Thermal Compound Short a SMPS?

[DreadPirate]

This is a meanwell S-50-24. I got it recently, tested fine. I installed it in a plastic case and may have caused the thermal compound on the rear switcher (Fairchield 5L0380R) to squeeze onto its leads while I was manipulating it in a tight space. Now, I get 52VAC on the output (125 VAC input). Outer case removed. Fuse did not blow. LED lighst up. I plan on replacing the diodes, perhaps both chips, though these seem hard to find singly. Anyway, could the compound have been the cause of failure? I'm pretty sure neither the inputs or outputs shorted, nothing else looks awry.

http://us.100y.com.tw/pdf_file/S-50.PDF

[star882]

You were probably measuring capacitive coupling. Modern multimeters are very high impedance.

[wwenze]

Thermal compound are mostly non-conductive, most commonly used version being the white one based on silicone medium containing ceramic powder. Metal-based ones exists but tend to not look white.

[DreadPirate]

I'm frequently amazed at my own stupidity. I had the dc voltage setting at 20 (instead of 200, next higher on the meter), so I got no reading (its a 24 v unit). I then switched to AC and measured 52 v. This prompted me to take the thing apart. For some reason, meters are thrown off here, I had a similar experience with a 12v meanwell unit. Anyway, all is well, unit works perfect, I'm getting 24v dc reading, but I'm still a bit confused as to why the ac would read as 52. These things are certainly cheaply built, I think I got it for $14 shipped, but they do have a high voltage rubycon ce cap.

[pjp]

The cheaper multimeters all do that. 24VDC will read ~50V on the AC range.

[DreadPirate]

True. My Fluke 77 doesn't do this.

[BADFLOPPY]

The surely feedback loop fails.

During design, the output kept a little bit higher voltage than required than lowest input voltage. The oscillation gets off when final voltage achieved, (Sensed by 431 IC, a small 3 pin transistor like) using resistor biasing and optocoupler.

So, when you give full voltage it generate 52 voltage output.

If feedback fails, PWM Control (?) or on-off control gots damaged and kept on all the time. So full voltage.

So,
Check step by step.
1. Biasing Resister of TL431 or 431 IC. (During operation, sense pin voltage 2.5 Volt)
2. Replace 431IC .
3. connect LED in opto-coupler input terminal to check 431 or series resistance.
4. Replace opto-coupler (if optocoupler fails, shorting high voltage side, oscilation stops and output voltage falls to zero.
5 If shorting opto-isolator, gives no result, replace the FPS.

Hope i explained simply.

Thanks