Balmer HtoD History

1
Balmer H-to-D History

• We generally prefer H/D lt 0.05 for good rf
  coupling to the core plasma --- would much prefer
  control with H2 fueling
• It has become difficult to keep H/D low during
  the run (between-shot ECDC usually
  required)
• Ratio tends to go up with rf power, length of
  plasma discharge, and number of shots during run
• There have been long run periods with very low
  H/D (0.02 to 0.04) with little variation
  during the run day
• First shot effect
• Try to identify what changes we have made
  invessel to produce this change in behavior

2
H/D data from 1999 to 2001
t gt 0.5 s Ip gt 200 kA nsum2
8/2000 conditioning
2/2000 conditioning
DNB beamline
3
BN Tile History
Antennas as of 02/16/2000
Antennas as of 07/27/2000
Still no septum
4
BN at E-Port
BN at all antennas
Septum at J-Port
ECDC required to keep H/D below 0.10
5
Solutions?

• Most likely cause of the problem is the addition
  of BN tiles to protect the antennas
• Better ways to bake out the tiles (invessel)
• add heaters to tiles where possible
• low level rf to assist bake
• Better plasma conditioning
• Pre-installation processing
• bake
• backfills other than air
• Some vacuum test stand work is being planned
• BN analysis