Easy Repair of a Faulty Power Supply;


or, Just a Lucky Break?  8^)


01 powersupply small


23rd October 2015. A friend recently acquired a quantity of equipment related to his own field of interest. However, the lot included a number of items of other kinds, principally semi-vintage electronics & amateur radio. He very kindly unloaded these onto me, saying that he didn’t know much about that sort of thing, but they just might be of use to me, perhaps to break up into their components & use them in odd projects &c. One of them you see above. It’s a chunky 30 Ampčre variable 13.8V power supply by Zurich, which, oddly, is a Chinese company. Nevertheless, Zurich have made many different power supply units for a long time, and indeed still do. My friend expressed some diffidence in cluttering up my living room with this stuff, but I reassured him that each item would be rigorously examined, and that on the one hand, any which did not work (or were unduly silly) would be straightway discarded; but that on the other, anything that worked (and was of use to me) would find a good home here, and some cash recompense forwarded to him in respect of each such item.


Now as you know, one can hardly have enough nominally 12 Volt bench power supplies, right? Very often, one has to drive more than one piece of gear with 12V from a separate source. In fact, we have two already. The first is fixed 13.8V at 10 Amps. This is an utterly basic requirement. The second is variable from zero to 30 Volts at up to 10 Amps. The two together would generally be considered, shall we say, adequate. However, the possession of a third variable voltage supply, especially one rated at 30 Amps, might best be described as sheer swinish self-indulgence. Oh, we tried to resist the temptation for ever such a long time - several hours at least! But we were hooked. As soon as my Benefactor had departed, we fired up the Zurich DPS-2512M which sprang into life, showing 13.8 Volts on its meter, which was confirmed by our test meter. We danced around the room, crowing and gibbering with delight.


At last comes the first serious bit of this page. Had we really tested the unit? NO! We had merely established the fact that a variable voltage was present at the output terminals. A full test would consist of feeding the output into a dummy load at full power for an hour or so, and seeing if there was anything left of the unit at the end of that time. If it was still in one (rather hot) piece at the end of an hour, we might legitimately conclude that it was working OK.


The dénoument was rather different.


We decided to test it by firing up our long-neglected 144 MHz (2m) FM amateur rig – which it seemed to do – and call CQ on 145.50 MHz asking for a rig check. Even taking into account the relatively low activity on 2 metres these days, we were eventually puzzled by the lack of a reply. Perhaps our rig, unused for nearly 18 months, had stopped working? So we changed back to the old PSU to check the rig. We had an immediate positive response. The rig was working, and our signal was perfectly readable. We then changed to the Zurich PSU, and that amateur could not hear us at all; but another, much closer, reported that there was only a loud hum to be heard. Aiee! There must be something wrong with the Zurich. Oh, dear.


Recourse was made to Google, and we discovered, at the very useful site http://g3zjo-radio.blog.co.uk/2009/03/03/zurich-dps-2512m-30-amp-power-supply-5686675/ that there is what one might call a generic problem in the DPS-2512M.


02 control board small


The unit is massively constructed, and the transformer, bridge rectifier, regulator transistors & heat sink are most unlikely to fail in use. However, the variable voltage control board – above – has its own, on-board power supply. As the current required is very small, it consists merely of a single 1N4002 diode, and a 470µF 10V electrolytic capacitor. In one example on the G3ZJO page above, the diode was found to be a dead short, so the 470µF capacitor was heating up – always a fatal sign with an electrolytic. The replacement of the diode and capacitor cleared the problem.


So we dived into our PSU, and Lo!



A blow-hole in the top of the 470µF capacitor indicated that it had given up the ghost, either because the diode had gone dead short, or perhaps the capacitor had simply grown weary of its existence? Actually, we suspect the diode in the example cited above had gone down as a result of the death of the capacitor, for the current rating of the 1N4002 is 1 Amp, and the consumption of the control board would surely be only a few milliamps? In any case, we renewed both components. We only had 1N4007s to hand, but they are no larger and even higher rated. Also, we only had 470µF electrolytics of 25V working rather that 10V; but in this application, that wouldn’t have any significant effect.



The board was screwed back in place, and the PSU worked fine. A single call on 145.50 MHz for a rig check was immediately answered, and gave our signal a clean bill of health.


The only downside, is that I will now have to bung my friend a few quid for the unit! 8^)


Lastly, this failure of a small electrolytic is identical to another generic fault described elsewhere in this rambling site, concerning Indesit washing machines:






Page written 24th October 2015.