Two Switch Flyback App Note says Duty cycle must be less than 0.5

[treez]

Hi,
The following App note on 2 switch flyback says that the controller used needs a max duty cycle of 0.5 (page 7)..
AN5287

https://www.st.com/resource/en/application_note/dm00597621-170w-high-input-voltage-two-switch-flyback-based-on-l6565-and-1500v-k5-mosfets-stmicroelectronics.pdf

However, the App Note uses the L6565 controller, which doesnt limit its duty to 0.5
L6565 datasheet

https://www.st.com/resource/en/datasheet/l6565.pdf

Do you agree that a 2 switch flyback does not need a duty cycle limit as such?...it just needs enough off time for the leakage inductance to be discharged down to zero Amps.
In a 2 switch flyback, the voltage used to discharge the leakage inductor is Vin - Vout * NP/NS.

Surely its as long as the controller off time is longer than dt = (L.di)/V ?
where V = voltage across leakage L; L = Leakage inductance. di = peak primary current.

Thus surely the main thing needed for 2 switch flyback is an undervoltage shutdown which shuts down the 2 switch flyback whenever vin goes too low?

 

[FvM]

Do you agree that a 2 switch flyback does not need a duty cycle limit as such?

The exact statement would be: "A two-switch converter can't transmit power to the secondary when the duty cycle is > 0.5". You need to take care that the controller can't wind-up to a higher duty cycle.

 

[Easy peasy]

the tx reset time needs to be equal to the on time if the reset volts are the same, as they are with the classic 2 sw f-back, if you don't have the diodes but snubbers instead - then yes you can exceed 0.5 D-cycle ....

 

[treez]

And we agree that in a CCM flyback, the transformer never gets reset...it is always fluxed.?
Its just the leakage inductance that needs resetting in a flyback, surely?

And with many an offline 2sw flyback, the leakage inductance gets reset in 200ns or so after the fets switch off?

 

[mtwieg]

No, this is one of the downsides of the two switch flyback. If D>0.5 (plus a bit of margin for the catch diodes) then you will get staircase saturation. And Vout is therefore limited to Vin*Ns/Np. Same as the forward converter.

 

[Easy peasy]

all transformers have to have 0v ave applied, thus the reset v.sec must equal the driven v.sec to avoid staircasing ....

 

[treez]

Thanks, this is interesting , you are obviously correct. The "v.sec ON should equal v.sec OFF" law bypassed me for some reason.
I was kind of thinking in terms of "flux in the core".....if in every switching cycle, the core flux goes to zero, then the transformer cant possibly saturate......so i am asking myself why a 2-sw flyback with D>0.5 can't de-flux its core within the OFF time. Hopefully there will be time to look into this soon

 

[Easy peasy]

because of the Vclamp ( actually Vin ) provided by the diodes, this can be a blessing or a curse as it limits Vout to Vin x the turns ratio.

if you remove the diodes in a 2 sw f-back and put snubbers on the Tx instead, then you can have more flyback volts and more Vout and more D, but the diodes have to go ... ( on the pri ).