08-02-2017, 02:04 PM
2 power supplies.
The first is a boost converter. This sits between the bridge and the big smoothing capacitor, and is a switch and diode (on heatsink) and inductor. It steps up essentially unsmoothed 100Hz from the bridge to a nominal 380V or so. But because this is active PFC (rather than a straightforward boost converter), the control circuit varies the duty cycle to draw a current from the bridge that is approximately proportional to the voltage from the bridge. In other words, it tries to look like a vaguely resistive load to the mains.
General safety note: This boost converter does not offer galvanic isolation.
Point of interest: The boost converter can be bypassed completely - it sits between the existing bridge and smoothing capacitor in a standard SMPSU. Useful to know for fault-finding occasionally, but more interesting to reflect that active PFC is very easy for a manufacturer to retro-fit to existing designs. On many PC power supplies a while back, if you had it, you'd have a sticker on the outside proclaiming as such, and a small daughter board added inside for just that purpose. Just a MOS-FET and diode on a small heat sink, an 8 pin controller chip (e.g. the MC34262), and usually a half-decent 1µF poly cap on the input.
The next stage is the conventional step-down (and isolating) SMPSU. Often a flyback type; might be half-bridge. It looks like there's only 1 device on the appropriate heat sink (second from left), so I'd suspect the former. The big transformer and feedback optocouplers (and EMC cap) crossing the isolation barrier is clearly visible.
It does look like a well engineered PSU. Some Dell stuff is very good. I speak very highly of my newer U2412 (LED backlight, IPS screen).
The first is a boost converter. This sits between the bridge and the big smoothing capacitor, and is a switch and diode (on heatsink) and inductor. It steps up essentially unsmoothed 100Hz from the bridge to a nominal 380V or so. But because this is active PFC (rather than a straightforward boost converter), the control circuit varies the duty cycle to draw a current from the bridge that is approximately proportional to the voltage from the bridge. In other words, it tries to look like a vaguely resistive load to the mains.
General safety note: This boost converter does not offer galvanic isolation.
Point of interest: The boost converter can be bypassed completely - it sits between the existing bridge and smoothing capacitor in a standard SMPSU. Useful to know for fault-finding occasionally, but more interesting to reflect that active PFC is very easy for a manufacturer to retro-fit to existing designs. On many PC power supplies a while back, if you had it, you'd have a sticker on the outside proclaiming as such, and a small daughter board added inside for just that purpose. Just a MOS-FET and diode on a small heat sink, an 8 pin controller chip (e.g. the MC34262), and usually a half-decent 1µF poly cap on the input.
The next stage is the conventional step-down (and isolating) SMPSU. Often a flyback type; might be half-bridge. It looks like there's only 1 device on the appropriate heat sink (second from left), so I'd suspect the former. The big transformer and feedback optocouplers (and EMC cap) crossing the isolation barrier is clearly visible.
It does look like a well engineered PSU. Some Dell stuff is very good. I speak very highly of my newer U2412 (LED backlight, IPS screen).
