1
\$\begingroup\$

I was reverse engineering a battery charger (Lidl brand Parkside 20V battery charger PLGS 2012 A1, with smart functions, among them charge current and termination voltage settings) and came across this unexplainable-to-me flyback transformer setup. -I stand corrected, as this is a forward converter-

As you can see in my below schematic the main SMPS transformer has 4 windings: a switched primary (between pins 1 and 2), an output secondary(which does the actual battery charging), a grounded shield winding that is only connected on one end (pin 4), and my "rogue" winding (between pins 1 and 3). You can see that the latter shares the same starting point at the rectified mains rail as the primary winding, but on the other end only connects to the cathode of a diode (D2), whose anode connects to ground. Measuring with a multimeter there is no connectivity between pin 3 and 4 (at first I thought this was simply an AUX winding but it isn't).

The switching is done via NPN transistor, driven by a CP1252, which has secondary side feedback via an optocoupler so doesn't need an AUX winding, and derives its VCC from a small second SMPS (driven by a KP2130), which provides around 10V to the primary side ICs via its AUX winding, and 5V to power ICs on the secondary side via its secondary winding.

In the schematic I have marked in red my two oscilloscope probe points, 1 and 2 (with the scope grounded) for the two waveforms you can see below.

My "rogue" winding practically has the inverse waveform of the primary. I imagine this has something to do with efficiency/PFC, but I have never seen such a setup and do not understand what function this "rogue" winding is preforming. It seems to me that D2 never conducts, it certainly isn't snubbing the "rogue" winding?

I'd really appreciate some insight into this unique setup. Thanks for the replies!

Datasheets

CP1252

KP2130

enter image description here

Here is the PCB: enter image description here enter image description here

And the oscilloscope waveforms:

Probe point 1 (bottom of PRI winding, Pin 2): enter image description here

Probe point 2 (far end of the "rogue" winding, Pin 3): enter image description here

\$\endgroup\$
9
  • \$\begingroup\$ Are you sure that's not a feedback winding of some sort? \$\endgroup\$
    – Hearth
    Jan 2 at 20:19
  • 1
    \$\begingroup\$ Check you you have two diodes at the secondary side as is shown here for a forward converter with demagnetizing winding en.wikipedia.org/wiki/Forward_converter But your scops show that this it's a Forward converter \$\endgroup\$
    – G36
    Jan 2 at 20:28
  • 1
    \$\begingroup\$ why do you say that it is a flyback? \$\endgroup\$
    – Jasen
    Jan 2 at 21:12
  • 1
    \$\begingroup\$ Just an observation on the board layout. I see little evidence of ant attempt to separate the mains input from the low voltage side of the system. There is a small anti-tracking slot (I presume) in the lower RH corner and another at the bottom left, but nothing visible under the feedback opto couplers or around the +325V rectified mains input . It's probably safe, but I'd like to see better. \$\endgroup\$ Jan 2 at 21:42
  • 1
    \$\begingroup\$ In trace 2 the measurements don't add up. It has the Vmax as 661V and Vmin as -122V. Vpp should be 783V but instead it shows 539V. Also pin 3 should never be able to go negative by more than 1-2V because diode D2 would prevent that. Something's amiss. \$\endgroup\$ Jan 2 at 23:03

2 Answers 2

3
\$\begingroup\$

That winding is there to reset the magnetic flux in the core when the transistor turns off.

That arrangement is more commonly used with single switch forward converters rather than flyback. In flyback version it would suppress any excessive flyback voltage when the battery is not connected.

\$\endgroup\$
2
\$\begingroup\$

Actually this is only a stretched repeat of already given comments.

Your rogue winding protects your switching transistor. The primary winding has limited inductance and it has enough time to collect some amount useless DC when the transistor conducts and the secondary outputs the useful pulse to the output.

When the transistor is turned OFF the useless DC current in the primary doesn't stop immediately. As the induction law says the primary current bulldozes its way through all insulations by developing just as high voltage as needed to let the current continue on and decay gradually. That's commonly called "inductive kickback".

The induction actually happens in the magnetic field and generates voltage to every winding. It's enough to suppress the kickback in one winding. Your rogue winding leads (=recycles) the inductive kickback pulse back to the capacitors of the input rectifier (not drawn, but surely exist) and thus prevents the high voltage pulse which would otherwise burn your transistor or other parts (= bulldoze through the weakest insulation).

As already said by others, your circuit is a "forward converter". In flyback converters the inductive kickback would be the one which would be used as output in the secondary, but here it's only a harmful side product caused by non-infinite inductance.

There's also a snubber (=diode +resistor) in the primary. It's needed because the magnetic field coupling is not full 100% between the windings.

\$\endgroup\$
2
  • \$\begingroup\$ Thanks for the explanation. At what point does D2 conduct, as I don't see a moment when the voltage at ground is lower than the other end of the winding (which to me never seems to dip below 0V, no?)? That seems to be the only moment when D2 could conduct? \$\endgroup\$
    – parkside
    Jan 2 at 21:00
  • 1
    \$\begingroup\$ D2 conducts a short moment just after the transistor is turned OFF. Measuring it is difficult because there's also the induced pulse when the transistor conducts and the input DC. \$\endgroup\$
    – user287001
    Jan 2 at 21:13

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.