T
treez
Guest
Hello,
Do you know what all the fuss is about Quasi-Resonant Offline Flyback converters for wide mains input (85-265VAC)? Do you agree that Quasi Resonant Flyback controllers are massively over-hyped?
In my humble opinion, a Fixed Frequency, DCM Flyback is usually better. –Because it allows a smaller transformer design, and doesn’t end up with very low switching frequency at low mains (85VAC).
Take this Flyback Design case for example….
Spec:
Vin = 85-265VAC
Vout = 25V
Iout = 1.3A
Here it is fleshed out with a (1) Quasi Resonant Flyback and a (2) Fixed Frequency Flyback……
(1)…With DCM , Current mode, Fixed Frequency Flyback (100kHz)
Lp = 380uH, Ls = 23.8uH (Ns/Np = 0.25)
Max I(pri)pk = 1.43A
Max I(sec)pk = 5.68A
….goes into CCM but only when in low USA mains (<110VAC)
(2)….With Quasi-Resonant Flyback…
Lp=940uH, Ls = 58.8uH (Ns/Np = 0.25)
Frequency varies from 30kHz to 90kHz over the 85-265VAC mains variation
Max I(pri)pk = 1.64A
Max I(sec)pk = 6.56A
………………………………………………………………………………………..
As you can see, the Quasi Resonant Flyback is obviously going to have the bigger transformer. This is because it has to operate at a significantly low frequency at low USA mains……we cannot design the Quasi-Resonant Flyback to have a higher frequency at low USA mains, because if we do that, then the switching frequency is then too high when running off high UK mains (265VAC).
…………………………………………………………
I appreciate that the Quasi Resonant Flyback benefits from a slightly lower switch-ON switching loss due to the FET switching on after the drain voltage has fallen somewhat, however, the fall in drain voltage won’t be that much, because to get a big fall in drain voltage, one needs a transformer with a low Ns/Np value, and that unfortunately reflects back too much voltage to the primary. In the above example the drain voltage cannot fall more than 100V below the input voltage anyway.
Another point, is that Quasi Resonant Flybacks only reduce switch-on switching losses, and in most cases including the above, they end up having worse switch-off switching losses than the Fixed Frequency Flyback (because the peak FET current at turn-off tends to be higher with a Quasi Resonant Flyback design.) Another point, is that the Quasi Resonant Flyback only reduces “capacitive” (Cds) switch-on switching losses….this is because with a DCM or Quasi Resonant flyback, there are no “overlap” switching losses anyway. –And “overlap” switching losses tend to be the worst type of switching loss.
Add to this the fact that few Quasi Resonant Flybacks actually manage to switch-on at the absolute trough of the drain voltage “valley”, (-at least not over the entire input voltage range) and you can see that the Quasi Resonant Flyback really is quite a let down.
With the Fixed Frequency Flyback, you can actually “cheat” and design it to go into CCM at low USA mains…this significantly helps in reducing the max. I(pri)pk value. Of course, this cannot be done with a Quasi-Resonant Flyback…the Quasi Resonant controllers on the market don’t ever allow CCM operation.
So do you agree that Quasi Resonant Flybacks are a big let down compared to Fixed Frequency Flybacks?
It is very strange that virtually all the new Offline Flyback Controllers being brought out to the market are all Quasi Resonant Flyback controllers. It is really hard to find a decent Fixed Frequency Offline Flyback controller. There is the NCP12XX range by onsemi, but these are susceptible to noise due to their low feedback current level. They also only have the standard “overload protection timeout” method for output short circuit protection. (ie they don’t have a ‘second’, higher current limit to prevent the ‘staircasing’ primary current that occurs before the “delayed” overload protection kicks in.)
The UCC28C4X chips are Fixed frequency, but have no internal start-up regulator, and no internal overload protection.
Why is all the effort being put into making Quasi Resonant Flyback control chips when they are not that good?
Do you know what all the fuss is about Quasi-Resonant Offline Flyback converters for wide mains input (85-265VAC)? Do you agree that Quasi Resonant Flyback controllers are massively over-hyped?
In my humble opinion, a Fixed Frequency, DCM Flyback is usually better. –Because it allows a smaller transformer design, and doesn’t end up with very low switching frequency at low mains (85VAC).
Take this Flyback Design case for example….
Spec:
Vin = 85-265VAC
Vout = 25V
Iout = 1.3A
Here it is fleshed out with a (1) Quasi Resonant Flyback and a (2) Fixed Frequency Flyback……
(1)…With DCM , Current mode, Fixed Frequency Flyback (100kHz)
Lp = 380uH, Ls = 23.8uH (Ns/Np = 0.25)
Max I(pri)pk = 1.43A
Max I(sec)pk = 5.68A
….goes into CCM but only when in low USA mains (<110VAC)
(2)….With Quasi-Resonant Flyback…
Lp=940uH, Ls = 58.8uH (Ns/Np = 0.25)
Frequency varies from 30kHz to 90kHz over the 85-265VAC mains variation
Max I(pri)pk = 1.64A
Max I(sec)pk = 6.56A
………………………………………………………………………………………..
As you can see, the Quasi Resonant Flyback is obviously going to have the bigger transformer. This is because it has to operate at a significantly low frequency at low USA mains……we cannot design the Quasi-Resonant Flyback to have a higher frequency at low USA mains, because if we do that, then the switching frequency is then too high when running off high UK mains (265VAC).
…………………………………………………………
I appreciate that the Quasi Resonant Flyback benefits from a slightly lower switch-ON switching loss due to the FET switching on after the drain voltage has fallen somewhat, however, the fall in drain voltage won’t be that much, because to get a big fall in drain voltage, one needs a transformer with a low Ns/Np value, and that unfortunately reflects back too much voltage to the primary. In the above example the drain voltage cannot fall more than 100V below the input voltage anyway.
Another point, is that Quasi Resonant Flybacks only reduce switch-on switching losses, and in most cases including the above, they end up having worse switch-off switching losses than the Fixed Frequency Flyback (because the peak FET current at turn-off tends to be higher with a Quasi Resonant Flyback design.) Another point, is that the Quasi Resonant Flyback only reduces “capacitive” (Cds) switch-on switching losses….this is because with a DCM or Quasi Resonant flyback, there are no “overlap” switching losses anyway. –And “overlap” switching losses tend to be the worst type of switching loss.
Add to this the fact that few Quasi Resonant Flybacks actually manage to switch-on at the absolute trough of the drain voltage “valley”, (-at least not over the entire input voltage range) and you can see that the Quasi Resonant Flyback really is quite a let down.
With the Fixed Frequency Flyback, you can actually “cheat” and design it to go into CCM at low USA mains…this significantly helps in reducing the max. I(pri)pk value. Of course, this cannot be done with a Quasi-Resonant Flyback…the Quasi Resonant controllers on the market don’t ever allow CCM operation.
So do you agree that Quasi Resonant Flybacks are a big let down compared to Fixed Frequency Flybacks?
It is very strange that virtually all the new Offline Flyback Controllers being brought out to the market are all Quasi Resonant Flyback controllers. It is really hard to find a decent Fixed Frequency Offline Flyback controller. There is the NCP12XX range by onsemi, but these are susceptible to noise due to their low feedback current level. They also only have the standard “overload protection timeout” method for output short circuit protection. (ie they don’t have a ‘second’, higher current limit to prevent the ‘staircasing’ primary current that occurs before the “delayed” overload protection kicks in.)
The UCC28C4X chips are Fixed frequency, but have no internal start-up regulator, and no internal overload protection.
Why is all the effort being put into making Quasi Resonant Flyback control chips when they are not that good?