I came across an old Apple mini USB adapter CLONE. It even had the green dot from the 2008 recall, so it's about a decade old. I'm actually glad I took it apart because one of the prongs fell off very easily and based on what I've seen, it's so simple compared to Apple's original that I doubt it was very efficient or gave a very clean output. On the bright side, its simplicity makes it a good circuit for learning.

What I'm curious about is the setup of the two transistors used to control the output voltage as I've never seen this kind of arrangement before. After desoldering all the components so that I could reference the PCBs, I've been able to simulate the circuit in LTSpice. Although the steady-state output overshoots to about 5.7V, the simulation is good enough to advance this discussion.

You can download the LTSpice circuit. All custom models are embedded so it'll hopefully run on your end without issues.

Adapter Circuit


Transformer, Switching

Fig. 1 - Transformer Fig. 2 - L1 Switching

The only markings on the (flyback?) transformer were "W" and "28" so the best I was able to do was to empirically determine the orientation and approximate inductance of each of the 3 coils. From what I could tell, it seems like L1 is the primary, L2 is an auxiliary and L3 is the secondary. It seems that L2 provides isolated power to the feedback circuit.

A quick test with a 170V, 33kHz pulse with 30% duty cycle gave a 5V output (Fig. 1). Update: It turns out that this can also be achieved with a much lower frequency of around 1.2KHz and approximately the same duty cycle. This is more in line with what I'm seeing in the simulation.

Diode D2 blocks half the AC signal and along with capacitor C2 provides a fairly constant voltage of 170V DC to the top pin of inductor L1 (red line in Fig. 2). The switching takes place via the bottom pin with a frequency of ca. 1.2KHz when the output voltage stabilizes, shown in grey in Fig. 2. If you look closely, you can see that the frequency decreases around the 140ms mark when the voltages reaches its target goal.

Zener Diode and Feedback Voltage

Fig. 3 - Zener Voltage Fig. 4 - L2 Feedback Voltage

I modelled a simple Zener diode with a reverse breakdown voltage of 4.3 but in the simulation it maxes out at 4.7V as shown in Fig. 3. This is in fact what I'm seeing with the actual diode, maxing out at around 4.6V.

Also, the voltage drop across the diode in the optocoupler maxes out at 1V. The 4.7V from the Zener plus the 1V from the optocoupler explains the resulting 5.7V, I guess. It should also be noted, however, that dropping the load resistance to 400Ω results in a voltage output much closer to 5V.

Musings and Questions

  • The adapter has been completely dismantled and won't ever be back in operation and I am left wondering if the output voltage was ever a stable 5V or if it varied depending on the device it was connected to. Thoughts?

  • I'm observing that the optocoupler is proving feedback to the transistors and adjusting the frequency of the controlling pulse but having trouble wrapping my head around the details of this functionality. Fig. 4 shows the voltage at the top pin of the aux L2 and the voltage drop across it. Any details on this would be appreciated.

  • The observed voltage at pin #1 (optocoupler emitter) swings from 0.7V to -7V. Why does it make sense to connect pin 1 directly to the base of Q1? Is the 0.7V switching Q1 on or are we seeing the 0.7V because Q1 was switched on?

  • It looks like Q1 is a low-voltage switch for Q2 which in turn is a high-voltage switch for L1 but then why is it connected back to Q1's base via R1?

  • The closest configuration I've found to the dual transistor configuration is from this post, entitled "2 NPN configuration that apparently works by small Recombination Currents" but it's not quite the same. It's not a push-pull; it's not a darlington; what is it and is there a name for this configuration?



Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.