Abstract :  Details about the conditions and gradual improvements of pulses from #227. This pulse is the first after some improvements in various systems.

  List of pulses and main parameters

   The next are pulses after the improvement of the data acquisition system, vacuum pump and ECRH. In general de objective is the test of the new systems and the improvement of plasmas. The ultimate objective is to prove that the stellarator fulfil the specifications, mainly the calculated plasma T and n .

   The ECRH system was calculated and tested in  [1] and another publications cited in  [1].

   The pulses from #211 to #226  are explained in  [2]. The first plasma was obtained in this group of pulses.

Meaning of columns:

# = number of order of the pulse

t = pulse length in seconds

B= Magnetic field at axis  at the beginning of the pulse (about 8% less at the end )

It  = kAmpere-turns in each coil

P = Pressure in the vacuum vessel in milli-Pascals at the beginning of the pulse

Pp = Pressure in the vacuum vessel in milli-Pascals during the plasma life. 

Wf =   Forward power from the directional coupler.  In italics (mV)

Wr =  Reflected power (W). In italics (mV)

Contractions:

Simul = The pulse was simulated later to compare with experimental data.

Previous plasma pulses in  [2]

 Results and improvements

Pulses #255 to #260

    Conditions of the experiments:

* A Langmuir probe has been installed in the plasma on a middle point beween the plasma centre and the plama edge. The Langmuir probe meassures 1.4mm length and 0.42mm in diameter. It is a second hand good quality probe.  +-15V are applied to the probe, one second +15V and -15V the other second.

* The current of the Langmuir probe is taken by means of a resistor of 10Ohm in series in the probe circuit. A multimeter is used for this test. The electronic circuit for the A/D and I/O cards will be designed and assembled.

* A 'capillary valve' has been installed to introduce gases and adjust the gas pressure in the VV. It is a mixture of a capillary and a miniature silicone tube with variable compresion that trhothe the capillary.

* Objectives of the pulses :

- Obtain an order of magnitude of the probe current to define the electronic circuit to receive the signal and produce the sweep of voltage for the probe.

- Test the hypotesis of good plasma in the range 

10¹⁸m^-3 to 10¹⁹m^-3  .

 * Experiments produced at neutral pressure of 18mPa and 29mPa equivalent to a density of neutrals of  2.9e18m-3 and 4.7e18m-3.

Results :  

- A plasma is obtained in all the pulses except for the  pulse #259. It is produced with a different adjustment of the stub tuner and at 18mPa. More experiences are necessary to understand the effect of the stub tuner.

-  The regulation of the pressure is possible and accurate with the 'molecular valve'.

- Current are measured with the multimeters as a test,

Pulses #255 to #258 : A value in the interval 50microA to 60microA has been obtained in all the pulses for U=+15V. The negative current could not be observed.

Pulse #260 : A probe  current of +190microA (at + 15V)  and  -30microA (at -15V)

- Some first calculations have been done to test the expressions to obtain Te and ni (pg 238 in Dolan) but at least one more point of the curve I(U) is necessary to calculate the temperature and density.

- Pulse #260  :  The plasma extinguish exactly when the TF coils are set to OFF. It further proves that there is resonance heating.

Necessary improvements and next experimental set-up:  

*  Finish and install the electronic circuit and lines for the  data acquisition of the signal from the Langmuir probe.

* Obtain Argon as a first gas. Helium the second, hydrogen also possible. Now air is used for the pulses.

* Calculate Te and ni  of the plasma.

* Exchange the stub lengths of matching.

Pulses #248 to #254

    Conditions of the experiments:

* Experiments at pressure from around 10mPa to 30mPa.

* 7 batteries, B produce resonance heating at half of the pulse. 

* The double Stub Tuner is installed, previous to the  Coupler. The L2 of the stub tuner (at hole nº 1) is 20.0mm external  + 26mm internal. The L1 (at hole nº 3) is on 10.0mm external height.

* The visible spectroscope is installed in the 3 last pulses to test its feasibility.

Results :  

- Very probably, the plasma is easily started at pressures from 2 x 1018m-3 to 1 x 1019 m-3. The reason is unknown.   A plasma is not created in all the cases with higher vacuum

-  The instabilities are caused likely by the pulsing RF heating and it was analysed in the previous pulses.

-  The antenna needs to be improved or a stub antenna, used previously, could be reinstalled.  Maybe it could be useful to separate the plasma from the internal coaxial line. The HN connector can be installed at the bottom of the port although the space is very reduced. Thus the coaxial line will be outside the vacuum and  the plasma will not appear inside the coaxial line.

   - Plasmas are created regularly. It is a first condition to try to improve the plasma. 

- Pulse # 254 : A faint spectrum in the visible has been recorded. It is only a test and the focus of the assembly is not perfect. In spìte of this, the intensity of light is enough to have a faint spectrum. It is concluded from personal information received from researchers of the CIEMAT [4] and from [5] that the visible spectrum is not the best way to obtain plasma parameters, specially temperature and density. So the spectroscope system will not be improved now.

More details in other next publications.

Necessary improvements and next experimental set-up:  

*  Change the antenna. A new stub antenna or an internal connector. 

*  Install a miniature valve to introduce different gases in a controlled way. It should control the plasma pressure as well. The diffusion pump has no serious method of pressure regulation. Now the manual valve is used to regulate the VV neutral pressure but it is inaccurate.

*  Start the design of the Langmuir system and order the components. It will be the only method to measure the T and density.

* Obtain Argon as a first gas. Helium the second, hydrogen also possible.

Pulses #240 to #247

    Conditions of the experiments:

* 7 batteries, B produce resonance heating at half of the pulse. 

* The double Stub Tuner is installed, previous to the  Coupler. The L2 of the stub tuner (at hole nº 1) is 20.0mm external  + 26mm internal. The L1 (at hole nº 3) is on 10.0mm external height.

* High vacuum.

* Both Reversed ('Herotek') and Forward ('Pasternak') power is recorded.

* Sampling rate is 500 samples/second now.

* Objective : Improve and better understand the plasma ans systems

Results :  

- Important discovery : The pulsing (50Hz cycle) power from the microwave oven is creating, very probably, most of the plasma instabilities.

   It was observed in pulse #246, and later in others, a periodical variation of the plasma intensity. Previously it was considered a random behaviour of the plasma but partially it is caused by the pulsing RF power in combination to the recording speed of the camera. Now it is clear that the same general behaviour occur even in the intervals with apparent random instabilities.

   The sequence of frames, recorded at 30 frames/sec, are observed as :  low plasma intensity, high, high, low, high, high, low... The RF power is a senoidal signal from 0 to Max. at 50Hz due to the half wave rectifier. The sampling rate of 30Hz (33.33ms) together with the 50Hz (20ms cycle) results in the recorded images. The plasma characteristic time is very low in UST_1 and 20ms is much larger than it so the oscillations are observed.

-  The maximum of 10 samples of Reversed  and Forward RF power is an important parameter. 10 samples at 500samples/s is a 50Hz cycle. The average of 10 samples was used before but some phenomena are hidden.

A series of pulses starting from low vacuum to high vacuum is produced :

-  Pulses #245 :  AIMS gauge at 9.95V =  ~500Pa. The plasma is a middle point between pulses #231 to #233 and the high vacuum plasmas.  A phenomenon is observed : When the internal RF coaxial line of about 5cm length is full of plasma the effective RF power is reduced approximately 3 times. It is a negative effect. When it happens good plasma does not appear.

-  Pulse #246 :  The effect of the pulsing RF power was discovered in this pulse. It was produced at neutral pressure = 15mPa equivalent to a density of   2.4 x 1018 m-3 at 298K. Curiously it is in the gap of best plasmas for pulses #223 to #226. However plasmas at high vacuum (about  1017 m-3) were sought recently. Perhaps pressure corresponding to AIMS around 9V should be sought.

- The PC pressure signal is accidentally not recorded but it is recorded manually, as always.

Necessary improvements and next experimental set-up:  

*  Work around AIMS = 9.2V.

*  In a long term, improve the power supplies of the microwave. An excellent explanation in [3] but it needs some time.

*  Reconnect the AIMS signal.

*  Change the hard disk to reduce the periods of saving data. The input is stopped when saving.

* Delay the TF  ON to have no B field at the start-up of the RF power in order to obtain a non-resonant RF heating at the beginning and at the end of each pulse.

Pulses #234 to #239

    Conditions of the experiments:

* The double Stub Tuner is installed, previous to the  Coupler.

* The diffusion pump is ON, so high vacuum.

* 6 batteries.

* The two Schottky Diodes Detectors are installed. Multimeters are measuring voltage and also the reversed power is connected to the A/D cards. The 'Herotek' is at the reverse power and the 'Pasternak' at the forward power. Only reversed power signal is received.

* Objective : Test and improve the A/D and ECRH systems

* TF coils are ON except for pulse #234

* Sampling rate is 500 samples/second now.

Results :  

- The end of the pulses have a TF current of ~8% less than at the beginning of the pulse (due to heating of the TF coils). So B should be adjusted to produce resonance at half of the pulse. 

-  A small  interference on the AIMS gauge signal remains.

- The apparent interference on the RF power signal was really a software 'interference'. It was suspected on the previous pulses. Really the ~50 samples/s produced a low frequency signal which was the composition of the ~50Hz pulses of the RF from microwave oven (it is common in all normal MW ovens with a diode rectifier) in combination with the sampling rate. Now the sampling frequency is 500samples/s and the 50Hz cycles are distinguished. It will be shown in a next publication.

-  Pulses #234 - #236 :  The L2 of the stub tuner (at hole nº 1) is 30mm external  + 26mm internal. The L1 (at hole nº 3) is on 0mm external height.

- Pulse #236  : A plasma with intense visible light radiation is created,  exactly during the period with TF ON. It indicates that there is resonance heating. The L2 of the stub tuner (at hole nº 1) is on 19.3mm height, the L1 (at hole nº 3) is on 9.3mm height, the same for pulses #237 and #238. The input data and graphs for this pulse will be shown and analysed in a next publication.

- Pulse #237  : The reversed power is higher than the previous. No plasma is produced. It is not clear yet why some pulses produce a plasma and others not while the conditions and the input data from the sensors is very similar.

- Pulse #239 : The TF current has decreased slightly. B on axis is only 41mT and on the outboard plasma edge only 38mT. It seems far from the resonance at 43mT. The L2 and L1 of the stub tuner are located randomly in this pulse.

- The only difference observed between the pulse #236, with plasma, and the pulses with no plasma is that in the pulses #235 - 237 - 238- 239  a faint light appeared (in general a weak plasma inside the RF coaxial line inside the VV) and it produced a small increase in the neutral pressure. In each pulse the behaviour is very similar. It did not happened in pulse #236. In this pulse the plasma was produced and the pressure did not increase in an appreciable manner.

Necessary improvements and next experimental set-up:  

*  Calibrate the AIMS gauge signals.

* Remove the delay of 1 second the TF ON rise. Perhaps the start-up of the plasma is more regular. Try both methods to know which one produces more proportion of pulses with a plasma.

*  Try to produce plasmas in a regular manner.

* Start the Langmuir probe system.

* Adjust B to produce resonance at half of the pulse. 

Pulses #231 to #233

    Conditions of the experiences:

* The double Stub Tuner is installed, previous to the  Coupler. The L2 (at hole nº 1) is on 30mm height, the L1 (at hole nº 3) is on 0mm height.

* The diffusion pump is OFF, only the mechanical pump..

* 6 batteries.

* The two Schottky Diodes Detectors are installed. Multimeters are measuring voltage and also the reversed is connected to the A/D cards. The diodes are interchanged from the last pulses, so the 'Herotek' is at the reverse power and the 'Pasternak' at the forward power. This is useful to control the usual high inaccuracy of these detectors.

* Objective : test and improve the A/D and ECRH systems

* A new Type N connector with Teflon (TFE) insulator is manually made and installed. New good quality Type N connectors are ordered. Teflon withstands up to 270ºC before melting while PE only around  125ºC.

* The coax. line and connectors from the coax-to-waveguide adapter to the stub tuner has been disassembled, revised, improved and reassembled. The coax-to-waveguide adapter has been cut 10mm to have a straight connection to the coax. line.

Results :  

- The new connector seems to withstand the high RF power.

- No RF leaks are observed. A mobile phone at 5cm from the RF leak detector emits from 10 to 100 more intensity.

- Pulse #231 and #232 : The TF coils are off. A plasma near the antenna is created. Because the vacuum is low the plasma is near the antenna (low electron mean free path). The microwave oven takes more than 2 seconds to reach full power. 3 seconds will be OK.

- Pulse #233 : The TF coils are ON. A plasma near the antenna is created.  No arcing inside the coax-to-waveguide adapter. The intense plasma only starts exactly when the TF coils are switched OFF -strange behaviour.

Necessary improvements and experiments:  

*  Connect the remaining  RF power signal to the A/D cards. 

* Produce pulses with high vacuum.

* Delay 1 second the TF ON rise.

* Increase the sample rate to discover if the RF interference is a software one (220V ~50Hz line, with ~50Hz sample rate)

Pulses #227 to #230

    Conditions of the experiences:

* The diffusion pump is OFF, only the mechanical pump..

* 6 batteries.

* The two Schottky Diodes Detectors are installed.. Forward detector may be partially broken. Multimeters are measuring voltage. 

* The double Stub Tuner is installer, previous to the  Coupler. The L2 (at hole nº 1) is on 30mm height, the L1 (at hole nº 3) is on 0mm height.

Results :  

- Pulse #227 - #228 : A plasma near the antenna is created. Because the pressure is high this behaviour is correct. The objective of this pulses is to improve the systems.

- Pulses #229-230 :  Pulse #229 finished with arcing inside the coax-to-waveguide adapter. Pulse #230 create no plasma at all and arcing and sparks were intense. The adapter was inspected and the Type N connector melted. This particular Type N connector was tested and the plastic is of very low quality, it melts at very low T in comparison to Teflon. The central conductor of the line was short circuit with ground inside the coax-to-waveguide adapter.

Necessary improvements :  

*  New high quality Type N connector will be ordered.

* The length of the coax-to-waveguide adapter will be reduced 10mm in order to have more straight and correct connections. 

* Cable RG393 would be better for the coaxial line near the coax-to-waveguide adapter.

* Two Schottky Diodes Detectors for the forward power and for spares will be ordered.

* A flat led will be installed to increase the precision of the synchronism between the cammera and the control and data acquisition PC. Now a halogen bulb is used. It has a slow response.

 References

[1] "ECRH system in UST_1 and first tests" , Vicente M. Queral. See List of all R&D
 

[2]  "Pulses from #211 . Improving the plasmas, ECRH and vacuum systems", Vicente M. Queral. See List of all R&D

[3] "Simple microwave preionization source for ohmic plasmas" W. Choe et al. REVIEW OF SCIENTIFIC INSTRUMENTS VOLUME 71, NUMBER 7.

[4] Personal comunication .  E. C.

[5] "Temperature-Dependent EUV Spectra of Xenon Plasmas
Observed in the Compact Helical System" SUZUKI Chihiro, NISHIMURA Hiroaki, OCHIAI Masayuki, KATO Takako, et al.

Date of publication 03-06-2007. Continuous updating