Hardware Insights Forum

I know this is rather far back but I have a question about something regarding the review, to some extent, that maybe someone could answer here...

I noticed with that generation of 300W PSUs from Hipro, Lite-on, and Bestec, they list the +5V and +12V combined rating instead of +5V&+3.3V&+12V... unless the former is just short for all those rails combined? Is the +12V rail limited by the amount of power on the +5V rail (and vice versa) since they share the same toroidal coil in a group-regulated design or is there something else involved? Is it the same thing as, in very old PSUs, the amount of power on the -5V and -12V rails combined (since they also share the same toroid)?

The amount on the 12V rail will be somewhat limited by what is being drawn from the 5V rail, since they do share the same toroid coil. The coil can only handle a given amount of power being drawn through it without overheating and burning up. The 12V + 5V rating on these PSUs is simply what the toroid coil can handle. The 3.3V rail isn't counted as it has its own coil. Some manufacturers don't state this limit and simply state what the main transformer can handle (the 12v + 5v + 3.3v limit)

278.6 / 0.66 = 422.1

You must have confused input with output power when you said 309.1W.

Also, could you fix the old forum link???

^
I think you're getting the power factor confused with the efficiency. 278.6W (Power draw) x 73.22% (efficiency) = 203.99W, rounded to 204W.

I have fixed the link

You explicitly stated:

So if your system draws 204W, the PSU is really drawing 278.6W from the wall, but the power grid will think your pulling 309.1W.

Ah, yes, I thought you were referring to the load table. Fixed.

...we have a pair of 16A parts on the 12v.

Might be worth mentioning that those are only "ultra-fast", not Schottky.

The original caps were a mixture of four 2200uF 16v Teapo on the 12v rail, a pair of 2200uF 10v Teapo caps on the 3.3 and 5v rails and a 1000uF (which was bulged) on the 5vsb rail.

• Of the 2200µF 16V caps, two are for the +12V and two for the -12V (as mentioned in the poly-mod article).
• The +5VSB has two caps, a 1000µF and a 470µF (also mentioned in the poly-mod article).

The bridge (not mentioned), a GBU806, is rated for 8A at 100°C with a heatsink. But rectifiers can only deliver 80% of their nominal rating into a highly capacitive load (e.g. no-PFC and passive PFC PSUs) - the nominal rating assumes a resistive or inductive load.

LongRunner wrote:The bridge (not mentioned), a GBU806, is rated for 8A at 100°C with a heatsink. But rectifiers can only deliver 80% of their nominal rating into a highly capacitive load (e.g. no-PFC and passive PFC PSUs) - the nominal rating assumes a resistive or inductive load.

Do those capactive loads apply for both the primary and secondary rectifiers? Just to be clear, as to my understanding the secondary side works a bit differently. And are Active PFC PSUs also highly capactive loads for the bridge rectifier?

EDIT: Also, this is a very interesting article on the matter: http://sound.westhost.com/lamps/pfc.html

Wester547 wrote:Do those capactive loads apply for both the primary and secondary rectifiers? Just to be clear, as to my understanding the secondary side works a bit differently.

Not as far as I know. The two operate under very different conditions, and the half-bridge and forward topologies have a large inductor between the rectifiers and capacitors anyway.

And are Active PFC PSUs also highly capactive loads for the bridge rectifier?

Not if working correctly. (In standby mode, the PFC is inactive, but then the rectifier is operating far below capacity anyway.)

By the way, I already knew the difference between connecting capacitance directly to AC, and feeding it through a diode bridge.

Each phase to neutral gives 230V RMS, and between phases is 398V ( 230V * √3 ).

Actually, nominal phase-to-phase under that system is exactly 400V, and the "230V" in question is rounded down from 230.94010767585030580366. I can't think of anything else that would explain such a weird number.

The exact peak-to-peak voltage from phase-to-neutral under that system is, therefore, 326.59863237109041309297.

Of course, there's no practical value in being that precise. Just mathematics.

By the way, what that article refers to as "2-phase" is actually called split-phase. 2-phase is actually the name of an obsolete system used to power motors, which had the two phases 90° apart instead of being mirrors (which wouldn't aid motor starting).

2-phase motors themselves technically still exist, only now they're called "split-phase" (this gets better and better!) and powered from single-phase mains with a capacitor in series with one winding to provide phase shifting. (In a few applications, such as ceiling fans, the capacitor can be switched between the windings to change the rotation direction.)