2
\$\begingroup\$

I've been having some issues with a Philips TV power supply (715G6677-P02-001-0020 PSU) which was burnt.

The old controller (LD5532) was burnt and I couldn't find any supplier of that exact IC, but, there was an equivalent IC for selling, PF6003AHS, which was supposed to be equivalent.

And Philips used this IC in another power supply (715G6934P01002002H) which uses basically the very same topology, with a single difference: The pin 7, which sould be a NC, is actually not connected to anything in the first one but grounded in the second one.

Here is a section of the one using LD5532:

enter image description here

And here using PF6003AHS:

enter image description here

But, somewhy, grounding it does make a difference.

Before grounding, the voltage in the VCC pin starts rising as soon as it powers up, and slowly goes up to 36 VDC. That's due to those high resistances to the AC input which serve to bootstrap it.

When I ground the pin 7, it doesn't rise above 1.5 V.

And in both cases (grounding or not), the voltage before diode D9393 is 0 V. There's only voltage after it (on the cathode side), showing that the IC is not booting up and switching mosfet Q9301 at all.

The voltage on its VCC comes solely from those resistances to the AC in at all times.

So, now I'm wondering: Could there be any large discrepancies making it fail?

And btw, I've tested nearly everything else.

There's voltage in the HV capacitors, and D9393, R9307, D9302, C9304 are all fine as well.

And it's really, really nonsensical that grounding a NC should make any difference. It should be an NC, but clearly isn't.

When measuring with a multimeter in diode mode, it displays 0.930 V with the positive lead in pin 7 and negative lead in pin 8 (in the inverse polarity it doesn't measure anything).

What could be happening here?

Thanks!

\$\endgroup\$
3
  • \$\begingroup\$ A datasheet is unlikely to give clear detail about NC pins, which may be related to internal or test functionality. It might not be possible to get a clear answer without intense reverse engineering. \$\endgroup\$
    – nanofarad
    Commented Jul 13, 2021 at 4:27
  • \$\begingroup\$ If it says NC meaning no connection, any schematic showing a connection could be in error. Your tests suggest this also \$\endgroup\$
    – Tony Stewart EE75
    Commented Jul 13, 2021 at 4:45
  • \$\begingroup\$ @TonyStewartEE75 yes, but I bought two of these and both give the same measurement. Plus all datasheets and schetmatics that use this IC call it NC, but ground it anyway. And It's also NC in the LD5532. Something is sketchy here. \$\endgroup\$
    – user2934303
    Commented Jul 13, 2021 at 4:51

0

Reset to default

Browse other questions tagged or ask your own question.