AC/DC converter problem

Question

EDITED to update diode spec and schematic...

I have two failed outside security lights each with PIR circuit with relay to switch 240V AC mains voltage to another circuit which looks to be a simple AC/DC converter which then feeds a matrix of 60 LEDs wired in series.

Looking at the (very) simple circuit board, the AC/DC converter looks like the image below...

Apologies, the image isn't great. List of components...

L1 4.0 is all it says on the side. Just a radial inductor. R1 510K (1/4W by the look of it) C1 0.68uF metallized film

U1 4x 1N4007 diodes arranged as a bridge rectifier

C2 4.7uF 400V axial electrolytic R2 15R (also possibly 1/4W)

R3 UNKNOWN

OK, so R3 is unknown as it has been fried and the outer resin coating along with colour coded bands has turned to dust (the PCB also has a nice blackened look to it). Interestingly though, the resistor is still mostly intact and my meter reads it as 54K but I don't know if that was its original value or if it has degraded due to partial burn-out.

So, what's my question? I've wired the input directly to the mains and left the LED matrix in place. I measure 239V AC input (close enough) but I'm seeing 292V DC on the output. How is that possible?

The LEDs don't light up but I've since worked out that's because one of the LEDs has blown (all others test OK).

I'm assuming that R3 is under-rated and should be at least 1W if not 2W but even at that rating, what should its value be based on the general design (54K?, something else).

I've dealt with this sort of circuit before but normally with input fed from the secondary of a 12V transformer (for example) and output going to a voltage regulator to create a regulated PSU. I've never dealt with a straight 240V AC-DC converter before.

Throw it away - if it's using 1N4001 diodes these are only rated for a maximum reverse voltage of 50V - it has been designed by an idiot. — Andy aka, Sep 04 '16 at 10:38
@Andyaka : Actually I feel a bit of an idiot - I looked closer at the diodes - they're actually 1N4007 but the markings are slightly faded so looked a bit like 1 instead of 7. The circuit (and quality of manufacture) is however seriously substandard. — MisterSquonk, Sep 04 '16 at 12:36
Klap Lubbish ! there is nothing to protect against inlush current , — Autistic, Sep 04 '16 at 13:47
@MisterSquonk -- see if you can measure the forward voltage of a LED? (If they're white, they could be about 3.5V apiece which'd make 270V DC needed to light the LED string, and means R3 is a high value bleeder resister of say 1M) — ThreePhaseEel, Sep 04 '16 at 16:04

score 1 · Accepted Answer · answered Sep 04 '16 at 12:02

That circuit is a joke. As Andy mentioned in a comment, 1N4001 diodes are inappropriate here. They should be 1N4004 at least. The rest of the circuit is nothing to be proud of either, unless you're the manufacturer and you're raking in the extra few pennies per unit you saved by buying cheap parts and hiring incompetent engineers, and your business model is to dump cheap crap out there, then run away before it catches up with you. Take a look at the unit and see if there is even a manufacturer listed. Even if there is, see if you can find them. Probably not.

To answer your question about voltage: The input voltage is 239 V, but that's a sine wave. When you measure AC voltage like that, you get the RMS value. The peak value of a sine wave is sqrt(2) higher than the RMS. The peaks of your input power is therefore 338 V. You are only seeing 292 V because various parts of this circuit drop some voltage.

Thank you for reminding me about AC theory. I'm actually a qualified electronics craftsman but that was 30+ years ago and I side-stepped into IT support. I'm trying to relearn electronics as a hobbyist. Apologies, for my mistake in the original question - I've looked more closely at the diodes and they're 1N4007. You're right about the quality of design and manufacture - I don't have a manufacturer name but these were supplied by a mainstream UK "Home and Garden" store - part of this exercise is to prove just how bad these units are and send a letter to their head office. — MisterSquonk, Sep 04 '16 at 12:46

AC/DC converter problem

1 Answers1