s. The “power of ten” factor on the OCXO’s would be a bit different ….. Bob > On Nov 28, 2017, at 3:10 PM, Skip Withrow <skip.withrow@gmail.com> wrote: > > Hello Time-Nuts, > > I recently characterized a lot of 29 used (surplus telecom) > Symmetricom SA.22C oscillators. These units never left the U.S. and I > removed them from the original boards. I have attached a JPG of the > plotted results. I also have a PDF scan, but it is too big to attach. > > Basically, it looks like there is a bias for the oscillators being a > bit fast. I think this correlates very well with the data that has > been published regarding helium permeation of the Rb gas cell > (http://freqelec.com/space/ieee_perfrmnc_qz_rb_space.pdf). Most of > the oscillators are 2011 manufacture and have probably reached > equilibrium, it would be interesting to plot the data versus date code > for a range of oscillators less than five years old. > > One question is - what about those slow units? - where did they start > out, and are there other things going on? > > I think Corby has done some great work with his HP5065. And recently > has shown his efforts to enclose the Rb cell and purge it with dry > nitrogen (is it He free?). My thinking is that it should be pulled to > about 10 torr vacuum (about the same as the Rb cell pressure) to > create a He equilibrium to minimize aging. I'm sure there are lots of > thoughts on both sides. > > My reasoning is that for an oscillator that ages towards fast at > atmospheric pressure, and ages towards slow in space - there should be > some pressure that aging would be minimized. Whether this is the case > for the 5065 is also a good question. > > The good news regarding the SA.22c lot is that the DDS can be set to > put the oscillators back on frequency such that temperature, voltage, > and retrace dominate the frequency error. > > Enjoy! > Skip Withrow > <SA22-2.jpg>_______________________________________________ > 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.