Recent changes

• Both mixers (A and B) are now tunerless.
• The problems with the DBI (Dual Beam Inteferometer) have been fixed.

Double-sideband Trec's are about 110-120K for both mixers across the band, so in SSB mode a reasonable Trec would be 200-250K or a little more. Frequencies above about 360 GHz exhibit sharply worse performance for channel A.

Both channels of B3 are capable of observing a window 920 MHz wide. Channel B must be used for wideband work with the spectrometer maximum of 1.8 GHz instantaneous bandwidth as the bandwidth is limited by the HEMT amplifier in channel A.

Note that in wideband mode both IF channels to the DAS are used even though only one mixer is in use; the IFS does not autolevel IF2 and this channel should be set manually to the roughly the same level as IF1 (around 40%) using the attenuator switches.

Suggestions for normal operation

• Double-check the Gunn operating voltage, and set to 6.0 volts if necessary. No need to adjust the DBI or hysteresis any more.
• Use the debug option; to get this, at the bottom of the B3 screen
  • If you don't have the option list, type "Q" in the B3 window.
  • Select O(ther), then
  • D(ebug Auto Tuning)
  The debug field at the lower left of the B3 window then shows the progress of tuning: F(ail), T(une), N(ext), S(ucceeded). A line of all S's is what you want.
• Quite often B3 will tune up without any fuss, however. If adjustments are necessary, rxb_retune is preferred over rxb_auto in general.

Known Problems and Possible Fixes

• "Power output is unstable on IF2"
  • Cause: Instabilities in the mixers
  • Fix: Lower the bias voltage for the unstable mixer either by going up into the cabin and turning it down on the dial at the top right corner of the appropriate mixer control module or via the PF3 screen on the B3 engineering screen. A reduction of approx 0.1V seems to usually cure the problem. We will make a software change at the unstable frequencies once we are confident this has no effect on the receiver performance. For now you should probably take a standard spectra after changing the voltage to check the performance is still OK.
• "Receiver locked but not fully tuned"
  • Cause: unknown
  • Fix: rxb_retune
• Gunn may not lock near LO 346 GHz
  • Cause: PTS unlocked
  • Fix: tune to other sideband or tune from the other direction (i.e. tune to a higher frequency, then come down from there). Also check that Gunn voltage is 6.0 volts.
• Tuning fails because of low mixer currents
  • Cause: software bug
  • Fix: It may be necessary to manually tune the halfwave plate, but first try tuning to a totally different frequency and when you get a good result, retune back to your original frequency. It may work this time and it saves a trip to the cabin.

    If you have to resort to manual adjustments see the last section of the manual tuning procedures for instructions. Basically you will have to use the waveplate to provide the demanded currents. If too little current is available the Level Control pot should be used to provide more. It may well help to leave the modulator module remote/local switch in local mode once one have obtained the correct mixer currents.