TRACO DCDC Converter Startup Tests
        ------------------------------------


                                  Initial Rev.  1-May-2025
                                  Current Rev.  5-May-2025


There are at least 2 aspects of the operation of the TRACO
96 VDC to 5 VDC DCDC Converter that need to be verified by
testing them:

1. When the TRACO converter first starts up its output is
   looking into a 99% reactive load.   Recall that the main
   load on the TRACO Converter is the 6 DCDC Buck Converters
   that supply all of the actual operating voltages to the
   circuits on the DK Board.  The input to each of these
   converters includes an LC Filter that has a 10 uH inductor
   and 4x 150 uFd Tantalum caps plus 4x 10 uFd Ceramic caps
   all 8 caps in parallel to Ground.

   These LC Filters go series resonant at about 2 kHz or at
   about a 0.5 msec period.   There are 6 of these series
   resonant filters all in parallel basically shorting the
   TRACO output to Ground at the 2 kHz frequency and thus
   potentially causing a serious problem with its feedback
   servo loop.

   The Q of these filters at 2 kHz is about:
   2 kHz x 2pi x 10 uH  /  50 mOhm   =  about  2.5

   This Q is low enough so that there will not be a
   significant noise peak at 2 kHz.

   The 50 mOhm is made up from about 41 mOhm DC copper
   resistance in the inductor winding, and a minimum of
   10 mOhm in the 4 terminal current shunt series resistor.
   Note that there are other losses so the Q should actually
   be lower, e.g. the winding's  AC copper resistance at
   2 kHz,  most of the current shunt resistors are actually
   20 or 25 mOhm, and hysteresis losses in the inductor.

   For the first approximately 1 second of operation of
   the TRACO Converter it has very little Real load.  After
   charging them it sees just the Reactive load of these
   6 series LC Filters all connected to the TRACO's output
   in parallel.

2. When the DK board is operating the Real DC load on the
   TRACO Converter is expected to be about a steady  1.5 Amps
   from our circuits on the DK board.  But there can be
   additional loads of approximately  1.5 Amps  that could
   switch ON or OFF with fast transition times.  Will these
   fast approximately 2x changes in real load current upset
   the TRACO Converter when it also has the large reactive
   load from the LC Filters on it ?    Needs to be verified.


In the following tests with scope shots labeled:  traco_...
or linear_...   the following loads were used:

 - Always connected to the output of the supply is:
     4x LC Filters of   600 uFd Tantalum in series with 10 uH
     2x LC Filters of  1000 uFd Aluminum in series with 10 uH
   All 6 of these are in parallel across the supply's output.

 - Sometimes a DC load of  1 Amp  is used:  this is 2x 10 Ohm
   power resistors in parallel.

 - Sometimes a dynamic load of about  1.3 Amps is used:  this
   is a 3.8 Ohm power resistor in the drain of an  IRF540ZPBF
   power FET that is turned ON & OFF by a bipolar square wave
   from an HP function generator.


Startup testing was done both with and without the 1 Amp DC
load also connected to the supply's output.

Dynamic testing was done both with and without the 1 Amp
DC load also connected to the supply's output.

The Linear supply  (Power-One 5V 12 Amp)  is used
only with the initial testing of DK board  SN #0.

So far both the TRACO and the Linear supply look stable
at start up and with a sharp load change.