Converting an offline Flyback to using the Topswitch?

Hello,
We have an offline flyback currently using the onsemi NCP12XX range of controllers that we would like to convert to a Topswitch-jx design, because we believe it will be easier to maintain and modify the design within the Power Integrations ‘Development system’.
However, does our primary di/dt mean that we can’t use Topswitch-jx?

The Flyback spec. is as follows:-
Flyback
Vin = 85-265VAC
Vout = 25V
Pout = 32.5W
F(sw) = 66kHz
Discontinuous Conduction Mode
NS/NP = 0.4535
L(pri) = 270uH

We see that our peak primary current of 2.1 Amps means that we need to use the TOP266 part..

However, on page 27 of the Topswitch-jx datasheet, it says that the peak current level of 2.55 Amps (for TOP266) depends on the primary di/dt being 530mA/us.
This seems to be a problem because in our design, the primary di/dt is 1370mA/us (when at 265VAC).
In fact at any DC bus voltage above 145VDC, our primary di/dt is above 530mA/us.

So does this mean we cannot convert this particular design to a Topswitch-jx design?
(I appreciate that we could re-design it to get a lower primary di/dt but that would then mean needing to re-do the transformer)

Topswitch-jx datasheet:
https://ac-dc.power.com/sites/default/files/product-docs/topswitch-jx_family_datasheet.pdf

- - - Updated - - -

Please find attached an "equivalent" flyback schem and LTspice sim as the NCP12XX one that we wish to convert to Topswitch

This is an issue because you've chosen to design it to operate at a very low duty cycle (10% at max line). Only easy way I can see is to use the higher frequency version, so that your required peak current will be lower.

Also keep in mind that when your current ramp is faster, the effective peak current limit will be higher. But you don't want to push it so far that it exceeds the SOA of the FET.

I think the problem is that the primary inductance is too low for the topswitch design, and brings about a too high di/dt, since as you know, di/dt = V(pri)/L(pri).
This appears to show a limitation of the Topswitch, however, as discussed, it can be solved by re-doing the transformer with a higher primary inductance, -I just wondered if anyone has data for the topswitch current limit for higher di/dt's than the ones in the datasheet?

I wouldn't say the topswitch isn't cabable of working with higher di/dt, but it will affect its ability to limit peak currents due to the leading edge blanking and current limit delays, so short circuit protection may not work as well.

You can probably approximate the effective peak current limit vs di/dt by using the current limit delay time given in the datasheet. If your di/dt is 1.37A/us and the delay is 0.1us, then the peak current should be higher by about 0.137A.

Sorry but I don’t think this is to do with the Leading Edge Blanking time. This is to do with the actual current limit after the L.E.B. has elapsed.

Yes I agree that there are situations with L.E.B. that might well pertain to use of certain di/dt’s , but that isn’t the question I’m asking here...sorry about that...actually thinking about it the effect with L.E.B. is indeed a very good point, its just not the one i'm asking about here.

The fact that the datasheet gives certain, particular di/dt values makes me suspicious that going above these di/dt values is not advisable.
I couldn’t be sure because the topswitch datasheet doesn’t elaborate.

- - - Updated - - -

Actually ive just done a Flyback design on the Power integrations PI Expert software, and at max vin , they have a primary di/dt of 670mA/us even though the datasheet of topswitch-jx page 27 states “270mA/us” for the TOP264 that their software selected.

So it “appears” that you can have a di/dt above what the datasheet states, but who knows how far above it one can go, and what it means in terms of actual peak current limit.

I am trying to extrapolate what the actual peak current trip point will be for a TOP264 with di/dt = 1.37A/us. Page 27 of the datasheet states that there is a 100ns “current limit delay”.
Now, the TOP264 is said (page 27) to have a peak current trip, when di/dt = 0.27A/us, of 1.3A.

-Therefore, one presumes that the actual trip current is 1.27A {1.3A – ([di/dt] * 100ns)}

And therefore, if the primary current di/dt = 1.37A/us, then presumably the trip current will be 1.41A? {1.27A + (100ns * [di/dt]) }

Do you think my linear extrapolations of the current trip point are correct?

As you can imagine, it is essential to know what is the actual current trip point.


Topswitch-jx datasheet:-
https://ac-dc.power.com/sites/default/files/product-docs/topswitch-jx_family_datasheet.pdf