Switchmode PS caps

[cyclones]

Hello,

I have been around this forum on and off for a number of years. In the last year I have really been working on my troubleshooting skills for electronics. I have repaired a car battery charger (bad diodes), a linear power supply (bad diodes), a computer power supply (bad input capacitor and a blown SMPS control chip that literally blew up!). All have worked great after my repairs.

I am currently working on a hot-swap power supply for some network equipment. The SMPS was dead. I opened it up and found a domed and leaking (from the bottom) main filter cap and a blown fuse. I replaced the crap brand cap with a Panasonic ED series and then substituted a 100W light bulb for the fuse to see what else was wrong. To my surprise, the light bulb stayed dark and 12.01 volts appeared on the PS's output. It appeared that the capacitor going out caused the fuse to blow. I then put a new fuse in and checked -- 12.01 volts on output. There was just one problem. The capacitor I had was too tall for the enclosure. I ordered a shorter one with the same rating with my next digikey order, but I couldn't find it in the same series. The original one was panasonic ED; the one that has the lower height is a panasonic EE. It looked like the EE was an improved ED from the literature.

When I tested the ED and got the 12.01 volts, the input draw I noted as 6 watts, by the way, after an initial brief (one second or so) higher draw of about 20. I desoldered and then soldered in the EE series, but when I went to check the output, I was getting 0 volts! I noted that the current draw was initially high for about a second like before, but it then went down to 1 or 2 watts after that.

I verified the solder work, and it seemed okay. I checked for continuity to other points along the traces for each lead, and it's showing continutiy.

I haven't tried to swap the previously working capacitor back in yet to see what happens, but I am really puzzled by this.

Any ideas what could be wrong?

[Per Hansson]

Except for something simple like the cap being in backwards it could also be a coincidence and something went marginal after the first failure and now failed...