Started experiment essentially at the 16:48 UTC change to 4cm.
Recording DAS1 to ATNF V002B (xraid0 on cavsi1).
Tied array was 12345, reference antenna CA03 on W84.
Focussing to default values on frequency change (only CA03 and CA05 affected).
Keeping the same DAS profile for both bands: at16x2_f.pro
Dropped to antennas 345 in the tied array at 17:46:26 since the phase stability on the long baselines at 15mm is poor.
Dropped to antennas 45 in the tied array at 17:48:32, the phase stability continues to deteriorate.
For 4cm frequencies, using 12345 in the tied array.
I can't really tell what happened, but sometime after the initial calibration, the cross-pol phase at 4cm got changed from -90 degrees to +90 degrees. So the RCP/LCP for ATCA will be backwards for 4cm.