mixing up

[RichD]

Say you want to generate a very narrow light
band in the UV, with precise desired frequency.
Is it feasible to do so, by starting with a reference
microwave signal, then heterodyning up and up,
doubling in successive stages?

What is the limit on this process? Assume each
stage utilizes a technology appropriate for that wavelength.

--
Rich

[Phil Hobbs]

Post by RichD
Say you want to generate a very narrow light
band in the UV, with precise desired frequency.
Is it feasible to do so, by starting with a reference
microwave signal, then heterodyning up and up,
doubling in successive stages?
What is the limit on this process? Assume each
stage utilizes a technology appropriate for that wavelength.
--
Rich

You actually go the other way, starting with a femtosecond Ti:sapphire
laser, pushing it through a holey fiber to broaden the spectrum by self
phase modulation, and then lock a line at the blue end to the second
harmonic of the red end.

It's a 1:1 lock, so there's no phase ambiguity. You can lock the beat
frequency to an RF signal, and get e.g. x10**6 frequency multiplication
without the accompanying 120 dB phase noise penalty.

Jan Hall and Ted Haensch got the Nobel prize in physics for that bit of
extreme cleverness.

Cheers

Phil Hobbs

[Joe Gwinn]

On Tue, 21 Mar 2023 19:30:50 -0400, Phil Hobbs

Post by Phil Hobbs

Post by RichD
Say you want to generate a very narrow light
band in the UV, with precise desired frequency.
Is it feasible to do so, by starting with a reference
microwave signal, then heterodyning up and up,
doubling in successive stages?
What is the limit on this process? Assume each
stage utilizes a technology appropriate for that wavelength.

It is possible, but extremely difficult to achieve in practice, and
only national standards labs could do it.

What Phil H discusses below replaced the above approach overnight. It
was definitely worth a Nobel Prize. Now, locking light to microwave
is easy.

Post by Phil Hobbs

Post by RichD
Rich

You actually go the other way, starting with a femtosecond Ti:sapphire
laser, pushing it through a holey fiber to broaden the spectrum by self
phase modulation, and then lock a line at the blue end to the second
harmonic of the red end.
It's a 1:1 lock, so there's no phase ambiguity. You can lock the beat
frequency to an RF signal, and get e.g. x10**6 frequency multiplication
without the accompanying 120 dB phase noise penalty.
Jan Hall and Ted Haensch got the Nobel prize in physics for that bit of
extreme cleverness.
Cheers
Phil Hobbs

Here is the Nobel Lecture (scroll down):

.<https://www.nobelprize.org/prizes/physics/2005/hansch/biographical/>


Joe Gwinn