where M&C refers to monitoring and control. Credit: Andrew Williams.
Layout of the dipoles on the tile:
0 1 2 3
4 5 6 7
8 9 10 11
12 13 14 15
The dipoles are also referred to as A through P, in the same order. This order is used for beam codes; i.e. delay corresponds to dipole A in the top left-hand corner.
Dipole delay values are integers from 0-31, representing the total delay in units of 435.0 picoseconds. This is implemented using analogue switches that send the signal through zero, one, two, three, four or five loops of track on the PCB. The shortest track adds a delay of 435 picoseconds, and the delay from each successive track is twice as large as the previous one. Note that manufacturing and component tolerances mean that the actual delays vary from beamformer to beamformer, more so for dipoles using the longer delay lines.
See the MWA 'sweet spots' and gridpoint numbers page for a description of how delay values are chosen for an observation, given RA/Dec, Az/El or other source position.
A special delay value of 32 turns off the entire dipole signal chain at the final summing junction, so no signal from that dipole contributes to the final tile output. This delay is applied automatically by the M&C system when a dipole is known to have a bad band shape.
This photo of a beamformer delay board shows the loops of PCB track used for delays, and the analogue switches (small black chips) used to direct the signal around the loops, or straight through a short impedance-matched bypass.