Power brick squeezes, low output voltages

[Behemot]

I got this Dedicated Micros VS-PSU65W-EXT power brick. it is supposed to have three output voltages (+5, +12 and -12 V). When I got it, it squeezed and had almost no outputs (maybe some tenth of volt).

It seems there is no real ground rather than it has flying ground, that possible? The +5 V output rectifiers (FMB24L) showed aprox. 40 on diode tester. I took them out, it still shows 40. But I think they were partially opened anyway, showed 96-99 on the internal diodes, good shottkies show around 140.

I have swapped them with MBR20100CT, now it still squeezes but I got something, aprox. +1,8 V, +4,4 V and -4,4 V. All other semiconductors (diodes, rectifiers, one zener) measure fine.

I took out one ceramic cap, thought it may be open, it checks fine, however, now the +5V measures 43 or so against common. There is one other cap, small blue, it may be open maybe? Other than that, I do not know what is wrong. Caps are good, Chemi-Con, all measured good (in circuit). Is it possible the secondary transformer winding is shorted? Or may optocouplers be bad?

[shovenose]

Re: Power brick squeezes, low output voltages

Squeezed? Am I missing something here?

squeezed past participle, past tense of squeeze (Verb)
Verb
Firmly press (something soft or yielding), typically with one's fingers: "Kate squeezed his hand affectionately".
Extract (liquid or a soft substance) from something by compressing or twisting it firmly: "squeeze out as much juice as you can".

[rogfanther]

Re: Power brick squeezes, low output voltages

The resistors and those diodes near D2 ? Also, which is the controller ?Could it be the startup cap is failing ?

[Behemot]

Re: Power brick squeezes, low output voltages

…

Oh man. Gues the was too much in this heat Ment whistles

rogfanther: what about them? Not sure if it even has a startup cap, it uses Fairchild KA3882 PWM controller. There is only small cap to filter it's input and that is most likely NCC KMG.

[budm]

Re: Power brick squeezes, low output voltages

Since all the output voltages are only 36% of what they should be, I also then suspect start up cap, something like TL431 ref diode, opto in the feedback.

[Behemot]

Re: Power brick squeezes, low output voltages

Seems that you can't even believe te Chemi-Con these days. KME 47 uF/25 V, measured 0,1 uF/25 ohm in circuit, out of circuit the ESR mciro does no even read *anything*.

Works liek charm, thanks guys, I would not even think about Chemi-Con getting bad…

[rogfanther]

Re: Power brick squeezes, low output voltages

I probably would not suspect Chemi-Con either, but in similar psus from D-Link network switches, its almost always that small cap that fails.

[shovenose]

Re: Power brick squeezes, low output voltages

You meant squeal? Oh... LOL

[Behemot]

Re: Power brick squeezes, low output voltages

What?

[tom66]

Re: Power brick squeezes, low output voltages

All electrolytics suck...
but some suck less than others.

[ben7]

Re: Power brick squeezes, low output voltages

Seems that you can't even believe te Chemi-Con these days. KME 47 uF/25 V, measured 0,1 uF/25 ohm in circuit, out of circuit the ESR mciro does no even read *anything*.

Works liek charm, thanks guys, I would not even think about Chemi-Con getting bad…

KME was a bad choice for this power supply. General purpose, 1000 hour series. Yeah, I am not blaming NCC here! There should have been a low esr 5000+ hour rated cap there.

[Behemot]

Re: Power brick squeezes, low output voltages

Quite often GP caps are used, I guess you just can't use low-ESR everywhere, right?

But the 1000 hours is MINIMUM what it should handle and I took NCC (besides KZG/TMV and so) as a quality manufacturer who makes *only* stuff which survives lot more than it should. And it was not being cooked here or something…there is nothing near what should bake it that much. You see, the whole brick was closed in aluminium cover connected with the heatsinks inside and all this was packed in plastic than. So the heatsinkes trasffered the heat onto the cover, they could not reach extreme temperatures.

But when we opened that…is there anything like the Nichicon VR series but for 105 °C? It has ESR somewhere in middle between GP and low-ESR. I think it is not being made anymore. The current VZ line is GP and PW is general purpose.

ADD// OK, found Nichicon SV. Does not have all the values, but it's a start
ADD2// mhmm, I DO LIKE the NCC FL and LE series, also Rubycon ML,

[ben7]

Re: Power brick squeezes, low output voltages

But the 1000 hours is MINIMUM what it should handle and I took NCC (besides KZG/TMV and so) as a quality manufacturer who makes *only* stuff which survives lot more than it should. And it was not being cooked here or something…there is nothing near what should bake it that much. You see, the whole brick was closed in aluminium cover connected with the heatsinks inside and all this was packed in plastic than. So the heatsinkes trasffered the heat onto the cover, they could not reach extreme temperatures.

Wrong! Those bricks get hot inside! The metal heatsinks do a poor job at heat dissipation with NO cool air flowing around them, and with the plastic around them. The plastic is a poor conductor of heat. So many people get fooled, thinking plastic is better because it seems to stay cooler. The reason is because it is a poor thermal conductor! You can't possibly use a sheet of plastic as a heatsink. If you don't believe me, try it! :P

The capacitor doesn't have to be next to anything hot to get cooked - heat is transferred through the pcb substrate to the capacitor. Also heat is transferred to the capacitor through the air inside the PSU.

All capacitors can last a long time, it just depends how you use them. Even fujyyhu caps can last a long time when kept cool and at a low ripple current.

Also, all capacitors can last a short time. Some are engineered badly. At a high temperature, and high ripple current, even a well engineered capacitor won't last as long.

[Behemot]

Re: Power brick squeezes, low output voltages

I know that but usually the bricks don't have such big aluminium surface inside which could transfer heat to the whole plastic cover. So they get hot only at some places. Also there is almost no discoloration but around transformer primary side.

[tom66]

Re: Power brick squeezes, low output voltages

The ambient air inside the brick will easily heat the electrolytics to well over 90°C. The surface of your charger is only 60°C you say? That's because the plastic is a good insulator...

[Behemot]

Re: Power brick squeezes, low output voltages

One thing I haven't mentioned, there are perforations in the plastic cover so air may flow at least a bit around the outter aluminium cover. Not to mention it is twice the size of todays 65W bricks. Simply I don't believe just this one could go so hot.

[tom66]

Re: Power brick squeezes, low output voltages

The perforations let you touch the aluminium? If not, they're not going to have much, if any real effect -- plastic is -really- good at insulating heat. If they are, I hope the semiconductors have a good mica pad because otherwise that's a shock hazard.

Only way to be sure is to stick a thermocouple inside the unit while it is running.

[Behemot]

Re: Power brick squeezes, low output voltages

They don't let me touch it ofc (even though it is grounded) but air may still get through.

I don't have the device it is powering so I cannot load it.

[tom66]

Re: Power brick squeezes, low output voltages

Air cannot permeate plastic. Plastic is a solid. Plastic insulates well and air flowing over it will have minimal effect on cooling the internal circuitry. It will have -some- effect, but it will be minimal. Primarily the patterning is likely to make the power supply easier to grip (friction on hand) or to make it look a little nicer, it costs nearly nothing to do it.

[Behemot]

Re: Power brick squeezes, low output voltages

I said "perforations", doesn't it mean "holes through the material"?