WP6: Active element

February 12, 2010
This Work Package of the EISCAT_3D Design Study covered the design of the active component of the radar including the generation and distribution of the RF signal, the transmitting antennas and the necessary control and monitoring systems.

A summary of the conclusions from this Work Package follows:

Active element power amplifiers

  • For the full-size EISCAT_3D system, an active subsystem based on a more powerful, higher-frequency, higher-gain FET device than the BLF248 is recommended. The BLF369 may be a good choice; this device would provide at least 2 dB of dynamic headroom above the 300-Watt level and would thus normally operate in an essentially linear part of its transfer function.
  • To minimise temperature-dependent output variations, the low-level stages should be temperature compensated and an automatic gain control loop should be added around the whole amplifier chain.

Active element transmit/receive switching performance requirements

A T/R switch constructed with UM4906 and UM7006 PIN diodes and low-loss λ/4 line sections and bias chokes, will meet or exceed the EISCAT_3D target specifications for:

  • TX-to-RX attenuation (required ≤-69.2 dB, achieved ≤-90 dB)
  • TX loss (required ≤-0.1 dB, achieved ≤-0.02 dB)
  • RX noise-contribution (required ≤25K, achieved ≤24.3 K)

Active element antenna array

  • The coupling between nearest-neighbour linearly polarised antennas at mid-band is on average about –25.3 dB. In circular polarisation, it is on average –29.0 dB, i.e. about 4 dB less than the linear polarisation value. However, the circular polarisation values are still almost 10 dB greater than the NEC-derived values for antennas mounted at 0.7 λ above real ground; these range from a maximum of –39.9 dB to a minimum of –61.7 dB.
  • While a mutual coupling between nearest neighbours in the order of –29 dB is may be manageable from the transmitter point of view, the observed discrepancy between the NEC values and the actually measured ones is not presently well understood and sufficiently great to warrant a careful second look at the issue, including more detailed NEC runs.

The main work in this Work Package was performed by the Swedish Institute of Space Physics.