Schottkys in series

Indeed. I have an about 10 year old Kodak picture frame, it is powered by a little 5V wall wart. We have tropical heat alert here, with temperatures above body temperature. This weekend I noticed the Kodak thing had stopped working. Worked OK on the lab supply at 5V. Opened the wall wart, 2 680 uFcaps in parallel were all swollen, classic rectifier circuit, 1 schottky. Replaced the caps with 2 470 uF low ESR mobo caps, fixed it.

But why would they fry? Sure if the diode gets hot there is a large AC component.. Weather was a bit too much for it it seems. Maybe it will fry the new caps again the next few days, must be very hot in that super small wallwart. Else replace with some better diode. Dead caps is quite a common fault in wallwarts and power supplies, and almost always some Schottkies are used. Maybe those diodes are responsible, not the caps. In a high power design I did long ago I had those on heatsink.

Thank you.

At 125C and DC back-bias of 35 volts, that could dissipate over 8 watts from reverse leakage. Without serious heat sinking, it could easily go into thermal runaway. Being designed for use in switchmode power supplies, it wouldn't be expected to run at high DC back-bias, I guess.

This is a very low-barrier part, so it conducts really hard in both directions. Vf is only 0.4 at 10 amps.

Right. The low Vf was why we picked it for a previous design, a buck switcher with 5V output where the efficiency was important. It was a lot less stressed in that application. Nowadays we'd probably go for a synchronous topology. In hindsight a regular PN diode would have probably worked just as well for the boost switcher, where the Vf is a much smaller fraction of the output voltage. But Schottkies do also have the advantage that (IIRC) they have no reverse recovery mechanism. I tried a bunch of different silicon diodes in simulation and the t_rr was responsible for about half the power loss.

Unless you push them so hard that the PN guard ring conducts, schottkies have no reverse recovery. The 0.4 volt drop part probably won't turn its guard ring on.