Post on 21-Oct-2020
CCL Workshop on comb technology, March 2003
Bundesamt für E ich- und Vermessungswesen
Experiences with a commercial comb generator at the BEV
Michael MatusBundesamt für E ich und Vermessungswesen,Arltgasse 35, A-1160 Wien, Austriam.matus@ metrologie.at
CCL Workshop on comb technology, March 2003 2
Outline
1. Motivations for ins talling a comb generator atthe BEV
Primary standard for length metrology. T ype of instrument chosen
2. S pecial problems (and solutions) in lengthmetrology
low-power and frequency modulated lasers (mise en pratique)
3. Laser-frequency measurements as a s tandardservice to the public?
current s tatus and outlook
CCL Workshop on comb technology, March 2003 3
Motivation
T raditional laser intercomparisons campaigns:• mostly limited to 633 nm (2 I2-s tabilized lasers as primary
standards at the BEV)• rather time consuming
• will come to an end in the near future anyway
Optical femtosecond comb generator:• direct realisation of the definition of the meter
• wavelength range extended• no need for specialized laser sources; working standards can be
calibrated directly (gauge block interferometer)• service for the “public”, i.e for other NMI´s
• other goodies (e.g. Cs-controlled ECDL)
CCL Workshop on comb technology, March 2003 4
S etup
magic magic fiberfiber
NL -NL -interferometerinterferometer
pump-laserpump-laser
AOMAOM
T iT i:S -L aser:S -L aser
experimentexperiment
ins talled: Jan. 2002price: ~ 300 k
S er.Nr.: 001
Offspring company of MPQ
CCL Workshop on comb technology, March 2003 5
S pecial problems in length metrology (1)
ProblemsFeatures of some CCL lasers :• Low output power• S trong frequency modulation
Features of the comb:• Low output power per comb
mode (~10 nW)• High background (~1000 comb
modes reach the APD togetherwith the laser)
• S pectral “holes” in the fibrestretched light (near 633 nm)
• Wavefront dis tortion due to NL-interferometer for l > 600 nm
S olutionsF irs t s teps for 633nm• Powerful backup lasers• T esting of different fibres (trial
and error)
CCL Workshop on comb technology, March 2003 6
HeNe - 633 nm via 2-mode laser
Due to experimentalboundary conditions
only component “c” canbe measured without
us ing an additionalmixer and local
oscillator.
CCL Workshop on comb technology, March 2003 7
HeNe - 633 nm via ECDL
T he frequency of the ECDL is locked to a particular comb modeand can be tuned over some nm (“amplified comb mode”).
In practice it is difficult for a s ingle operator to lock the ECDL to themode des ired. Also the high output power is partially wasted
because of the rather poor beam geometry.
CCL Workshop on comb technology, March 2003 8
S pecial problems in length metrology (2)
ProblemsFeatures of some CCL lasers :• Low output power• S trong frequency modulation
Features of the comb:• Low output power per comb
mode (~10 nW)• High background (~1000 comb
modes reach the APD togetherwith the laser)
• S pectral “holes” in the fibrestretched light (near 633 nm)
• Wavefront dis tortion due to NL-interferometer
S olutionsF irs t s teps for 633nm• Powerful backup lasers• T esting of different fibres (trial
and error)
Current system (minor modificationsto original system)
• Optical separation of spectralregion of interest from the NL-interferometer
• Mode matching of combradiation and laser beam
• T echnique for checking thevalidity of the beat frequencyreadings
CCL Workshop on comb technology, March 2003 9
Present setup
Laser to becalibrated
ReplacedmirrorDichroic mirror, hightransmittance at 45°from 600 - 900 nm
CCL Workshop on comb technology, March 2003 10
Minimal laser power
Laser power atAPD even lowerby approx. 40 %
S ens itivity sufficient for most lasers !
CCL Workshop on comb technology, March 2003 11
F ilter band width
Narrow band RF-filter not necessary!
Us ing narrow bandfilters together with
frequency modulateds ignals can cause big
errors .
Passband ~25 MHz
CCL Workshop on comb technology, March 2003 12
Cycle s lip detection
S imple method toensure validity of
beat-frequencyreadings
drawback:laser power must be
s lightly higher
CCL Workshop on comb technology, March 2003 13
Results obtained during EUROMET #647
532 nm CMI YAG1 [32-0 R(56) a10] (~500 µW)563 260 223 514.3 kHz ± 0.4 kHz
543 nm PLG1 [28-0 R(106) b10] (~15 µW)551 580 162 396.6 kHz ± 0.6 kHz
633 nm BEV2 [11-5 R(127) f ] (~90 µW)473 612 353 604.36 kHz ± 0.14 kHz
CCL Workshop on comb technology, March 2003 14
F requency s tability
T wo samples variance as measured with the comb for sample times>10 s as good as one can expect from the standard!
CCL Workshop on comb technology, March 2003 15
Present s tatus
• Instrument is working fine and reliable. Eventhe fibre survived all experiments so far.
• Calibrations on 543nm and 633nm lasers arenow performed as a regular service.
• Already applied for entry in the cmc tables(520 nm – 1150 nm, U rel=10-9)
• Comb generator now included (as the nationals tandard) in the hierarchy scheme for lengthmeasurements .
CCL Workshop on comb technology, March 2003 16
hierarchy scheme for dimens ional measurements
Overview of traceability for dimens ionalcalibrations (and verifications) covered by
BEV‘s quality sys tem.
CCL Workshop on comb technology, March 2003 17
T op-level services
T he optical femtosecondcomb generator is now the
primary (national)s tandard for all length
measurements traceableto the BEV.
T he I2-s tabilized lasers BEV1and BEV2 can still be
used as primary s tandardsbut it makes more sense
to employ them as transferand travelling s tandards
CCL Workshop on comb technology, March 2003 18
Conclus ion
• For ins titutes with the need for more than as ingle wavelength, the femtosecond combgenerator as a length s tandard is an option!(as opposed to build and maintain a variety ofCCL lasers).
• Calibration of secondary s tandards (2-modestabilized etc.) is now not more elaborate thanthe traditional way. For this lasers even thebeat frequency measurement us ing aspectrum analyser might be adequate.
• We are confident in the reliability of thistechnique to offer it as a service to the public.