High voltage startup circuits of flyback controllers (NCP1239 & NCP1219)

You need to consider the voltage ratings of the resistors, generally higher wattage resistors have higher voltage ratings.

You also need to consider reliability, designers with only theoretical experience often put resistors in series as shown - these tend to be VERY, VERY unreliable. Resistors in series in this way commonly tend to go high or O/C, as new TV's, VCR's etc. were released you could predict which ones would have failing PSU's because of the use of resistors in series.

If they are specifying particular sizes of resistors it would be foolhardy to make them smaller!.

**..Question 2..*
Hi,
Page 18, fig 41 of the NCP1219 datasheet shows a circuit to stop the NCP1219 from switching.
(by pulling down the FB pin to ground)
Does this same circuit also work on the NCP1239?
Its just that we are trying to make a single PCB which we can populate with either NCP1239 or NCP1219B.

NCP1219

NCP1239

AND9296

AND8393

Flyback said:

NCP1239 or NCP1219B.

Click to expand...

They both have the same FB mechanism. When the output voltage is too high the FB pin gets pulled low.

**..Question 2..still*

When the output voltage is too high the FB pin gets pulled low.

Click to expand...

Thanks, yes, though i am not speaking of that particular use of the FB pin....that one may not shut it down..it may still switch on and just switch straight off....still letting some power through...like the ucc28c43 when you pull its FB pin low...it still lets thru some power. I speak of actually shutting it down, and making it not switch at all. In NCP1219B, pulling the FB pin to ground "does" shut it down...but does it with the NCP1239?

As you know , there is a difference between "shutdown" and "demand value pulled low but its still actually running".

Its like pulling the FB pin down doesn't always shut down a PWM controller. By pulling FB down, what you often do is just ground one input to the PWM comparator....and the demand input is also at 0V....so you have a comparator with both inputs at 0V...and the noise often makes it trip....the noise can make either input look higher than the other, depending on what ambeint noise gets picked up by either input.

**..Question 3..*
Hi,

Pages 11 & 12 of the NCP1203 flyback controller datasheet say that the connection of the NCP1203 to the HVDC rail via its HV pin (so that it can start itself up)
can also bring about catastrophic latch up of the NCP1203.
It says that series resistor values of >4k7 must be added in series with the HV pin to avert this danger.
One presumes the entire NCP12XX range of controllers are also susceptible to such disaster?

Though why not just insert a 400V SOD123 diode in series to the HV pin instead? The diode would stop the reversing current inherent in latch up.
Also, it would dissipate hardly anything since the startup current is a maximum of 18.5mA and diode Vf is only 1V.

So in the schem of page 2 of AND8393, why do they use D13 & R41 & R42? All they needed to do was connect D13 from the HVDC bus to the NCP1219 HV pin and its job done?
A SOD123 is only 3.85mm by 1.7mm, so its smaller than the two 0805 resistors that onsemi always recommend with the NCP12XX application schematics.
So why are they using resistors when they don't need to?
It cant be too disallow startup below X volts because 3.64kOhms x 5.5mA is only 19.8 Volts.

NCP1203

NCP1219

NCP1239

AND9296

AND8393

Why have we thirty threads on the SAME SUBJECT..

I'll need to trim afew.

Leave this one and aske the questions here!!

Thanks very much .. and well spotted, sorry for spreading them too widely......as you obviously know well, NCP12XX is a great family of controllers, so these qu's will be of value to a great many folks
**Question 4..**
Hi,
Would you say the below output overvoltage
protection circuit is OK? How much current is it safe to
put into the latch/skip pin? Datasheet doesnt say.

Output overvoltage protection for NCP1219B...

NCP1219B