Switch-Mode Mains Power Supplies

Follow-up to G3SEK's articles in RadCom, November and December 2004.
Click the links to download PDF files.

WARNING - these power supplies are DANGEROUS!

Use information on this page at your own risk.



In Fig 7 (December), R8 should be 2K2 and RV1 should be 500R. With 1.0mA flowing, these values allow the voltage at TL494 pin 1 to be adjusted to exactly 2.5V.

Modifications to 13.8V Output 

(See circuit diagrams in the original articles for part-number references.) 

These modifications come in three basic levels:

  1. Simply accept the existing +12V DC output, or tweak RV1 to increase it. To reach 13.8V, you sometimes need to change R8 to the next-lower standard value. This will give up to 10A (the existing rating of the +12V output) with a voltage drop of less than 1V on full load.
    Note: you will always need a load resistor on the 5V output - see the articles.
  2. Go for higher current capability and better voltage regulation. This generally involves:
    1. Swapping D3 and D4 so that the 13.8V output uses the larger, lower-resistance diode D3.
    2. Modifying filter choke L2 (which can be a problem because there are no other windings to reduce the core flux)
    3. Modifying the fault detection circuits so that they monitor the new 13.8V rail instead of +5V. 
    The weak point then is the secondary winding of T1, so the next level is...
  3. Rewinding the secondary of T1, which requires you to take full responsibility for all the output and control circuits.

All of this can be done - the article by VK6APH is an excellent example of a 'Level 3' mod - but letís also keep some perspective:

  • A typical 100W-class transceiver requires 13.8V at 20A peak (260W), which is beyond the capabilities of any PSU rated for a total output of only 200W. Even if a 300W PSU can be modified to deliver 13.8V, it will be marginal - and may be stressed towards the limits of reliability.
  • Are you seriously proposing to operate maybe £1000-worth of transceiver from a PSU that wasnít designed for the purpose, that you have modified, and whose circuit you donít quite understand? Does that seem like a sensible risk? It doesnít to me. 

The photograph in Part 2 shows a ĎLevel 1í modification of the PSU pictured in Part 1:

  • A mains rocker switch and Powerpole DC connectors are epoxied into existing slots in the case
  • A better mains input filter is bolted directly to the case. This seems to be sufficient to kill all the output noise - nothing is audible on an HF transceiver powered from this particular unit.
  • Further options to reduce output noise would be to wind both output leads on a  ferrite toroid; ground the negative lead to the case immediately behind the output socket; and bypass the positive lead to the case, also immediately behind the output socket. I have not tried any of these options. 
  • The metal-bodied 10W load resistor for the +5V rail is bolted to the case, underneath the mains filter. The case provides the heat-sink that these resistors need.


Note the widespread use of heat-shrink insulating sleeving in the photograph.

This PSU is reasonably safe when its cover is replaced, but don't forget that it was originally designed for use inside a PC case. Without this additional protection, it does not meet modern safety standards because it's possible to contact mains and other high voltages through the ventilation slots... but the PSU will overheat if those slots are blocked. TAKE CARE!


Back to   G3SEK's Amateur Radio Technical Notebook

'In Practice' pages

Updated 01 November 2008
Page © 2004 IFWtech.