Intermittant/pulsing power supply

[mb300sd]

I have power supply out of a DGS-3224 network switch. The part number is Skynet DLN-Z097, but I cannot find a suitable replacement. It'll work once in a while, but will usually output 1v with quick pulses up to 8v. I believe it is the SMPS trying to start up. I have already replaced the output capacitors, which did not solve the problem.

Does anyone know of either a replacement part number or have any hints on repairing it? It seems that it's a common pinout among power supplies for network equipment, but I can't find one that is the same size and output.

Similar:
http://www.ebay.com/itm/DELTA-ELECTRONICS-DPSN-100BP-AC-CISCO-34-0966-04-POWER-SUPPLY-PS339-/251161708364?pt=LH_DefaultDomain_0&hash=item3a7a67874c
http://www.ebay.com/itm/151146265995?_trksid=p2055359.m2763.l2649&ssPageName=STRK%3AMEBIDX%3AIT

Datasheet for another similar supply with proper pinout: http://www.usbid.com/assets/datasheets/34/dpsn-150bp_a.pdf

The second ebay link looks like it will fit, but is only 4.5A when the original is 7.5A. I will never use all 24 ports on the switch, so it'll probably work, but I'd like to repair it or replace it with the same capacity if possible.

[wraper]

Sounds more like small capacitor on primary side failed. Big high voltage capacitor is possible too. Check if rectifier diode on secondary side is not shorted, however not likely as you get some voltage on the output.

[mb300sd]

Sounds more like small capacitor on primary side failed. Big high voltage capacitor is possible too. Check if rectifier diode on secondary side is not shorted, however not likely as you get some voltage on the output.

I think your right. Scope of the HV capacitor, should I pretty much assume its blown? I don't have anything similar on hand so I'll have to order one.

Edit: Yep, popped it out and it's measuring under 100nF on the multimeter.

[SeanB]

It's dead. Look out for smaller electrolytics as well, as they also fail open as well and it will not start up.

[AKM]

If it is trying to start (Hiccup mode), I would check the secondary side for shorts first; even the cable.

[mb300sd]

Well, I managed to fix it for about 2 seconds before I broke it again Got all excited after the new capacitors finally showed up and went to put them in...

Accidentally left a piece of desoldering braid under the board and powered it up, something in this area let out a small puff of magic smoke. But I can't seem to find any burnt parts.



I'm pretty sure I blew something in the feedback curcuit, because it's now outputting 7V at enough current to spin a small motor. Voltage drops with load.

I'd hate to have put in all this work (and waiting!) and toss it now

[amyk]

Am I the only one wondering about the heatsinks' fins covered in hotmelt glue?

[Shock]

Don't forget to discharge the caps to be safe.

You now need to see if you still have a shorting component or it's shorted and now open. Look close up for blackened or broken components and check for burnt traces all over.  Check fuses are good.

[mb300sd]

Am I the only one wondering about the heatsinks' fins covered in hotmelt glue?

There was a plastic card glued on top, I assume to protect from shorting to the case if it gets dented. I had to pull it off to get to the caps

[Richard Head]

The pulsing symptom is a common one. It's normally caused by either the high value bleed-on resistors from the HVDC bus going high in value, or the small value electrolytic capacitor across the controller power pins getting a bit leakey or low in value. This is how it's meant to work:
When power is first applied to the unit the bleed-on capacitor starts charging through the series resistors. When the voltage across the cap reaches the turn-on threshold of the controller chip the bleed-on cap starts powering the controller chip. The voltage feedback (from a feedback winding generally) then powers the chip for the time it takes for the feedback voltage to climb above the turn-off threshold.
If the cap leakage current starts to increase (with age) the bleed-on resistors struggle to provide sufficient current to overcome the leakage and the cap voltage never reaches the turn-on threshold. Result is PSU never starts up.
If the cap value decreases then the controller supply pin voltage drops to the turn-off threshold before the feedback voltage has a chance to power the chip. Result is that the PSU pulses as it continually tries to start up.
Sounds more complicated than it actually is.

[mb300sd]

Well, I fixed the original problem but couldn't undo the damage I did from shorting it. So, alternate solution: cheap 8A 12v supply and a few wires from a dead ATX supply. The switch dosen't seem to care about the ps_on or status lines. Now to deal with the unexpectedly loud fans