I have another reading key. When the manufacturer publishes the specs of its crystals, it alrady took into account all of the possible known variables, some of them very tiny such as dependences on orientation (which do exist), so freeing the assembler (and the end-user) from caring of. (Anyway, on the issue of dependence on orientation, I feel the standpoint of this group is a bit superficial. I would suggest a test. Those of you who have GPSDO or cesium and are tracking the control voltage, please try rotating your oscillator, and make the voltage trace available to the group). Antonio I8IOV Rick Karlquist wrote: >......... However, the 10811 crystal > package is free to rotate to any random angle, so the > crystallographic orientation of a 10811 crystal would only be > consistent if the production line proactively installed them > in a certain orientation. I don't remember ever hearing about > anyone paying attention to how the crystals were oriented, > in all my years of dealing with the 10811. If it were important, > knowing how HP did things, the crystal package would be indexed > so that the orientation would not be up to the whim of the > assembler.

iovane@inwind.it wrote: > I have another reading key. > When the manufacturer publishes the specs of its crystals, it > alrady took into account all of the possible known variables, > some of them very tiny such as dependences on orientation > (which do exist), so freeing the assembler (and the end-user) > from caring of. > > (Anyway, on the issue of dependence on orientation, I feel > the standpoint of this group is a bit superficial. I would > suggest a test. Those of you who have GPSDO or cesium and are > tracking the control voltage, please try rotating your > oscillator, and make the voltage trace available to the > group). Since the gravitational/acceleration sensitivity is a vector property, it should not be hard to make a few measurements to establish this vector sensitivity for a number of oscillators and see if there is a good correlation or not. I have enougth 10811s on the bench that I could test with (8 would be sufficient to get some indication). The turn-over test needs to be done over all three axis. So top/bottom turn, long-side/long-side turn and short-end/short-end turn. The crystal blank is mounted parallel to the top and bottom sides BTW. Cheers, Magnus

I have seen one instance where crystal orientation does matter. Among other things, I'm interested in WW II military radios. Admittedly long ago. The Canadian Wireless Set 19, the standard radio for many tanks, etc., in WW II had an external crystal calibrator. It put out 1 MHz, 100 KHz, and 10 KHz. Interestingly, the 100 KHz and 1 MHz were generated bu the same crystal. The calibrator was in its own case about 1.5 x 4 x 10 inches and could be mounted on edge or side. The manual specifically stated the crystal MUST be mounted in the correct orientation within the calibrator before the calibrator is mounted to the set. -John ============ > I have another reading key. > When the manufacturer publishes the specs of its crystals, it > alrady took into account all of the possible known variables, > some of them very tiny such as dependences on orientation > (which do exist), so freeing the assembler (and the end-user) > from caring of. > > (Anyway, on the issue of dependence on orientation, I feel > the standpoint of this group is a bit superficial. I would > suggest a test. Those of you who have GPSDO or cesium and are > tracking the control voltage, please try rotating your > oscillator, and make the voltage trace available to the > group). > > Antonio I8IOV > > Rick Karlquist wrote: > >>......... However, the 10811 crystal >> package is free to rotate to any random angle, so the >> crystallographic orientation of a 10811 crystal would only be >> consistent if the production line proactively installed them >> in a certain orientation. I don't remember ever hearing about >> anyone paying attention to how the crystals were oriented, >> in all my years of dealing with the 10811. If it were important, >> knowing how HP did things, the crystal package would be indexed >> so that the orientation would not be up to the whim of the >> assembler. > > > > _______________________________________________ > 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. > >