4

I'm repairing the circuit for a motion-activated LED light. The PCB uses a capacitive dropper circuit as the power supply. Photos of the PCB and the corresponding reverse-engineered power schematic are shown below.

PCB

Schematic

In this case, the two glass diodes in the bridge rectifier have failed short (D4, D5). Before I replace them, I'm wondering what's the significance of using two different diodes for the bridge rectifier? I would have thought that for such a simple circuit, any efficiencies saved by a potentially lower diode forward voltage drop wouldn't be worth it.

Edit Here's a closeup of the glass diode markings. The band appears to be green, which, according to some of the links posted in the comments, would imply a Schottky diode. Update after finding some more information online regarding this specific circuit, I can confirm that the glass diodes are indeed zeners.

MELF Diodes

Update I'm happy to report that after replacing the two glass diodes with 24V 500mW MINIMELF diodes, as well as transistor T2 (BC81-40), the light now works fine.

Amr Bekhit
  • 1,278
  • 10
  • 24
  • This system is galvanically connected to the AC mains, it seems. Assuming an intelligent designer and not some arbitrary reason, I'd have to offer the suggestion that there may be a difference in the way these diodes may fail that led the designer towards these differences. But now I feel ignorant and also curious, too. Perhaps someone here will be able to provide the missing piece and we'll both learn something new! Good question. +1. – jonk Aug 22 '20 at 19:46
  • Only the neutral is galvanically connected, the high side is dielectrically coupled. – Tony Stewart EE75 Aug 22 '20 at 19:58
  • I'm thinking that MAYBE the two larger diodes are actually TVS diodes that would clamp any surges, yet act as normal diodes for the normal voltages. – John D Aug 22 '20 at 20:13
  • 2
    The [diodes marked M7](https://www.vtrons.com/images/DIODE%2520M7.pdf) are plain rectifier diodes. – JRE Aug 22 '20 at 20:20
  • 1
    The MELF diodes have a blue band. [That says they are Zener diodes.](http://www.marsport.org.uk/smd/sod80.htm) – JRE Aug 22 '20 at 20:37
  • 1
    @Tony, or only the live is galvanically connected ... – Transistor Aug 22 '20 at 21:01
  • @Transistor good point, can't remember how many outlets I've seen wired backwards (here in the US anyway) so even if the plug is polarized it guarantees nothing. I'm not a fan of these non-isolated mains supplies. – John D Aug 22 '20 at 21:03
  • @JRE Right! So the MELF diodes are for clamping transients and even regulation, and the others are regular rectifiers. Apparently the zeners encountered a transient overvoltage with more energy than they were capable of absorbing. – John D Aug 22 '20 at 21:09
  • @Transistor the low impedance DC ought to be polarized to the large tab Neutral and the high impedance to Line – Tony Stewart EE75 Aug 22 '20 at 22:34
  • @JRE From what I can tell, the band is actually green, indicating Schottky diode. – Amr Bekhit Aug 23 '20 at 10:03

3 Answers3

2

The two glass diodes are zener diodes.

They serve two purposes

  • as one pair of a bridge rectifier

  • as a 24V bridge input voltage clamp.

A couple of 1N4007s (the marking 'M7' is clearly visible on one of them) make the other pair.

enter image description here

vu2nan
  • 15,929
  • 1
  • 14
  • 42
  • In your circuit, you've placed the glass diodes at the bottom, connected to 0V, but on the PCB, the glass diodes are actually the top half of the rectifier, connected to the 24V out. – Amr Bekhit Aug 23 '20 at 10:04
  • 1
    That would make no difference to the functioning, which would be the same in either case. – vu2nan Aug 23 '20 at 10:11
  • Ah yes, you're right. In any case, I think the assumption that the glass diodes are zeners may not be correct - the band is actually green, which may indicate a schottky diode according to the links posted by some other users. – Amr Bekhit Aug 23 '20 at 12:01
  • 1
    But, I suppose, the probability of those being zeners is more. – vu2nan Aug 23 '20 at 18:11
  • I managed to find some more information online and it confirms that the glass diodes are indeed zeners, as you correctly identified. – Amr Bekhit Aug 24 '20 at 09:34
  • 1
    Thank you, Amr Bekhit. – vu2nan Aug 24 '20 at 11:04
0

The larger diodes in the schematic are 24V Zener Diodes. They clamp the incoming line voltage to 24 V pulses.

The small diodes in the schematic take those pulses and feed them to the bulk reservoir capacitor..

This is similar to a standard capacitive dropper except that it is full-wave.

Dwayne Reid
  • 23,390
  • 2
  • 35
  • 67
-2

Assuming they used the cheapest diode 1N4148 300mA < $0.01 in vol. There is no doubt a line transient would blow the diode junction. The resistor has a larger junction but might also blow open from 9A

The 1N4148 is rated for impulse current rate of 4A single shot 1us. Abs Max. A Resistor might have a 0.5 s time constant

A cheap 1N4007 1kV diode can handle 30A 1.2kV half-sine 1 shot but then the SMD resistors will blow. Install it with bent leads carefully with 2 tweezers folded to fit on SMD pads or any other 1A SMD silicon diode.

Check all components in the front end for component values open or short by some sort of test. [![enter image description here][1]][1] https://store.comet.bg/download-file.php?id=13687 The large M7's are similar to 1N4007 1kV silicon diodes rated for 1A, or 30A impulse

All 4 diodes should have been M7's and 100R . A line filter would reduce the impulse voltage but stretch out the duration. [1]: https://i.stack.imgur.com/WWweG.png

Tony Stewart EE75
  • 1
  • 3
  • 54
  • 182