I’ve been seemingly running in quicksand ever since returning from the 7th Symposium on Frequency Standards and Metrology, what with the pileup of work while I was away (and everything seemingly breaking during that period of time) and getting ready for our clocks to leave the nest. But now, as I’m burning up my comp time from all that, I’ve had a chance to look back.
The conference was really good, as conferences go. A little over 100 people attended, from labs around the world. I knew perhaps a third of them already (though a few probably did not remember me) and met a few more. I didn’t see any glitches except for one or two instances of technical difficulties, which speaks volumes for the organizers and support staff, because you just know there were issues, and since they didn’t become visible it means they were solved quickly. The accommodations were very nice and the food was decent as far as dining hall food goes. The whole thing came in under the government-rate per diem, and the government is actually pretty stingy about such things.
Many of the talks encompassed the recent push into optical transitions for timekeeping; the microwave transitions used in the established clocks of today run at something a little less than 1010 cycles per second, but an optical transition will be about 4 orders of magnitude higher in frequency. Even if your detection can’t be done to the same level of precision, owing to lower light levels and fewer atoms, the higher frequency represents 2 or 3 orders of magnitude improvement in the overall measurement. The enabling technology for this has been the octave-spanning optical frequency comb, made by pulsed lasers in some nonlinear medium. If you consider the time width of the pulses Fourier-transformed in the frequency domain, you see a whole bunch of laser frequencies separated by the pulse repetition rate, so it looks like a comb. As I’ve mentioned before, you can use these individual frequencies to interrogate atoms, meaning you can measure some narrow clock transition. This becomes really useful when the comb spans an octave, so the low-frequency end can be frequency-doubled and referenced to the high-frequency end, making the comb stable. The repetition rate can be tied into some stable RF or microwave source, and now you know what each frequency is to a very high level of precision. A lot of labs are now doing this.
Continue reading →