ECE spectrum of HSX plasma at 0.5 T
K.M.Likin, H.J.Lu, D.T.Anderson, F.S.B.Anderson, J.M.Canik, C.Deng1, C.W.Domier2, R.W.Harvey3, J.N.Talmadge, K.Zhai
HSX Plasma Laboratory, University of Wisconsin, Madison, WI, USA; 1UCLA, CA, USA; 2UC, Davis, CA, USA; 3CompX, Del Mar, CA, USA
1. ECE Radiometer
Fc=12.9 GHz
F=0.4 GHz
8-way
Power
Divider
Fc=12.6 GHz
F=0.2 GHz
20 dB
20 dB
Fc=12.2 GHz
F=0.4 GHz
Fc=11.8 GHz
F=0.4 GHz
20 dB
20 dB
Fc=11.4 GHz
F=0.4 GHz
R, m
Band-Pass
Filters
20 dB
Crystal
Detectors
Video
Amplifiers
Input
Low Field Side
Frequency (GHz)
Frequency (GHz)
Isolator
Directional Coupler
•Network analyzer is used as a
microwave source in CW regime
•The frequency is fixed within the
band width of each channel
Directional Coupler
•Calibration has been made for two sets of eight IF filters --- one set is for
receiving the electron cyclotron emission from the low magnetic field side and
the other one – from the high magnetic field side
•Sensitivity of each channel has been measured at different input power, i.e. at
2.5 mW, 5 mW, 7.5 mW, 10 mW and 15 mW, so that the output signal varies from
0.5 V up to about 5 V, respectively
•Sensitivity has been measured at various frequencies within the band width of IF
filters
Frequency (GHz)
•20 dB Directional Coupler has been
calibrated with a network analyzer, i.e.
transmission is measured into the
forward and arm sections
•8-way power divider followed by IF
filters have been calibrated with a
network analyzer separately for each
channel, i.e. a transmission coefficient
is measured in 1 GHz frequency
intervals within the corresponding
frequency band
Coeff. (keV / V)
Low Field Side
High Field Side
DWS Channels
UPS Channels
•Low frequency (< 70 kHz), narrow bandwidth (< 5 kHz), low amplitude (less than
5%) fluctuations have been measured by the ECE radiometer in QHS
•Amplitude is almost independent of
•The phase difference between ECE
the plasma density
signals on the either side of the
Low Ne
High Ne
magnetic axis is about 180 degrees
Power Spectrum Density
Frequency vs. Ne
Frequency (GHz)
•The integrated response has been calculated assuming that the plasma is a
black body. In these calculations an etendue of the horn antenna is estimated to
be 5.84 cm2*sterad
•The central frequency has been calculated based upon the measured sensitivity
of each channel
Frequency (kHz)
Frequency (kHz)
ECE
Data
Tece (keV)
Frequency (GHz)
•Solution of the radiation transport equation is used to find a
population and a temperature for energetic electrons
•Bulk plasma density and electron temperature are from TS
• The best fit is for the following parameters:
Ntail = 0.1∙1018 m-3 in QHS, 0.2∙1018 m-3 in Mirror and
Ttail = 2.5 keV for both modes
5. CQL3D code
3.3 Plasma Density Fluctuations
Channel Number
Frequency (GHz)
4.2 Mirror Spectrum
Frequency (GHz
r/ap
r/ap
P (a.u.)
Power (dB)
2.3 Integrated Response
Coeff. (keV / V)
Power Meter
F (kHz)
ECE
Part
P (a.u.)
Scope
Network analyzer
High Field Side
S (V / mW)
ECE Radiometer
1 2 3 4 5 6 7 8
S (V / mW)
+12V -12V +12V
#2
#1
2.2 Measured Sensitivity of the Channels
•TS and ECE data are shown in the quasi-helically symmetric (QHS) and 10% Mirror magnetic configurations in the case of
the central resonance (CR) heating
•Plotted ECE data are based on the calibration presented above
Te in QHS, CR
Ne in QHS, CR
•ECE temperature at UPS frequencies is almost the same in
QHS and Mirror while there is a clear difference at DWS
UPS Channels
DWS Channels
frequencies between the two configurations
Measured ECE Spectrum
•Higher ECE
DWS Channels
UPS Channels
Mirror
temperature in
QHS
the Mirror mode
r/ap
r/ap
is due to a lower
Ne in Mirror, CR
Te in Mirror, CR
central plasma
Fg
density in Mirror
UPS Channels
as compared to
DWS Channels
QHS
ECE
Data
Frequency (GHz)
3.2 ECE Spectra in QHS and Mirror configurations
2. Calibration
2.1 Lay-out on the bench
4.1 QHS Spectrum
Tece (keV)
20 dB
Tece (keV)
Fc=13.2 GHz
F=0.2 GHz
20 dB
Relative Amp. (%)
Fc=13.5 GHz
F=0.2 GHz
Horn
Antenna
Vacuum
Vessel
Flux
Surfaces
20 dB
To Data Acquisition System
Z, m
Fc=13.8 GHz
F=0.2 GHz
Tece (keV)
Window
Relative Amp. (%)
32 dB
Te (eV)
22 dB
IF Amplifier
•ECE channels have been also calibrated against the electron temperature measured by Thomson scattering (TS) diagnostic in
the plasma with off-axis heating
•In order to minimize a contribution from high energetic electrons to the ECE signals, the resonance is located inboard at
r/ap = - 0.2 in case of the down-shifted (DWS) frequency channels while the outboard heating is made at r/ap = + 0.2 for the
up-shifted (UPS) channels, respectively
ECE Spectrum
Plasma Density
Electron Temperature
•TS and ECE data are
averaged over 5 shots
ECH @ r/ap = +0.2
ECH @ r/ap = +0.2
•Optical depth is
ECH @ r/ap = -0.2
calculated based on the
ECH @ r/ap = -0.2
TS profiles
•ECE temperature is
corrected on the
calculated optical depth
r/ap
r/ap
Frequency (GHz)
Te (eV)
IF Amplifier
•The spectrum of the Electron Cyclotron
Emission (ECE) from HSX is measured
with an eight channel radiometer at a
magnetic field of 0.5 T
•The central region of the plasma is not
accessible because the plasma is heated
by the extraordinary wave at the second
harmonic
•60 dB stop-band filter at 28±0.3 GHz is
used to suppress the non-absorbed
microwave power
•A low-pass filter with 40 dB attenuation
in the band of (50-95) GHz reduces
pick-up from the heating source at its
second harmonic
Te (eV)
40 dB @
F > 50 GHz
Local
Oscillator
@ 42.5 GHz
P (a.u.)
Fc=28 GHz
F=0.6 GHz
Mixer
Diode
ne∙1018 (m-3)
Pin
Switch
Diode
ne∙1018 (m-3)
1
B 3
Quartz R
Waveguide
Line
LowPass
Filter
P (a.u.)
Mod B
60 dB
Band-Stop
Filter
3.1 ECE vs. Thomson Scattering
ne∙1018 (m-3)
1.2 Block diagram
1.1 HSX Cut
4. Bi-Maxwellian
plasma model
3. Results of measurements
<Ne>∙1012 (cm-3)
Channel Number
5.1 Maxwellian Electron Distribution Function
5.2 Non-Maxwellian Distribution Function
•The frequency
decreases with
plasma density.
See the HSX
posters by
C.Deng & S.Oh
47th Annual Meeting of the Division of Plasma Physics, October 24-28, 2005, Denver, Colorado
•CQL3D code is used to simulate the
electron cyclotron heating in QHS and to
calculate a radiation intensity at the second
harmonic of wce
•On the left figures the results of two runs are
shown
•The plasma parameters are as follows:
central Te = 0.4 keV, central Ne = 1∙1018 m-3
and input power is 100 kW
•On the left bottom figure it is clearly seen
that the ECE temperature for the nonMaxwellian electron distribution function is
above 4 keV
•Further work should be done to understand
why CQL3D predicts low ECH absorption in
HSX plasmas