Offline Flyback: NCP1239 datasheet interpretation?

Hi,

I expect the 1.2mA to be the nominal current that needs to be applied to the FAULT pin.
So, in order to SAFELY activate the OVP function I´d apply "1.2mA PLUS SOME MARGIN"

Let´s say your external circuit activates at about 36V.
Let the voltage across CE of Q4 be 0.5V
so the current through R32 is (36V - 0.5V - 3V) / 27k = 1.20mA
The extra current through R23 is 3V / 22k = 0.14mA
leaves a pin current of only 1.06mA

In my eyes not a safe level to activate OVP.

***
The max pin voltage is 5.5V ... thus you may apply (5.5V - 1.35V) / 1.35kOhm = 3.07mA (nom)

--> I see no problem in driving the PIN with 1.5mA...2.0mA (at least 20% margin)
If you still want to protect the pin you could add an external 4.7V zener.

****
Disclaimer: I did not read the complete datasheet. Thus I may have missed some informations.

Klaus

Wonder why are you using a zener, 5 resistors and 2 transistors where the datasheet application uses a zener and an optional series resistor for OVP fault detection?

Datasheet says:

To reach the upper threshold, the external pull-up current has to be higher than the pull-down capability of the clamp, i.e. 1.2 mA.

Click to expand...

Unfortunately there's no maximum rating for current into fault pin, you may want to ask Onsemi for a specification.

Can't make sense of figure 39...
If there really is a 1.35V zener on that node, then it's impossible for the OVP comparator to ever trip (threshold is 3.0V nominal)
I think that Rfault(clamp) must actually be in series with the Vfault(clamp) zener, rather than in series with the Fault pin.

You are right. There´s something wrong.

From the formula and math example I guess the +IN of the OVP comparator needs to be connected with the left side of the resistor.

Klaus

Fault detection schematic doesn't look completely plausible, I assume that threshold specification is nevertheless correct.