The topology that wins in one enclosure loses in the next. Two programmes with the same output chose differently: the one with a volume constraint took the more complex architecture, the one with a cost constraint stayed with the simple one, and both met their specification.
Charger topology: the four decisions that choose it
The output range and power level, the enclosure volume, the efficiency target and the electromagnetic requirement.
The four are read from the operating envelope rather than from a preference, and each one eliminates a class of architectures. Reading them together is what turns a topology discussion into a design decision with reasons attached.
The common architectures are not ranked; each sits in a range of power, volume and cost where it is the efficient answer, and the boundary between them moves with the switching technology.
Decision 1: output range and power level
Power level narrows the field faster than any other input.
At low output a simple flyback arrangement is usually sufficient; as the power rises, the losses in the primary switch and the leakage energy in the transformer become difficult to manage with the simple arrangement, which is where quasi-resonant and active clamp variants appear, followed by resonant architectures at higher levels. The boundary is a design judgement, and it moves with the switching devices and the magnetics available.
Decision 2: enclosure volume
The volume decides how much magnetics the design can carry.
A resonant architecture can reduce the size of the magnetic components considerably, which matters when the enclosure is fixed by the plug mechanism and the port layout. Where volume is not constrained, the simpler architecture with a larger transformer can meet the same specification at lower cost. The volume, not the efficiency, is often what forces the more complex choice.
Decision 3: efficiency target
An efficiency target at defined load points, not a general aspiration.
Efficiency requirements usually appear as a figure at a stated load and input, and they decide whether the losses in the primary switch are acceptable in the available volume. At high line and full load the difference between architectures is largest, which is why the target has to state its conditions before it can select anything.
Decision 4: the electromagnetic requirement
The pre-scan margin decides how much filter the design carries.
Architectures with softer switching transitions generally produce a lower signature, which can reduce the filter and the layout work. Where the enclosure is small and the filter has to fit inside it, that difference can decide the architecture. The requirement is the margin the pre-scan has to show, not a general statement about emissions.
| Decision | What it selects | Where the boundary moves | Common mistake |
|---|---|---|---|
| Output and power | The architecture class | Higher with better switching devices | Choosing by habit |
| Enclosure volume | The magnetics the design can carry | Smaller enclosures need higher frequency | Treating volume as a packaging decision |
| Efficiency target | Acceptable loss in the volume | Stated at a load point and input | A target without conditions |
| Electromagnetic margin | The filter and layout effort | Softer switching lowers the signature | Emissions addressed after the layout |
The common architectures and where each sits
Four arrangements, described by what they do rather than by a ranking.
A simple flyback arrangement for lower output, where cost and simplicity matter more than efficiency; a quasi-resonant variant that reduces turn-on losses by switching at a valley, which suits moderate output with an electromagnetic requirement; an active clamp arrangement that recovers leakage energy and suits higher output with a GaN switching stage; and a resonant architecture for the highest output, where the tank and the control carry the design effort.
Where the choice is genuinely open
Two cases, and both are decided commercially.
Where two architectures meet the same specification, the choice usually falls to the cost of the additional engineering against the saving in magnetics, the filter and the volume. And where the platform already exists in one architecture, reusing it may be cheaper than an optimal new design, provided the envelope fits inside its characterised range.
What to ask a supplier for
Four answers that show whether the topology was chosen or assumed.
Which architecture the platform uses and why; the efficiency at the required load points; the pre-scan margin; and the magnetics selection with the volume it occupies. A supplier that answers all four has matched the architecture to the envelope; one that answers with a power level has described a category rather than a design.
FAQ
Is a resonant architecture always the best choice at higher power?
It is frequently the efficient answer where the volume is tight and the efficiency target is demanding. Where the enclosure has room and the target is modest, a simpler architecture can meet the same specification at lower cost. Ask what the enclosure volume and the efficiency target actually require, because a simpler architecture can meet a modest target at lower cost.
How much does the switching technology change the boundary?
It moves it upwards: better switching devices reduce the losses that made higher frequencies and resonant designs necessary in the first place, which extends the range where simpler architectures remain viable. Ask whether the chosen switching device moves the boundary far enough to keep the simpler architecture viable at the output being specified. Ask whether the boundary has moved past the output being specified.
Does the topology affect certification?
It affects the electromagnetic profile and the thermal behaviour, and therefore the evidence that has to be established. The construction requirements follow the product and the market rather than the architecture. Ask which evidence the architecture changes: the electromagnetic pre-scan and the thermal measurement, while the construction requirements follow the market. Ask which of the two the new architecture changes.
Can a platform change topology later?
That is a new design rather than an update: different magnetics, different layout, a new pre-scan and usually a new enclosure. Reusing the existing platform is often cheaper than an optimal architecture. Ask what a topology change costs: different magnetics, a new layout, a new pre-scan and usually a new enclosure, all of which is a new design. Reusing the existing platform is often the cheaper answer.
Next step
Send the output range, the volume target and the efficiency requirement to Wecent Group through the contact page. The reply names the architecture for the envelope, the efficiency at the required load points and the pre-scan margin behind the choice. Related reading: GaN charger applications, OEM and ODM charger manufacturing and charger quality control. External references: USB Power Delivery specification, IEC standards, IEC Webstore, EU energy-efficient products, UL standards and certification and Intertek testing services.
{“@context”:”https://schema.org”,”@type”:”FAQPage”,”mainEntity”:[{“@type”:”Question”,”name”:”Is a resonant architecture always the best choice at higher power?”,”acceptedAnswer”:{“@type”:”Answer”,”text”:”It is frequently the efficient answer where the volume is tight and the efficiency target is demanding. Where the enclosure has room and the target is modest, a simpler architecture can meet the same specification at lower cost.”}},{“@type”:”Question”,”name”:”How much does the switching technology change the boundary?”,”acceptedAnswer”:{“@type”:”Answer”,”text”:”It moves it upwards: better switching devices reduce the losses that made higher frequencies and resonant designs necessary in the first place, which extends the range where simpler architectures remain viable.”}},{“@type”:”Question”,”name”:”Does the topology affect certification?”,”acceptedAnswer”:{“@type”:”Answer”,”text”:”It affects the electromagnetic profile and the thermal behaviour, and therefore the evidence that has to be established. The construction requirements follow the product and the market rather than the architecture.”}},{“@type”:”Question”,”name”:”Can a platform change topology later?”,”acceptedAnswer”:{“@type”:”Answer”,”text”:”That is a new design rather than an update: different magnetics, different layout, a new pre-scan and usually a new enclosure. Reusing the existing platform is often cheaper than an optimal architecture.”}}]}