Switch mode power supplies and protection class.

[Selfmade]

My understanding is that a SMPS does not separate, like a SELV transformer, the secondary from the primary. Therefore presumably the equipment supplied cannot be IEC protection class III. So what class is the equipment? For example a laptop with an SMPS, metal body and high resistance functional earth? Or a mobile phone with metal body when charging? These devices are surely too small to fit double insulation, but do not appear to have protective earths.

You'd think it would be a simple question, but I've trawled the internet for hours and not found an answer. I want to make it clear that I'm talking about the protection class of the equipment supplied here, not the power supply itself (assuming each have separate enclosures and are connected together by a flex). I've found lots of sources talking about the protection class of the power supply being either class I or II, but these don't answer my question.

Interested mainly out of curiosity as I don't currently work on this type of equipment.

 

[Lucien Nunes]

My understanding is that a SMPS does not separate, like a SELV transformer, the secondary from the primary.

They can, and usually do, provide isolation. An SMPSU has electronics on both the primary and secondary sides, but the two are isolated from one another by one or more transformers (to get power from primary to secondary) optocouplers (to get control feedback from secondary to primary) and low-value class Y capacitors (to bypass RF)

Not all PSUs are designed to (or need to) provide an equivalent degree of safety to a double-wound transformer to BS EN61558-2, for example the units built into other circuitry within a Class I device. But that is not a limitation of the SMPSU in principle and all stand-alone units such as phone chargers should do so.

One possible reason that the answer is not very clear is that the new standard for energy sources, BS EN62368-1 etc does not refer to SELV etc, but to the energy source classes (different from protection classes) and it is these definitions that indicate suitability for use with class III loads.

 

[Selfmade]

They can, and usually do, provide isolation. An SMPSU has electronics on both the primary and secondary sides, but the two are isolated from one another by one or more transformers (to get power from primary to secondary) optocouplers (to get control feedback from secondary to primary) and low-value class Y capacitors (to bypass RF)

Not all PSUs are designed to (or need to) provide an equivalent degree of safety to a double-wound transformer to BS EN61558-2, for example the units built into other circuitry within a Class I device. But that is not a limitation of the SMPSU in principle and all stand-alone units such as phone chargers should do so.

One possible reason that the answer is not very clear is that the new standard for energy sources, BS EN62368-1 etc does not refer to SELV etc, but to the energy source classes (different from protection classes) and it is these definitions that indicate suitability for use with class III loads.

Interesting, thanks. It was one of JW's videos about SPDs where he stated that a key reason for needing SPDs is that modern electronics don't have transformers and so the entire devices is exposed to any surges. He described how an SMPS works and it didn't seem to me to offer separation. But perhaps I misunderstood.

 

[Lucien Nunes]

and so the entire devices is exposed to any surges

Not the entire device, but the electronics on the primary side of the SMPSU, which actually constitute the bulk of the innards of the PSU itself. Often with a small unit such as a charger, the only electronics on the secondary side are the rectifier and smoother for the output, and the feedback circuit to tell the control chip on the primary side about the load current and voltage it is producing.

The power electronics on the primary side, being the main chopper transistors that turn the bus DC into a drive waveform for the transformer, and the boost converter (if the unit has active PFC) are highly stressed and vulnerable to damage by overvoltage. These are the parts that will be saved by SPD. A really serious spike such as a nearby lighting strike can cause flashover to the secondary side, especially via capacitors that bypass the isolation barrier. However the same possibility exists for wire wound isolation transformers too.

Historically it was more true that some entire electronic devices were connected directly to the mains. Many 1960s/70s colour TVs for example, used rectified and chopped mains to produce the DC voltages for the set directly. Every part of the circuit was dangerous live and the only external connection - the aerial - had to pass through class Y or similar isolating capactors. As SMPSU technology developed, and as there was an increasing need to connect VCRs, home computers etc via SCART and other interfaces, TV makers moved to isolated power supplies.

Find an old power brick for a laptop you mo longer have, or some other reputable quality electronic device with an SMPSU, open it up and study the PCB. You will find a clear physical demarcation of the isolation barrier, straddled only by the transformer(s), optocoupler(s) and Y-cap(s).

 

[Selfmade]

Not the entire device, but the electronics on the primary side of the SMPSU, which actually constitute the bulk of the innards of the PSU itself. Often with a small unit such as a charger, the only electronics on the secondary side are the rectifier and smoother for the output, and the feedback circuit to tell the control chip on the primary side about the load current and voltage it is producing.

The power electronics on the primary side, being the main chopper transistors that turn the bus DC into a drive waveform for the transformer, and the boost converter (if the unit has active PFC) are highly stressed and vulnerable to damage by overvoltage. These are the parts that will be saved by SPD. A really serious spike such as a nearby lighting strike can cause flashover to the secondary side, especially via capacitors that bypass the isolation barrier. However the same possibility exists for wire wound isolation transformers too.

Historically it was more true that some entire electronic devices were connected directly to the mains. Many 1960s/70s colour TVs for example, used rectified and chopped mains to produce the DC voltages for the set directly. Every part of the circuit was dangerous live and the only external connection - the aerial - had to pass through class Y or similar isolating capactors. As SMPSU technology developed, and as there was an increasing need to connect VCRs, home computers etc via SCART and other interfaces, TV makers moved to isolated power supplies.

Find an old power brick for a laptop you mo longer have, or some other reputable quality electronic device with an SMPSU, open it up and study the PCB. You will find a clear physical demarcation of the isolation barrier, straddled only by the transformer(s), optocoupler(s) and Y-cap(s).

Thanks very much for this detailed and fascinating explanation. Really good to get the historical context too, thanks.