Message 7/486 dave                                       Oct 16, 2000 02:02:51
pm +0200

Subject: EVNtech: IF Switch report for TOG
To: evntech@jb.man.ac.uk
Date: Mon, 16 Oct 2000 14:02:51 +0200 (CEST)
Sender: owner-evntech@jb.man.ac.uk
Precedence: bulk


  Report on IF switchbox project

(D. Graham 16.10.00. Sorry for late report with little content)

AIMS: Allow flexible IF switching in MkIV terminals, implementing the 'patch'
  command to connect video converters to IF1 high/low and IF2 high/low,
  preferably also to IF3 (geo high IF for R and D frequencies).
  This would make possible easy repatching to allow astronomy at SX using
  geo stations and geo at astronomy stations.

INPUT so far: received from several people, useful comments
  from Brian Corey stating that isolation between IFs must be at least 60dB,
  implementation suggestions from Ari Mujunen.

RADICAL IMPLEMENTATION: One possibility is a complete rebuild of the
  IFD unit using modern, cheaper components, 4 IF channels instead of two
  to allow dual-channel SX, flexible switching on the output. This is
  rather a large project and difficult to prototype.

SMALLER-SCALE IMPLENTATION: keep existing IFD and existing 7-way splitters,
  Construct 7*switching box, each has inputs from 3 IFs, Low and high,
  6 in all. Internally inputs 1L, 1H and 2L , 2H are split with
  Minicircuits ADP-2-1W (USD 7 each), 1 output of each splitter goes
  via SPDT switch MSWA-2-20 , 3L and 3H inputs without splitter via SPDT
  switch. These signals go to an SP6T switch made of 2* GSWA-4-30DR to
  output #1. Output #2 is selected by a GSWA-4-30DR from the other splitter
  signals from 1H, 1L, 2H,2L. Isolation requirements are met by having
  at least 2 open switches between any IF inputs, therefore 70dB isolation
  in switches. Good layout needed to prevent coupling between signal
  lines. A cheap SMD broadband amplifier is needed on each output to
  make up for about 10dB loss in the switch. Serial commands from the
  FS would be interpreted by a PIC micro in each box. The #1 and #2 outputs
  from a typical box would e.g. go to VC1 and VC8, VC2 and VC9 ...VC7 and
  VC14, arranged so that all the VCs that need IF3 can get it. Each box
  would have a switch giving the identity of the VCs it is connected to.
  The estimated cost of each box is about USD 200, including 2 confectioned
  SMA cables per box, so USD 1400 per station. A switching box with only
  4 inputs would be about USD 30 cheaper.
  This approach has the advantage of being easier to prototype than the
  IFD replacement, but is really not optimal either.

PROBLEMS: First of all who would prototype and build, secondly to
  modify sched program to generate correct schedules for those stations
  which have this unit.