Switch mode PSU component identification.

[blueadept]

Hi EEVBloggers!

I'm currently looking at a switch mode PSU... I'm not sure how common this is, but there appears to be a small, separate supply using a VIPer22a chip driving the fans and also the control circuit for the main PSU... and it is this which appears to have failed... there's precious few components in this thing, but my attention is focussed on the shunt regulator in the feedback circuit... See attached image... I believe this is something not to far from a TL432, but can anyone explain the marking to me?

James

PS, Also, if it's a TL432, i wonder why they didn't use the TL431, it seems identical apart from the pinout, but much more common.

[blueadept]

"DaveCAD" skills not withstanding...  I think with no feedback there is some kind of failsafe mode in the VIPer22a chip as no explosion has taken place.... the VIPer22a chip and the optocoupler appear good.

[SeanB]

Change the 22uF 50V capacitor and see if it powers on then. That is the power supply of the chip, and if it is leaky ( or the 10R resistor has failed open or the diode or the transformer coil is open or shorted) then the chip will either sit in start up capacitor charge mode forever, or will run at single cycles every second or so.

[blueadept]

Change the 22uF 50V capacitor and see if it powers on then. That is the power supply of the chip, and if it is leaky ( or the 10R resistor has failed open or the diode or the transformer coil is open or shorted) then the chip will either sit in start up capacitor charge mode forever, or will run at single cycles every second or so.

That's absolutely great, it gave me the push to read the principle of operation of the VIPer22A chip properly... and realise that there is an internal supply that feeds OUT of Vdd during start up... which means there is something I can test in a stable state not just a few pulses as the power comes on.... from that point it's almost trivial, there's no power, and it can't be the transformer shorted, because there's a diode, so as you say, it's only the diode or the cap, and since it's sitting at 0.6v I picked the diode... spot on, switched it for one salvaged from a duff LED lightbulb and we're all working again...

Cheers for the push...

[blueadept]

Of course, I'm still in the dark about how I would identify the shunt regulator in the photo if it had been the fault.....

[Agent24]

Of course, I'm still in the dark about how I would identify the shunt regulator in the photo if it had been the fault.....

Look it up in an SMD codebook

[blueadept]

Of course, I'm still in the dark about how I would identify the shunt regulator in the photo if it had been the fault.....

Look it up in an SMD codebook

Where in the codebook does the symbol "Little house with a goat in it" come?

[Agent24]

Quote from: blueadept on Today at 11:02:05 AM

Where in the codebook does the symbol "Little house with a goat in it" come?

I am not sure what you mean. The part I see looks like it is written RA IK. But the R maybe 20 on its side. I am not sure.
One half of that code is probably a batch\date ID and the other the actual part number.

Can you confirm what the printing actually says or get a better photo?


Find a dubious SMD code website or try to get a professional codebook. Then look up RA IK, hoping you read it properly. Then hope you find it, and when you can't, realise why SMD markings are a PITA and why I love through-hole parts without cryptic lettering.