MSE Challenges - Plasma Science and Fusion Center (PSFC)

By Cheryl Ford,2014-01-20 02:19
5 views 0
MSE Challenges - Plasma Science and Fusion Center (PSFC)

    Challenges for MSE on Alcator C-Mod

    S. D. Scott

    October, 2005

    Updated December 12, 2005

    1. Calibration of an MSE channel using an APD detector.

? Calibrate an MSE channel in the usual way with the rotational stage.

? Replace the PMT with the APD.

? Re-calibrate.

? Check that the calibration with the APD doesn’t change if we vary the intensity.

    2. Why does the polarization angle measured with the WGP vary with radius?

    ? Is this effect real or does it indicate a problem with MSE?

    ? Approach: in-vessel calibration of the WGP for each MSE channel.

    ? Schedule: Jan-Feb 2006.

    3. Why does the measured polarization angle vary with intensity?

    ? Why is the effect of intensity most pronounced at low intensity rather than high


    ? Approach: illuminate a PMT with photodiode driven by a signal generator. Drive

    pre-amp directly with a signal generator.

    ? Also, see if effect goes away if the PMT is replaced by an avalanche photodiode.

    Schedule: December.

    ? Schedule: November

    3. Why is the measured PEM retardance not equal to the demand retardance?

    ? Is it due to the fact that the rays hit the PEMs at positions other than the center?

    ? Calculation: compute the average retardance for both PEMs for all MSE channels by

    launching rays from a given MSE channel (i.e. the backlit image area), recording

    where these rays hit the PEMs, and then assume that the retardance is parabolic along

    one axis of the PEM and independent of position along the other axis.

? Compare these computed retardances with the measured retardances … can we

    reproduce the channel dependence, which shows a lower retardance for core channels

    (7,8,9) than edge channels (0,1,2).

? Schedule: November

    4. How do mirror reflections affect the ‘rotation’ of polarization through the MSE

    optical system?

    ? Approach: ray-trace individual channels, recording k-vector and mirror

    normal at each mirror for all rays passing through lens L1. Apply formula (to

    be derived!) that determines the rotation of polarization for each ray as a

    function of the k-vector and mirror normal.

? Question #1: can we understand the channel-to-channel variation of the

    rotation, i.e. the fact that different channels observe 2-3 degree differences in

    polarization angle for the same incident direction of polarization. For this

    calculation, we should launch rays from the entire viewing area of a given


? Question #2: is this effect (rather than variation of PEM retardance)

    responsible for the variation in angle we observe when we move the light

    source within the viewing area of a single channel?

    ? Schedule: limited by derivation of the correct formula. Should take only a

    few days, once formula is derived.

    5. How much could we increase the MSE signal strength by replacing the PMTs with

    avalanche photodiodes?

? How much will the statistical error decrease?

? Tasks:

    o Select APD vendor and place order for one APD

    o Purchase pre-amp from Nova Photonics.

    o Disassemble one spare PMT assembly to see how it is constructed.

    o Disassemble assembly from Levinton.

    o Design optical system & housing for APD.

    ? Schedule: place order for APD and pre-amp in early November. Design

    optical system & housing in November December. Test in January.

    5. Measure Faraday rotation in-situ.

    ? Approach: position WGP in front of L1 and take data. Do this as early in Fall

    2005 run campaign as possible, i.e. before WGP potentially gets coated with


Plasma-only shots (no DNB): these are slightly less useful than the shots using the

    DNB described below, since the light source is possibly diffuse (but we think the

    MSE optics does a good job at rejecting light from other than its nominal field of

    view for each channel, so this shouldn’t be a big concern).

    ? Toroidal field ramp in a single shot, with Ip fixed. Use plasma as source

    of light.

    ? Ip ramp in a single shot with Bt constant.

Note: long-pulse DNB operation is needed for the following ramp shots:

    ? Beam-into-gas, Bt = 5 Tesla, EF ramp

    ? Beam-into-gas, Bt = 5 Teals, EF constant.

    ? Beam-into-gas, Bt = 3 Tesla, EF ramp (need to retune filters).

    ? Beam-into-gas, Bt = 3 Tesla, EF constant.

Potential benefit: if the measured profile of polarization angles matches the

    profile we measured in-vessel which had conditions of (1) atmospheric pressure;

    (2) no magnetic field; and (3) no tokamak-induced noise, power supply droop,

    interference with PEMs, etc. then we have effectively ruled out all of these

    potential issues as explanations of the MSE calibration difficulties.

    6. Spectrum measurements

? Objective: confirm that the spectrum matches the expected spectrum, and get

    information about ratios of upper state populations.

    ? Approach: Use long-pulse DNB to obtain improved spectra during beam-

    into-gas and beam-into-plasma at a few values of the TF / EF.

? Details to be determined with Bill Rowan.

    ? Schedule: as soon as DNB interlock becomes available. Hopefully, this can

    be done this run period, i.e. before January 5, 2006.

    Proposed Procedure to Calibrate the WGP In-vessel

    One of two things are true:

1. The WGP generates polarized light with about the same polarization angle

    along each of the 10 MSE channels, but MSE incorrectly measures this

    angle, thereby generating the a measured angle that varies with channel

    number; or

2. For reasons we don’t understand (and contrary to the lab measurements),

    the WGP generates different polarization angles along each of the 10

    MSE channels, and MSE correctly measures these angles.

    So I think we should perform an in-vessel calibration of the polarization direction

    of the WGP for each of the 10 MSE channels. Approach:

1. Position the WGP in front of L1.

2. Remove the linear polarizer from the MSE diagnostic, and turn off the


3. Mount the rotational stage with circular LED array etc. on the optical

    bench along the DNB trajectory as usual.

4. Backlight a selected MSE channel to position the center backlit image at

    the center of the rotational stage. Also, using the alignment fixture, aim

    the rotational stage directly at lens L1.

5. Perform a 36- or 72- point measurement as the rotational stage goes

    through 360 degrees and take data with MSE. But we will use only the

    amplitude, i.e. we will effectively determine the angle at which the

    measured intensity reaches a maximum.

6. Now insert the MSE linear polarizer and turn on the PEMs. Perform a 36-

    or 72-point measurement as the rotational stage goes through 360

    degrees and take data with MSE. MSE should measure the same angle

    irrespective of the angle of the rotational stage (except that when the

    intensity gets very small, when the rotational stage transmission axis is

    nearly perpendicular to the WGP transmission axis, we may get a small

    change due to the known variation of measured angle with intensity).

    Before step 7, mark the current position of the shutter on the MSE turret,

    so that when we make measurements in the future, we can return to

    exactly the same position of the WGP.

    7. Rotate the shutter so that lens L1 views the torus unobstructed, i.e. move

    the WGP polarizer out of the way. Repeat step 6, i.e. perform the usual

    in-vessel calibration of the channel.

    8. Now do an in-vessel check of the angle-of-incidence effect: with the WGP

    still out of the way, rotate the entire rotational stage about its vertical axis

    so that it is no longer pointed directly at lens L1. If possible, it would be

    great if we could rotate it by 10 or even 15 degrees. Do another 36- or 72-

    point measurement with the rotational stage. Ideally, we would take

    measurements at a rotational angle of say 5, 10, and 15 degrees.

    9. Return the WGP to its original position, i.e. to the fiducial mark as

    recorded just before step 7. Repeat steps 4-8 for each of the MSE

    channels or a subset of the MSE channels, as time permits. If time is

    limited, we should do the edge channel (0), a central channel (3 or 4) and

    channel 8.

If time is limited, we probably can’t all of the proposed measurements for all of

    the channels. Of the various measurements, step #5 is more important than 6, 7,

    and 8.

Step #6 is probably the least important, since we expect a null result. If in fact

    we see a null result on the first channel we measure, i.e. no change in angle as

    the rotational stage is rotated, then there would be little reason to repeat step #6

    for the other channels.

    MSE Calibration Issues Studied to Date

    Most of the issues discussed in this list have been described in Howard Yuh’s doctoral dissertation. Yuh’s dissertation is always a good place to start.

? Quantum mechanics: effect of adding B to E=v x B on Stark Effect

    Memo 34: Analysis of Breton angles

    Memo 39: Polarization angle for combined MSE + Zeeman

    Memo 40: computing intensity in e1 + e2 direction

    Memo 41: elliptically polarized emission

    Memo 42: effect of Zeeman splitting on MSE analysis

    Memo 46: Pressure and magnetic field in long pulse DNB

    APS 2004: Invessel Calibration of the Alcator C-Mod MSE Diagnostic

    DNB meeting (11/2004): Effect of Zeeman Splitting on MSE Analysis

? Retardance of photoelastic modulators

    Memo 7: Is the retardance correct?

    Effect of variations in PEM retardance (October 2005)

? Polarized impurity radiation from beam excitation

    MP436: Tests of fluorine and molecular emission contamination of MSE

    signals and calibration

    MP436 Summary (September 2005)

? Beam-excited molecular D emission during beam-into-gas calibration 2

    MP436: Tests of fluorine and molecular emission contamination of MSE

    signals and calibration

    MP436 Summary (September 2005)

? Non-ideal mirror properties

    Memo 2: Effect of polarizer angle on MSE analysis

    Memo 20: Proper treatment of effect of mirrors

    Memo 21: Angle of pi radiation in MSE reference frame

    Memo 23: More geometry

? Scattered light along MSE optical train

? Phase shift between PEM drive signal and measured Fourier components of the

    MSE signal

    Memo 35: Calculation of MSE offsets

? Faraday rotation in MSE optics and results of wire grid polarizer

    Memo 36: Axial magnetic field from two permanent ring magnets

    Memo 37: Invessel Verdet Measurements

    Considerations on MP to measure Faraday rotation

    Further analysis of data from MSE invessel linear polarizer (May 2005)

    Inferring Faraday rotation with the wire grid polarizer (July 2005)

    Faraday rotation measurements from July 18 (July 2005)

    Results from WGP study of September 16 (September 2005)

? Finite bit A/D resolution at low signal amplitudes

    Effect of finite bit resolution of A/D on MSE analysis

? Effect of circularly polarized light

    Memo 5: Effect of circularly polarized light on MSE analysis

    Memo 33: correcting for circularly polarized light in mse_2020

? Effect of Zeeman splitting on thermal H?.

    Memo 14: Thermal Dalpha emission due to CX on beam ions

? Faraday rotation including effect of ripple on MSE lens assembly L3

    Memo 37: Effect of TF ripple on MSE lens L3

? Apodization effects arising from short analysis durations

    Effect of Aliasing on MSE analysis

    Apodization effects on MSE analysis (may 2003)

    More on apodization effects on MSE analysis (May 2003)

? Effect of finite entrance slits on tilted bandpass filters

    Memo 25: Effect of finite entrance slit on tilted bandpass filters

? Bending of DNB due to reionization in drift duct and fringe fields

    Memo 29: Effect of torus pressure on MSE calibration

    Pressure and Magnetic field in long pulse DNB (March 2005)

? Effect of sigma contamination of pi line

    ..\Presentations\MSE Data Analysis 4 August 2003.ppt

? Proper treatment of subtracting background noise

    Memo 4: How to subtract background light

    Memo 9: corrected background subtraction

    Improved noise suppression in MSE analysis (November 2002)

? Digitizal FFTs replacing lock-in amplifiers to provide measurements of PEM

    retardance and circularly polarized light

? Polarization fraction of measured MSE signal

    Memo 3: how to compute polarization fraction

    Memo 13: How to compute polarization fraction with background

? Effect of non-normal incidence at the PEMs

    Memo 1: Effect of non-normal PEM incidence

    Memo 18: Linear polarizers and flat glass plates

? Effect of spatial resolution on MSE analysis for edge channels which see a large

    geometric multiplier between polarimeter angle and magnetic field pitch angle

    MSE spatial resolution and effect on analysis (July 2004)

? Effect of temperature on temperature-tuned bandpass optical filters

    Filter swap temperature scan (April 2004)

? Relationship between pitch angle and polarization angle in MSE reference frame

    Memo 32: The relation between real and MSE invessel optics pitch angles

? Understanding polarization angle of light transmitted through a linear polarizer at

    non-normal incidence

    Memo 48: Linear polarizers at oblique angles of incidence

    Memo 49: Tilted linear polarizers at oblique angles of incidence

    Memo 49c: Tilted linear polarizers at oblique angles of incidence

? Invessel MSE calibrations using a linear rotating stage

    Design of MSE calibration fixture

    Revised MSE calibration based on GA experience

    Memo 26: Reflections near the Brewster angle

    Memo 19: Preliminary polarizer calibration results

    Memo 27: Polarizer calibration of April 28-29

    Memo 30: Initial results from In-vessel MSE calibration (June, 2004)

    MSE Calibration Preliminary (January 2005)

    Final MSE calibration Results (February 2005)

    Revised final MSE calibration results (February 2005)

? Measurements of filter bandpass response

    Memo 24: Analysis of beam-into-gas (no TF)

? Results of beam-into-gas calibrations

    MSE calibration 2003 preliminary (May 2003)

    MSE calibration 2003 (June 2003)

    Summary of MSE calibration of January 21, 2004

    Summary of MSE calibration of March 4, 2004

    Revised summary of MSE calibration of March 4, 2004

? MSE measurements in plasma

    Unusual behavior of measured polarization angle profile at high density

    (October 2002)

    Shot-to-shot scatter increases with plasma density (October 2002)

? Nonlinear dependence of MSE signal strength on DNB current

    MSE signal strength vs dnb current

? Mechanical problems

    Disruption stresses on mirrors (December 2002)

    Proposal to use metallic mirrors for MSE (December 2002)

    MSE mirror repair (January 2003)

    More on Metallic Mirrors (February 2003)

? MSE spatial resolution and DNB aperture

    Preliminary DNB aperture proposal (February 2003)

    DNB size: what we can do about it (April 2003)

    DNB collimating aperture: preconceptual design (April 2003)

? Loss of MSE signal intensity

    MSE intensity study (June 2005)

    Effect of PMT voltage on sensitivity (July 2005)

    Recent studies of MSE signals (July 2005)

Report this document

For any questions or suggestions please email