[hpsdr] Demeter - Working principle

Jeroen Bastemeijer J.Bastemeijer at TUDelft.nl
Fri Jun 22 06:17:12 PDT 2007


Dear All,

As promised, I would inform you about the working principle of Demeter 
(as it is proposed now!)

It is based on two thoughts:

   1. A switching power supply is more energy efficient compared to a
      lineair power supply.
   2. A linear power supply provides a cleaner voltage and a nicer
      current limit behaviour (something which is important when things
      go wrong during experiments).

 

The configuration consists of a switching pre-regulator and a linear 
post-regulator. This combines the best of two different types of regulators.

 

The Demeter proposal as it is now is based on almost all the wishes. 
E.g. it is possible to power the HPSDR through Demeter from the mains, 
the shack power supply and a battery (mobile/emergency operation). On 
the other hand, the implementation will be quite flexible. E.g. people 
who don't need the battery option can leave out the charger. People who 
will not use the mains supply can reject the transformers, rectifiers 
and some other components.

 

Working principle:

 

The "Heart"of Demeter is based on a switching converter (step-up) with 
number 10. (The numbers refer to the schematic on: 
http://hpsdr.org/wiki/index.php?title=DEMETER) What it does: It boosts 
the voltage of an external power supply or the battery to a level which 
is high enough to ensure proper operation of the 12Volt linear 
regulator. The output of this step-up converter is used as well as input 
for the 5Volt step down converter and the -12V inverter. (The mentioned 
DC-DC converters have no isolation function!).

In case the mains power is used: This voltage will be the proper level 
for the linear regulator. So the DC-DC converter (#10) will be put in a 
power feed-through state.

During mains-supply use a negative voltage is derived from the 
transformer at the same time, so DC-DC converter (#17) will be in 
feed-through state as well. The only switcher active will be the 5Volt 
converter #14.

Blocks #3, 9 and 18 are used as "switches"to choose the "best"available 
voltage. Probably the implementation will just consist of a Schottky-diode.

Each output will be protected with a crow-bar circuit. This will 
probably be a crow-bar using transistors instead of thyristors (sturdy 
thyristors are difficult to get in SMD-housings).

The mains supply part consists of two transformers:

   1. Small transformer used for a "standby-voltage". The GPS and the
      OCXO will be powered. To allow software power-up the USB-hub and
      the PIC will be powered as well.
   2. The main transformer will be switched on by means of an
      (electronic) relay.

 

N.B. The main transformer will be a single secondary winding transformer.

N.B.2 The transformers will be normal transformers. Allthough a switcher 
could be build by amateurs there is a high risk of an electric shock or 
worse. Furthermore, a conventional transformer is more tolerant to input 
supply changes (like a brown-out).

 

Control and monitoring of the power-supply is done by a PIC. Through a 
multiplexer and a single DAC it will read the various voltage and 
currents. This will be helpful for (remote) monitoring and 
trouble-shooting the power supply. Furtermore, the PIC will be 
controlling certain parts of the power supply (e.g. switching the DC-DC 
converters in a pass-through state when necessary). This will be based 
on the raw voltage inputs which are measured as well.

The PIC can act as a frequency reference of the switchers as well. This 
will concentrate the "noise" at a defined frequency, which can be 
electronically adjusted as well. This might be convenient when "noise" 
is within the frequency band of interest.

The FAN will be controlled from the PIC as well. This allows a more 
intelligent control of the fan. E.g.  a delayed switching off after 
using the radio and a variable speed control.

Another suggestion was to add a display read-out on Demeter as well. 
This will make Demeter and the rest of the HPSDR less dependant on a PC. 
E.g. when used with Sasquatch.

The USB-hub is incorporated because Demeter will consume a USB port. 
Probably ;-) extra USB-ports will be necessary to control the HPSDR and 
some other periphery.

The hub will be arranged in such a way that it has a PC-input which will 
be distributed to four ports. One port will stay internal for Demeter, 
the other three will be available for external use. Standard IC's are 
available for building an USB-hub so implementation will be quite easy.

As an extra a V++ is available for powering some power-amplifiers or 
other circuitry which needs a (slightly) higher voltage which might be a 
bit noisy.

 

For those who are scared by the switchers in the deisgn: By using a 
proper lay-out and filtering on the right places it must be possible to 
make the power-supply noise-free enough for HPSDR use. Maybe whe need a 
Pandora-II.... A Pandora-II will keep out the heat and some noise of the 
HPSDR.

 

>From Henry Vredegoor I/we received an extensive E-mail with some nice 
suggestions which I will be looking into.

73, have a nice weekend, Jeroen PE1RGE

-- 
Ing. Jeroen Bastemeijer

Delft University of Technology
Department of Electrical Engineering
Electronic Instrumentation Laboratory
Mekelweg 4, Room 13.090
2628 CD Delft
The Netherlands

Phone: +31.15.27.86542
Fax: +31.15.27.85755
E-mail: J.Bastemeijer at TUDelft.nl
GPS: Lat N52.00002 Lon E4.37157 Alt 46.2m

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openhpsdr.org/pipermail/hpsdr-openhpsdr.org/attachments/20070622/625bdf5c/attachment-0003.htm>


More information about the Hpsdr mailing list