OpenBCM V1.07b12 (Linux)

Packet Radio Mailbox

DB0FHN

[JN59NK Nuernberg]

 Login: GUEST





  
G8MNY  > TECH     19.12.05 19:43l 162 Lines 7774 Bytes #999 (0) @ WW
BID : 9302_GB7CIP
Read: GUEST DL1LCA OE7FMI
Subj: Switch Mode PS Principles
Path: DB0FHN<DB0NOE<DB0GAP<DB0AAB<F6KFT<DB0GE<DB0ZDF<DB0LJ<DB0RES<ON0BEL<
      ZL2BAU<VK6HGR<GB7YFS<GB7CIP
Sent: 051219/1027Z @:GB7CIP.#32.GBR.EU #:9302 [Caterham] $:9302_GB7CIP
From: G8MNY@GB7CIP.#32.GBR.EU
To  : TECH@WW

By G8MNY                                                (New Dec 05)
There are 2 types, they work slightly differently but both use high frequency
LF into a ferrite cored transformer. With ferrite the eddy current losses are
very small the BH curve magnetic cycle area loss is also much smaller than for
iron.
                        __ 
   IRON      ³   __.--"/            A MODERN  ³     ..-'
   LAMINATE  ³./'  ./'              FERRITE   ³   // 
           ./³   ./'                          ³ //
      ÄÄÄÄ/ÄÄÅÄÄ/ÄÄÄ +H                ÄÄÄÄÄÄ/Å/ÄÄÄÄÄ +H
       ./'   ³/'                           // ³
     ./' _./'³                           //   ³    
 __.:..-"    ³-B                     .-''     ³-B

The high frequency means smaller lighter transformers. In fact for most
ferrites the power handling is proportional to the frequency. So a core running
at 250KHz can handle say 250W, but at 50KHz only 50W.

Transformer 8 ´               _ . -"""-._ eddy current &
   Power    4 ´       _ . - '            `-. BH losses 
            2 ´ . - '                       \ win out
            1 ÁÂÄÄÄÂÄÄÄÂÄÄÄÂÄÄÄÂÄÄÄÂÄÄÄÂÄÄÄÂÄÄÄ 
               5  10  20  40  80  160 320 640 KHz

The problems at these higher frequencies are the other circuit losses, not just
the increased eddy current & BH losses in the ferrite, but inter winding
capacitance, winding skin effect copper losses etc. External losses also
increase, such as transistor on & off times (FETs can be fast if driven well),
& diode rectifier on & off times & of course not to forget the increased
radiation losses from all the wiring as the frequency goes up.

MAINS CIRCUIT
Both types of SMPSU start with a MAINS FILTER (not always fitted despite QRM
laws!) the capacitors must be the X rated type for mains work if earthed there
will also be 4n7 caps to earth. The surge limiting resistor (say 5R 10W or a
thermistor) feeds the bridge rectifier into either 2x 200V electrolytics or 1x
400V one. With the 2 caps a 110v mains option can be provided with just a wire
link, that uses only half the bridge as a voltage doubler. (OK if mains 60Hz &
the 2 caps are large value).
                               BRIDGE
                       SURGE  ÚÄ´>ÃÄÄÄÂÄÄÄÄÄÄÄÂÄÄÄÂÄ +HT         ÄÂÄ +HT
 FUSE    MAINS FILTER  LIMIT  ³       ³       ³   ³200v           ³
LÄÄo-oÄÂÄÄÂÄÄ¿   ÚÄÄÄÂÄÄÄRÄÄÄÄÁÄ´<ÃÄ¿ ³     100K ===     OR WITH  ³
       ³  ³  (((()   ³              ³ ³ Link  ³   ³      NO 110V  ³ 300-380V
 Anti  ³ === =====  ===    ÚÄÄÄÄÄÄÄÄ)Ä)ÄÄo oÄÄÅÄÄÄ´      OPTION  === MAINS
 Shock R  ³  (((()   ³     ³        ³ ³ 110V  ³   ³200v  WITH A   ³ LIVE HT
NÄÄÄÄÄÄÁÄÄÁÄÄÙ   ÀÄÄÄÁÄÄÄÄÄÁÄÄÂÄ´>ÃÄ)ÄÙ     100K ===     400V CAP ³
           BALANCED           ³     ³         ³   ³               ³
            CHOKE             ÀÄ´<ÃÄÁÄÄÄÄÄÄÄÄÄÁÄÄÄÁÄ -HT         ÄÁÄ -HT

SINGLE ENDED
These use the back emf or "ringing up" of the stored energy in the ferrite
cored transformer. The transformer then does 3 jobs, provides isolation,
provides step down ratio, & provide choke input filter. 
 
                             FAST      Pi O/P Filter
+HT ÄÄÄÄÄÄÄÄÂÄÄÄÄÄÄÄÄÄÄÄÄ¿ RECTIFIER    ====
FERRITE     )|ÚÄÄÄÄÄÄÄÄÄÄ)ÄÄ´>ÃÄÄÄÄÄÄÄÂÄ(((()ÄÂÄÄÄÂÄÄÄÄÄÄ +
TRANSFORMER )|(          R           ===     ===  ³       LOAD
            ³ ÀÄÄÄÄÄÄÄÄÄÄ)ÄÄÄÄÄÄÄÄÄÄÄÄÁÄÄÄÄÄÄÄÁÄÄÄ)ÄÄÄÄÂÄ -
         ___³            ÃÄÄÄÄ¿                   ³    ³
        ³    \³    ÚÄÄÄÄÄÁÄ¿  ³+                  ³    ³
        R     ÃÄÄÄÄ´Control³ ===                  ³    ³
SNUBBER ³   e/³    ³  IC   ³  ³                   ³    ³
        ³   ÃÄÄÄÄÄÄ´       ³  ³   ________       set   ³
       ===  ³    C ´       ÃÄÄ)ÄÄ´  opto  ÃÄÄÄÄÄ>volts ³
        ³   R    R ÀÄÄÄÂÄÄÄÙ  ³  ³isolator³      pot   ³
-HT ÄÄÄÄÁÄÄÄÁÄÄÄÄÁÄÄÄÄÄÁÄÄÄÄÄÄÁÄÄ´________ÃÄÄÄÄÄÄÄÁÄÄÄÄÙ
                OSC

The phase of the transformer is arranged so that when the transistor is turned
off the rectifier on the secondary conducts. The pulse width control IC is
powered from a HT dropper R, sometimes there is a separate PSU or secondary
that takes over for this.

The OSC CR sets the switching frequency, the IC also monitors the transistor
current with a small value emitter R. Voltage feedback from the output drives
an opto isolator to control the pulse width of the IC to alter the energy
stored in the transformer for the next pulse. As only DC pulses are fed into
the transformer only half the core BH curve energy can be.

Secondary³    Rect__on
voltage  ³       ³  ³ ringing
         ³       ³  ³º 
      0v ´Ä¿pulse³  ³³ÀÄ¿               Height due 
         ³ ³width³  º³  ³       ³      ³ to mains V &
         ³ ³_____³  º   ³_______³     _³_turns ratio
         ³TRANSISTOR     pulse width
         ³   ON 

Due to the pulse current in the output is it usual to have a pi filter for the
DC output.
   
PUSH PULL
These are more complex & more powerful. The double action utilises the full
power handling of the transformer core, but that means the secondary must be
full wave & an additional HF choke (about the same size of the transformer) is
now needed is needed to convert the HT fixed height variable width voltage
pulse to a variable current feed to the secondary load.

The control IC now has 2 drive outputs with not more than 45% on, as each power
switching transistor has to have some to turn off before the other antiphase
one is turned on.

+HT ÄÄÄÄÄÄÄÄÂÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄÄ¿    PULSE     optional
        ÚÄÄÄ)ÄÄÄÄ¿    diodes   ³    CHOKE     pi filter
        ³   ³     )|| ÚÄ´>ÃÄÂÄÄ)ÄÄÄ¿ =====     ====
        ³   ³_____)||(      ³  R    ((((()ÄÄÂÄ((((()ÂÄÄÄÂÄÄÄÄÄÄ +
        ³         )|| ÃÄÄÄÄÄ)ÄÄ)ÄÄ¿         ³       ³   ³      DC
        ³         )||(      ³  ³  ³        ===     ===  ³    OUTPUT
        ³        ³    ÀÄ´>ÃÄÙ  ³  ÀÄÄÄÄÄÄÄÄÄÁÄÄÄÄÄÄÄÁÄÄÄ)ÄÄÄÄÂÄ -
        ³     ___³             ÃÄÄÄÄ¿                   ³    ³
        ³    ³    \³     ÚÄÄÄÄÄÁÄ¿  ³+                  ³    ³
      __³    R   T2ÃÄÄÄÄÄ´Control³ ===                  ³    ³
     ³   \³  ³   e/³     ³  IC   ³  ³                   ³    ³
     R  T1ÃÄÄ)ÄÄÄ)ÄÄÄÄÄÄÄ´       ³  ³                   ³    ³
     ³  e/³  ³   ³       ³       ³  ³                   ³    ³
     ³  ÀÄÄÄÄ)ÄÄÄÅÄÄÄÄÄÄÄ´       ³  ³   ________       set   ³
    ===     ===  ³     C ´       ÃÄÄ)ÄÄ´  opto  ÃÄÄÄÄÄ>volts ³
     ³       ³   R     R ÀÄÄÄÂÄÄÄÙ  ³  ³isolator³      pot   ³
-HT ÄÁÄÄÄÄÄÄÄÁÄÄÄÁÄÄÄÄÄÁÄÄÄÄÄÁÄÄÄÄÄÄÁÄÄ´________ÃÄÄÄÄÄÄÄÁÄÄÄÄÙ
                      OSC

                           ______
          ³               ³      ³
Secondary ³     ringing   ³T2 on ³
voltage   ³       º³      ³      ³
       0v ´¿pulseººÀÄÄÄÄÄÄÙ      ººÚÄ
          ³³width³               º³
          ³³     ³               ringing
          ³³_____³
          ³ T1 on

 Voltage ³ ______          ______        _
  from   ³³      ³        ³      ³       ³ height due 
         ³³ _ _ _³ _ _ _ _³ _ _ _³ mean  ³ to mains V &
  fast   ³³      ³        ³      ³ DC    ³ turns ratio
 diodes  ÁÁÄÄÄÄÄÄÁÄÄÄÄÄÄÄÄÁÄÄÄÄÄÄÁÄÄ     Á 
          on to off ratio depends
          on load & supply voltage

As the pulse choke can be made big enough to still give current between primary
pulses sometimes a pi DC filter is not used.

SAFETY
1/ Working on them it is recommended you use a mains isolation transformer.
2/ Use a high voltage 1/10 or 1/100 scope probe, rated at 600V @ 100KHz not one
   rated at 400v @ 100Hz MAX!
3/ If in doubt if HT is discharged when off use a discharge R 10K 5W.
4/ Use the 1 hand behind the back safety precautions on HT circuits!

FUTHER READING
See buls on "Stopping HF Tx/Rx SMPS QRM",  "Reducing Electronic RF QRM" & the
crowdbar part of "High AMP crowbar protected PSU".

Why don't U send an interesting bul?

73 de John G8MNY @ GB7CIP


Read previous mail | Read next mail


 04.04.2025 02:23:44lGo back Go up