Philips uses part exceeding max rating.

[Bud]

I was repairing a Philips 40'' flat screen TV when noticed a few switching regulators rated at 20V absolute max were powered from a 21V rail (I measured and confirmed the supply voltage was 21V). What is the deal with this, Philips?

[ataradov]

Yep, it is very common, and manufacturers like Philips don't do that on a whim, they actually talk to the part manufacturer and get real ratings. Manufacturer may provide higher failure rate or shortened life time, and the customer has to accept it.

[Bud]

Now we talking planned obsolescence 

[ataradov]

Now we talking planned obsolescence 

Yep. Some even reputable LED light bulb manufacturers use 85 C capacitors derated for 105 C operation. Their life span is reduced considerably, but it looks like it is about the same as the life span of a light bulb. From what I've seen LEDs die first anyway. So while this approach seems sketchy, it really is not, it is just balancing things out.

[magic]

I wonder if it could be a matter of binning, so that if you order the same part number from an ordinary distributor you may get a part which will fail in such out-of-spec conditions?

[Mr. Scram]

I was repairing a Philips 40'' flat screen TV when noticed a few switching regulators rated at 20V absolute max were powered from a 21V rail (I measured and confirmed the supply voltage was 21V). What is the deal with this, Philips?

Companies like Philips can and will do their own verification. Using something outside of the specifications isn't an issue if you've done the homework and understand what you're doing. The manufacturer establishes a baseline performance, but it's very possible that the specific scenario the part is used in leaves a little more room. Dave has done a video on this too. Binning would be another option. Some companies bin regular parts to eek out the higher performing parts which are sold at a premium.

[mariush]

There's a mm5z15b  zener diode there in the schematic, won't that cut the voltage down to around 15v ?

Also i would think traces being thin maybe there's some voltage drop due to trace resistance etc etc

Maybe Phillips ordered a custom version which can handle higher voltages?