August 2004

Modified Tx/Rx W-band circulator network to handle 5-port pulse-by-pulse switching between 4 antennas. Previously only 2 antennas or one antenna with two polarizations could be switched on a pulse-by-pulse basis and different antenna configurations had to be switched using manual RF switches. New version of the radar control board as well as a new version of the acquisition and radar control software cpas were installed in order to handle the all-antenna fast switching.

The modified network and the existing dual-polarimetric antenna (side/up pointing one) are not yet capable of providing sufficient isolation when using 5-port configuration (dual-side antenna with the side/up antenna using two ports for dual-polarization). Thus currently only 4-port configurations are supported.

March 2003

Installed new version (5.01) of the data acquisition program cpas. Rewritten the code for the control of the ARB function generators (fixed a bug not allowing long prt). Added pulse group time interval for SPPMAG mode. Added experimental code to tackle the time jumps and the synch problems of the tfp card when using an external irig signal.

W-band mixers for H- and V-channels: Repaired mixer is on V-ch.

WCR installation in the trailer completed. 24" cassegrain antenna is connected to H- and V- ports. The transmitted polarization is dependant on the position of the scanning dual-reflector orientation. It is designed to transmit H/V-pol via H/V-port when the beam is horizontal and normal to the axis of the trailer. The manual waveguide switches for the multi-antenna configuration of WCR on KingAir are removed.

PC-Linux host computer was upgaraded from a single-processor to a dual-processor system. This eliminated small data gaps due to OS interruptions.

March 2002

RF unit circulator network H and V ports are reorganized. First they are switched, meaning H-port becomes V and V-port becomes H. Thus the shortest waveguide path (minimum loss) to the 12" conical lens antenna carry H-pol signals. Second. Two mechanical waveguide switches (S1 and S2) are installed on the new designated H and V ports of the circulator network. S1 is a E-plane two-port switch and is used to connect to the 12" conical lens antenna (side/up pointing beam) and the 15" casegrain antenna (down forward pointing beam). S2 is H-plane (V-pol), 3-port switch and may connects to the 12" conical lens antenna (via OMT), 12" cassegrain antenna (side forward pointing beam) and 18" cassegrain antenna (nadir pointing beam). From the calibration stand point this configuration not only has the receivers swapped for H and V but have new waveguides and new additional antennas. Every antenna configuration has its own set of calibration coefficients.

VXI embedded controller (HPUX workstation) is replaced with external controller based on PC Linux OS. The GUI of the data acquisition is modified and upgraded to work on the Linux box. Few minor changes to the data header are introduced. The DSP downloadable routines (radar algorithms) are unchanged.

September 2000

The analysis of the MITEQ MLIF detectors showed (see Pulse-mode receiver calibration (pdf file)) that they exhibit higher non-linearities than MITEQ LIFD type detectors. Thus the original LIFD detectors were repaired to resolve the power slope dependance and were installed in June 2000. New non-linear lookup tables were installed as well.

March 2000

In March 2000 the V-channel log-detector was also replaced with a MLIF type detector. New lookup tables were prepared and used during WYICE 2000 Apr-May.

September 1999

In July 1999 the H-channel logarithmic receiver was upgraded with a new logarithmic detector (MLIF). The V-channel receiver will be upgraded by the end of 1999. New lookup tables (calibrating the receivers) were created and thus the calibration constants starting 1999 are very different and cannot be compared with the old ones. It was discovered that the old H-channel log detector (LIFD) and the current V-channel log detector do not perform, under pulsing mode, as expected (see Pulse-mode receiver calibration (pdf file)).

Information pertinent to the experiments before 1999

Receivers Imbalance (Hch-Vch received power) is about -5 dBm but vary slightly from experiment to experiment and is included in the calibration constants.

WCR Internal LookUp table for H-channel is incorrect for the experiments between 1995 and 1999. It is corrected in the post-flight processing software. The correction is with respect to the V-channel and is linear for the power given in dBm.
     The Slope is 0.95 (corrected before calibration)
     The Intercept is included in the cal. constants as part of the gain imbalance.

Cross channels are not calibrated for the experiments between 1995 and 1999 and the following assumptions are made:
     hvcal=hhcal, vhcal=vvcal and vhiso=hviso,
where {xy}cal is xy-channel cal. constant and {xy}iso is xy-channel isolation.

The two-way isolation measured from drizzle in May '97 is better than -17.5 dB.
In- and Out- isolations are assumed equal. Further analysis (May'98) with experimental data showed that the relative difference between VH and HV isolation is less than 1dB and that |vhiso| > |hviso|. In some cases weather data is used to fine tune hviso and vhiso.

Internal calibration channels of the WCR are not used for calibration and were disabled in 1998. They exhibit significant dependence on the transmitted power leakage. After 1999 the internal calibration is implemented by using additional W-band detector for monitoring the transmitted power. Installation of a noise source for receivers' calibration is in consideration.