Look up "burst mode" (I think)
(or possible discontinuous mode)
If there was no load and you kept putting out pulses the output voltage would rise...probably not what you want.
This is where some small SMPS have a minimum load requirement because they cannot go to 0% duty cycle.
(or possible discontinuous mode)
If there was no load and you kept putting out pulses the output voltage would rise...probably not what you want.
This is where some small SMPS have a minimum load requirement because they cannot go to 0% duty cycle.
The flyback stores energy in the core and airgap during on-time of primary power-MOSFET. During off-time that energy is transferred to the load. Hence there is always some magnetizing current present.
To keep track with various load input energy must be controlled. This can be done by controlling primary peak current, for instance.
In contrary to forward or buck converters, the duty-cycle relates to input voltage and primary flyback voltage, i.e. reflected transformed secondary voltage.
Btw, I designed lots of flyback converters over the last years. But I would not go for them to power any audio-amp.
To keep track with various load input energy must be controlled. This can be done by controlling primary peak current, for instance.
In contrary to forward or buck converters, the duty-cycle relates to input voltage and primary flyback voltage, i.e. reflected transformed secondary voltage.
Btw, I designed lots of flyback converters over the last years. But I would not go for them to power any audio-amp.
It's probably easiest not to contemplate on magnetising current for DCM flyback, but rather think more about leakage inductance and how that contributes to waveforms after the main primary to secondary energy has transferred on a cycle by cycle basis.
Each cycle can effectively have a different period (frequency) with DCM. There will be a little energy still flowing in the (typically primary) ringing waveform after the energy dump, which modern controllers then assess for the valley portion to next turn on the main switch, so as to reduce switching loss. That's a cute part of modern controllers - that was not available in the early days.
Each cycle can effectively have a different period (frequency) with DCM. There will be a little energy still flowing in the (typically primary) ringing waveform after the energy dump, which modern controllers then assess for the valley portion to next turn on the main switch, so as to reduce switching loss. That's a cute part of modern controllers - that was not available in the early days.
This is certainly a point.
All in all stray losses are a big issue with flyback transformers thus producing additional losses in the transformer and requiring a snubber circuit to dampen oscillations.
Power dissipation inside the snubber increases with power, adequate cooling it makes the converter bulky at medium power rates - this beeing one reason not to use flyback converters above power levels of let us say 50~100 watts.
Do not let you trick by the simplicity of the circuit - the flyback-converter is on of the most treacherous topologies.
Most of the design know-how resides in one part: The power transformer!
All in all stray losses are a big issue with flyback transformers thus producing additional losses in the transformer and requiring a snubber circuit to dampen oscillations.
Power dissipation inside the snubber increases with power, adequate cooling it makes the converter bulky at medium power rates - this beeing one reason not to use flyback converters above power levels of let us say 50~100 watts.
Do not let you trick by the simplicity of the circuit - the flyback-converter is on of the most treacherous topologies.
Most of the design know-how resides in one part: The power transformer!
Thank you for the replies.
I will try outputting a very short pulse when output voltage is correct and see how I get on.
I have also used a feedback signal for output voltage level but also one from the secondary to see when the output pulses subsides and so I can apply another pulse.
I will try outputting a very short pulse when output voltage is correct and see how I get on.
I have also used a feedback signal for output voltage level but also one from the secondary to see when the output pulses subsides and so I can apply another pulse.
I'm unsure what aspect of flyback operation you are trying to get a better handle on. Is it the way the primary and secondary currents start and stop in DCM, or what level they start at and stop at in CCM? Or is it a switching transition aspect, such as losses or emi related aspects? Or is it output ripple voltage characteristic, or magnetic core B-H locus.
Certainly in CCM the change in primary current, and secondary current, can be a lot less than the DCM current change that transitions between zero and some controlled peak value. But the core is only working in one-half of its BH for either mode.
Certainly in CCM the change in primary current, and secondary current, can be a lot less than the DCM current change that transitions between zero and some controlled peak value. But the core is only working in one-half of its BH for either mode.
Continuous mode implies hard switching turn-on needed to turn off the secondary rectifier - imho an ugly feature. In that case there is no way for quasi-resonant transition.
On the other hand core utilisation is better than with discontious mode thus reducing ferrite core size.
On the other hand core utilisation is better than with discontious mode thus reducing ferrite core size.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
- Home
- Amplifiers
- Power Supplies
- Flyback SMPS question.