spin noise

[George Herold]

(grumble, I wanted to cross post this on sci.optics but I don't know how todo it from google groups... could someone help?)

I was just sent a nice article on spin noise spectroscopy.
http://www.nature.com/nature/journal/v431/n7004/full/nature02804.html
There’s an Arxiv version of the paper too.
arXiv:cond-mat/0408107
or search for
“Spectroscopy of spontaneous spin noise”

So this looks like a fun experiment. A mixture of diode laser spectroscopyand noise. (what could be better?) The crucial piece of kit is a balanced photodiode.
(Well I’ve already got the diode laser and Rb cells..etc)
Cue Phil Hobbs (I hope I’m using cue in the right way)

So Phil here’s a noise question. Say instead of a balanced detector (twophotodiodes (PD) and one TIA opamp) I do this with two PD TIA circuits and then subtract the two signals in an opamp. How much of a noise hit do I take?
Certainly I get the extra amp noise for the second TIA.
Oh! I think I see the issue. If I’ve got a small difference signal I can crank up the gain of the balanced PD. Whereas I’m swamped with noise in the two PD-TIA case.. and then looking for a small difference.

(OK I’ll post this anyway... spin noise is a new idea for me, and perhapsothers)

 

[RobertMacy]

...snip...
(OK I?ll post this anyway... spin noise is a new idea for me, and
perhaps others)

Thank you for posting !! although $32 seems a bit high for an article

Interesting! Is there anything advantageous to 'applying' specific fields,
intead of simply 'observing' fields? Maybe, like pull out a molecule, or
hold it still while everything else is 'flushed' away?

 

[John Miles, KE5FX]

Thank you for posting !! although $32 seems a bit high for an article

It's on arxiv:
http://arxiv.org/abs/cond-mat/0408107

-- john

 

[RobertMacy]

It's on arxiv:
http://arxiv.org/abs/cond-mat/0408107

-- john

Thanks!

 

[George Herold]

Thank you for posting !! although $32 seems a bit high for an article



Interesting! Is there anything advantageous to 'applying' specific fields,
intead of simply 'observing' fields? Maybe, like pull out a molecule, or
hold it still while everything else is 'flushed' away?

Well when you apply the field (say the light field in optical pumping whichis the tradition way to see the above (magnetic) splitting of ground stateof the atomic system) then it turns into a system that is not in thermal equilibrium and this (I guess.. I not at all a theorist) means that the states are perturbed a bit. Where the field free method has no such perturbation... But the signals are much much smaller!! The other thing you might worry about is the the RF fields used to probe the states can also cause changes. Again I have no idea how big these effects are.

George H.

 

[George Herold]

Haven't got time to read the paper just now, but it sounds like a good
application for a laser noise canceller.

My handwavy understanding.
1.) there's fluctuations in the spins.. (noise)
2.) if the system has resonances, then noise driving those resonances will show up the in noise spectrum. (I think of white noise into a band pass filter... and what you see at the output.)
3.) 'Slightly' off resonance faraday rotation is used to detect the fluctuations in the spin states and the ground state resonaces are thus observed. (So there's a term in the denominator of faraday signal that goes as a resonant frequency (this is now an optical resonance.. or maybe a better word is an optical transition frequency) minus the applied optical frequency.. this kinda blows up when the two are equal.. and when they are near equal you get a big enhancement of the signal.)

Using two separate paths doesn't work well because you can't match the
phase shifts to the required accuracy. By wiring two PDs in series
(parallel at AC due to the supply bypasses), you get only one signal
path and one set of circuit strays, so it's only the transit time
behaviour of the PDs and optical delays that can differ.

Oh, thanks Phil I forgot about balancing the phase shifts. (path lengths)

The noise canceller is sort of an intermediate situation, where you have
very nearly the same strays but not exactly. (The high frequency tweaks
in my fancy ones are one answer to that problem.)

Say with two PD's in a box how do you do the optical alignment. (or do you 'cheat' and use fibers. I scribbled down this gizmo, where the entire box moves in x and y (assume light direction is z) and then one PD does and independent x translation, and the other a y- . A bit complicated.
Hmm maybe just do everything with mirrors, so a separate x/y mirror adjustment for each PD?

George H.