[evlatests] First Ka-Band Receiver on the EVLA

[Paul Harden]

Robert Hayward wrote:
> As predicted, there was not enough LO Ref power available at the output 
> of the UX Converter (we measured about 1 dBm at the receiver input, we 
> really need about +6 to +9 dBm to keep the downconverters happy).

The original PDR-era plan and system specs was to have +10 to +13dBm of 
LO1/LO2 power arriving at the UX Converter and for the receivers.  After 
the first couple of EVLA antennas, it was obvious the loss through the 
LO cables and switches from the IF/LO rack to the UX Converter/receivers 
was several dB more than anticipated.  The Sythesizers are running at 
near full output.  As a result, I had to add additional LO amplification 
to the UX Converter to drive the LO Multipliers - which fail to operate 
properly with an LO less than about +3dBm (and certainly by 0dBm).

The EVLA system specs were changed June 2005 by the Change Board to:
LO1/LO2 input to UX Converter: +7dBm -3/+4dB (+4 to +11dBm, 12-14GHz)
This was to satisfy the +7dBm nom. LO to the UX Converter and the 0dBm 
nom. LO power required by the receivers.

The statement "there was not enough LO Ref power available at the output 
of the UX Converter" is misleading in that the LO and UX Converter are 
doing their job per EVLA specs.  It is delivering 0dBm nom. to the 
receivers.

 From the antennas I have measured, there is typically +5dBm LO arriving 
at the UX Converter, meeting LO specs.  Accounting for 1-2dB loss 
through the LO1 switch on top of the UX Converter, and 1-2dB cable loss 
to the receiver, +1dBm LO at the receiver input is about right, and 
fully EVLA spec compliant.

Paul Harden