AOM & AOD
Forex
I'm just working over Acoustic Modulator and acoustic Deflectors, mainly
with Isomet, Neos and Crystal technologies.
There's one thing I don't understand.
Setting the right Bragg angle (about 2.9 °) it's possible to modulate the
1st order changing the amplitude of the RF acoustic wave or to deflect the
same beam changing the RF frequency of the acoustic wave injected into the
TeO2 crystal..
If I use a pure sinusoidal trasversal wave, the theory states that only the
first order should appear, while the other diffraction order should be
eliminated.
The problem is that wichever RF wave I use, I have always the beam
diffracted on at least 4 or 5 orders, limiting the efficiency of the
modulation and confusing the deflection, as the 1st order diffraciton goes
over the 2nd order and so on.
where am I wrong? or there's any suggestions?
What I'm trying to do is a raster laser video display using only
Acoustic-opto elements.. and no mechanical parts..
Forex (ebay user: gigilofisica)
Repeating Rifle
7/6/04 2:08 AM:
I have been away from this kind of work a long long time.
It would useful if you indicated the relative intensities of the various
orders compared to the input. Could it be that a diffracted wave acts as an
input wave as well? How are you sure that RF harmonics are absent? Can the
nonlinear materials themselves create harmonics?
Bill
Forex
"Repeating Rifle" <Salm...@sbcglobal.net> ha scritto nel messaggio
news:BD101B1F.1D363%Salm...@sbcglobal.net...
> It would useful if you indicated the relative intensities of the various
> orders compared to the input. Could it be that a diffracted wave acts as
an
> input wave as well? How are you sure that RF harmonics are absent? Can the
> nonlinear materials themselves create harmonics?
>
mmh, I've analized the wave and you're right, I had a lot of spurious.
Thanks for the suggest.
It's a little bit better, btw I'm interested on the second answer, how could
a diffracted wave acts as an input wave? you mean by reflection or internal
into the crystal?..is there any literature on this?.
Luigi
Repeating Rifle
7/6/04 10:32 AM:
> It's a little bit better, btw I'm interested on the second answer, how could
> a diffracted wave acts as an input wave? you mean by reflection or internal
> into the crystal?..is there any literature on this?.
The input wave is the source that produces the first order. As the first
order builds up, it can be the source for the second order, and so on.
As I said, it has been a long time for me. IIRC you may want to look at the
distinction between Bragg and Brillouin diffraction.
Bill
Lineshape
Hello -
I work with AO's on a daily basis in my lab and often see multiple orders,
usually under the following 2 circumstances :
1) If your beam is converging or diverging, even with a slight focus,
you'll produce a few multiple orders - the range of k-vectors in an
uncollimated beam allows some efficiency for diffracting into higher orders.
Furthermore, although focusing lightly into an AO for a spot size of ~100
micron is great for high switching speeds, a small spot also means you are
covering fewer acoustic wavelengths in the crystal so that multiple orders
become present (i.e. you are no longer in the Bragg scattering regime). For
the best efficiency I generally have collimated beams as large as will fit
in the entrance aperture without clipping. If I need switching speed, I
lightly focus to a few hundred microns. These results are
manufacturer/crystal dependent, since active areas, crystal lengths, and
drive powers vary from one company to another.
2) It's possible, but unlikely, that your RF Drive power is too high. This
can happen though if you mix drivers/AOs from different companies.
Hope this helps,
Frank
Forex
"Lineshape" <line...@cox.net> ha scritto nel messaggio
news:ouQGc.768$b95.368@lakeread06...
> 1) If your beam is converging or diverging, even with a slight focus,
> you'll produce a few multiple orders - the range of k-vectors in an
> uncollimated beam allows some efficiency for diffracting into higher
orders.
you mean I have to use a collimated beam?
I normally put the modulator on the waist and then separte the different
order by expanding a re-collimating.
> Furthermore, although focusing lightly into an AO for a spot size of ~100
> micron is great for high switching speeds, a small spot also means you are
> covering fewer acoustic wavelengths in the crystal so that multiple orders
> become present (i.e. you are no longer in the Bragg scattering regime).
For
> the best efficiency I generally have collimated beams as large as will fit
> in the entrance aperture without clipping. If I need switching speed, I
> lightly focus to a few hundred microns. These results are
> manufacturer/crystal dependent, since active areas, crystal lengths, and
> drive powers vary from one company to another.
>
interssting, it seem exactly what I do..
> 2) It's possible, but unlikely, that your RF Drive power is too high.
This
> can happen though if you mix drivers/AOs from different companies.
>
> Hope this helps,
Thanks, I used a spectum analyzer and I found many harmonics that were 10dB
less than the foundamental. too much, it seems the cable impedance or some
other things.
Many thanks for suggestion, it helped me..
Luigi
Lineshape
"Forex" wrote in message news:2l31ucF...@uni-berlin.de...
>
> "Lineshape" ha scritto nel messaggio
> news:ouQGc.768$b95.368@lakeread06...
>
> > 1) If your beam is converging or diverging, even with a slight focus,
> > you'll produce a few multiple orders - the range of k-vectors in an
> > uncollimated beam allows some efficiency for diffracting into higher
> orders.
>
> you mean I have to use a collimated beam?
> I normally put the modulator on the waist and then separte the different
> order by expanding a re-collimating.
To be precise, whenever you put the AO at the waist, you are using a
"collimated" beam. It's just not collimated for very long. I make this
point because it is definitely good practice to place the AO at the waist as
you do; this is just impractical if your beam has a long rayleigh length (z0
= pi*waist/wavelength). In this case, which usually gives the best
efficiency, as long as you are not clipping the beam with the entrance
aperture of the AO, its placement isn't a big deal.
This said, even if you are focusing to ~100 micron into the AO, you should
still see most (>60%, depending on the wavelength and RF drive power) of the
power diffracted to the first order. As a previous post suggested, better
diagnosis can be provided if you can say what your diffracted intensities
are, and what lens/waist combination you have.
Good luck,
Frank
Christian Kurtsiefer
Good ol' Born&Wolf has a whole chapter on it (number 12 in my 7th ed)
with all the different regimes. But maybe there are more compact
descriptions...
Cheers, Christian
Jamie Carter
Both the plane wave spectrum of the optical beam and the acoustic beam
must be well controlled. As another posted, a large range of
divergence angles in the laser (due to collimation or just small beam
size will provide energy to be coupled (or not *) into spurious
orders. Generally the difference between the waist and the region
close to the waist involves a sphericity to the wavefront. Both have
about the same angular extent in the plane wave spectrum of the beam
(search for Korpel's teatment here). *When the beam is focused too
tightly, it has angular spectrum that is not matched by the acoustic
beam (which also diffracts) and these angle components may not be
coupled into the 1st order or diffracted at all.
Generally, the spurious orders arise from non-linearites. There exist
a host of sources for non-linear phenomena in the AO cell. First, the
interaction is inherently non-linear. The relation between strain
field amplitude and optical coupling is sinusoidal; but this doesn't
induce spurious diffracted orders for a single tone. The next one on
the list is acoustic propagation. If the strain field has a very high
amplitude, then the acoustic velocity (or you can use acousic slowness
formalisms) picks up an additional (quadractic in strain field) term.
This generates harmonics of the acoustic frequency that can easily
couple the optical beam into the very same higher order directions.
Another source of spurious beams arises directly from the acoustic
diffraction as the sonic field propagates away from the transducer.
This why most quality AOMs have a diamond shaped tranducer, usually
with the long part of the diamond aligned to the direction of optical
propagation (note that the width of the optical beam compared to the
acoustic wavelength is not the big driver between Bragg and Raman-Nath
diffraction; it is much more dependent on the interaction length). As
the sonic field diffracts, the diffraction orders (lobes) tend to move
perpendicular to the edges of the transducer (when projected to the
transducer plane). This arrangement takes them out of the optical beam
path and there will be little acoustic diffraction along the optical
beam cross section within the acoustic interaction region. Many papers
have been published on this phenomena and novel transducer shapes
proposed that minimizes the effect. Look for papers by I. C. Chang, E.
Young and M. Shaw.
Generally, you should be able to get 95% in the blue (I have seen
modulation inversion in the blue with little power "lost" to the
(+/-)2 and (-/+)1 orders - obviously about 5% lost), 89% in the green,
and about 82% in the red with a standard TeO2 cell, 200 micron tall
tranducer (length varies) and about 125 to 150 micron waist. If
acoustic transition time is a modulation band limit, consider
searching for "Scophony Labs" for a cleaver fix if you are using an
opto-mechanical scan mechanism.
If you are trying to use the AOM as a deflector for more than a few
degrees, good luck. The transducer thickness, metalization layers
(acoustic antireflection "coating") and RF impedance matching are
tuned (with a high Q factor) to the AO cell cariier. Going more than a
few tenths of that center frequency away in freqency will result in a
drasic loss of efficiency between RF signal and acoustic strain
amplitude (and probably result in some non-linear effects as that
energy must go somewhere - usually back to the RF amp (generator) and
create problems for the amplifier - if you have enough RF power
consider a 3dB pad at the AO cell end of the cable).
Best of luck!
James Carter
http://www.jacarter3.com
"Forex" <gig...@lcnet.it> wrote in message news:<2kv8kaF...@uni-berlin.de>...
Forex
"Jamie Carter" <jaca...@onebox.com> ha scritto nel messaggio
news:4b78e1a9.04071...@posting.google.com...
> Best of luck!
>
> James Carter
> http://www.jacarter3.com
>
That was absolutely interesting an clear, thanks a lot..
Forex
Forex
"Lineshape" <line...@cox.net> ha scritto nel messaggio
news:rl1Hc.5$oh.1@lakeread05...
> This said, even if you are focusing to ~100 micron into the AO, you should
> still see most (>60%, depending on the wavelength and RF drive power) of
the
> power diffracted to the first order. As a previous post suggested, better
> diagnosis can be provided if you can say what your diffracted intensities
> are, and what lens/waist combination you have.
>
at the moment, I use 3 Viasho DPSS, 443 (50 mW), 572(100mW), 633(100mW).
but I'm waiting a white laser quantum OPO.
The AOM are Isomet 1205-B1 ( TeO2, 80 mhz center freq).
The modulator is Isomet 232
But now I've designed them by myself.
I've tested everything with a Rhode&Scwartz modulator, 1w power, sinusoidal
carrier, analog maodulation 0-1 V, e 2nd harmonics at -20db.
The efficency is 40% for 443, 55 % for 572, 50% for 633
All the 3 beams have a 2nd order diffraction af about 15 %.
Ok, now I think I understood better the problem, that is first on the right
cable impedance, second on the very low match of the input impedance of the
AOM, they are done with bad 50 ohms load ( 10%), so I have to tune the
impedance looking at the ROS on the cable.
At this moment I'm asking if there is a possibility to build and cut by
myself this kind of crystal (TeO2 or PbMoO4 or flint glass). (for sure not
exactly by myself, but as I want)..
The problem will be the transducer, but what I don't understand is the very
high price of this product, Isomet, Neos, AA, Crystal technology quote an
AOM over a 1000$.
Don't mention that I bought many AOM on e-bay for some $ :-) exactly the
same.
No problem for me to design and build a VCO or a modulator.
As I said, I'm a bit new of optics but I worked about 25 years on digital
electronics, but still I can't understand why an FM VCO or a amplitude
modulator like those cost so much..
BTW, many thanks to you and other guys for the answers, the optical
community is very nice..
Luigi