Cisco switch PSU dead

[sebr]

Unfortunately my 3yo Cisco Catalyst 3560 Series PoE-8 Switch died the other day. No signs of life anymore. The PSU of it has two separate connectors: one supplying 12V and the other 3.3V and -50V (thats for the PoE stuff I guess). I hooked up an external PSU supplying 12V and the switch boots up fine, web interface shows up
So i checked the PSU: no voltage on any of the supply rails (12V, 3.3V, -50V), fuse isn't blown, no visual damage or burnt components, no bulged caps. The discrete semconductors (diode, zener, transistor / FET) also all tested without short and the caps are in their range (capacity). Across the main cap i can measure stable 400V. Any ideas how to investigate this further? Any known problems with these PSU's (it's made by Liteon)?

[Khron]

Re: Cisco switch PSU dead

400v on the main cap means the PFC is most definitely working.

Have you identified / can you trace the output rails from the diodes all the way to the connectors? Try measuring the voltage on each of the output caps, see if / what you get.

Failing that, see if you can identify which of the blue Rubycon caps on the primary side are responsible for filtering the PWM chip's (IC802) supply line. What's the model number of that chip, by the way? Datasheet should indicate how many volts it should be getting, and where (ie. which pin).

[sebr]

Re: Cisco switch PSU dead

Yup, PFC seems to work. There is no short on any output caps but they all measure 0V. Seems like the PSU doesn't come come up.
Those blue Rubycon aren't for the PWM controller and on the bottom IC802 is just a voltage comparator. On one of these daughterboards (left) soldered on the board is the PWM controller L6598 and the orange cap for it's supply which is only arund 1V across that cap what doesn't look good at all. :-(

[Khron]

Re: Cisco switch PSU dead

I've found an application note:



Can you try to trace where the L6598 gets its supply voltage from? Maybe see if the general circuit resembles the one on page 14.

[sebr]

Re: Cisco switch PSU dead

Sorry no chance to trace it further on this board. But from what I can see the circuit is totally different than in the application note. Many components glued to the board and also two addon board plus many jumper wires - it's a mess.

[Khron]

Re: Cisco switch PSU dead

Even the Vcc line that supplies the PWM chip (via that orange cap)?

On one hand, that's supposed to come from the main high-voltage bus (to start up the PWM chip), after which it might/should be supplied by the rectified voltage from an auxiliary winding on the main transformer.

[sebr]

Re: Cisco switch PSU dead

I could trace this track on the PCB and the components along it up to the aux winding of the transformer - also without any component faulty. While measuring I was wondering about the readings that sometimes jumped like crazy. first i thought it was just bad contactof the probes but then i decided to resolder the whole board and that finally did it!
Before I inspected the solder joints even with magnifying glass but there was nothing suspicious!
Anyways I'm glad it's working now