Power supply build quality pictorial. part 2

[cheapie]

Re: Power supply build quality pictorial. part 2

i'm pretty sure it has dual 12V rails instead of a beefy single rail.

Unlike this thing - 1300 Watts with one(!) 108A 12V rail. I thought ATX12V v2 required at least two 12V rails...

[mariushm]

Re: Power supply build quality pictorial. part 2

afaik the latest atx standards have removed that restriction, no amps limit on rails.

[Wester547]

Re: Power supply build quality pictorial. part 2

and the primary caps will improve the holdup time.

Does this mean that how fast the main switchers (whether bipolar or FETs) can switch is ultimately limited by the holdup time of the voltage doubler or primary capacitor(s)? I ask because I noticed that holdup time in PSU reviews is measured in milliseconds (thousandths of a second) whereas bipolar transistors' switching time is measured in microseconds (millionths of a second) and MOSFETs in nanoseconds (billionths of a second), in their datasheets.

[Behemot]

Re: Power supply build quality pictorial. part 2

No. For the beginning, take it as if the caps cant handle, there is nothing to switch, no power.

In reality it works but some nasty effects can occur.

[PeteS in CA]

Re: Power supply build quality pictorial. part 2

Wester, hold-up time refers to how long a P/S will remain in regulation after AC input power is shut off or lost. The input capacitors and the load determine that. The voltage across the input capacitors decreases, while the switch duty cycle increases. When duty cycle reaches maximum the outputs will fall out of regulation. So the larger the capacitance of the input capacitors the longer it takes for the voltage to fall to where duty cycle is maximum. 16 or 20 ms is a couple blinks of your eye, but it's a very large number of processor cycles.

[cheapie]

Re: Power supply build quality pictorial. part 2

I used to have a PSU that could hold up for almost a second...

[Shocker]

Re: Power supply build quality pictorial. part 2

Under what load???

[ben7]

Re: Power supply build quality pictorial. part 2

Under what load???

I was thinking the same thing. All the power supplies I have hold up for like 3-5 seconds, no load that is!

[cheapie]

Re: Power supply build quality pictorial. part 2

Under what load???

About 150W. It was a 700W PSU.

Oh, and my mom's laptop charger can hold up for over a minute with no load.

[PeteS in CA]

Re: Power supply build quality pictorial. part 2

The 16-20mS number I refenced above assumes full rated load. With less than that hold-up time increaes, as the I/P lytics are discharged less quickly. With no load, what discharges the I/P lytics is keeping the P/S supply running plus the discharge resistors across those lytics.

[tleu8472]

Re: Power supply build quality pictorial. part 2

Is this Aywun A1-3000 unit good for 250-300W? Seems to have no MOV, but otherwise looks OK?

[mariushm]

Re: Power supply build quality pictorial. part 2

I wouldn't use them on a system that needs more than 200 watts.

The transformer is probably ERL-33, not 35, but otherwise it has inductors on the secondary side, heatsinks look reasonable, the rectifier diodes look capable enough, it doesn't look that bad.

[DJduck]

Re: Power supply build quality pictorial. part 2

4 diodes? Thats pretty crappy.

[mariushm]

Re: Power supply build quality pictorial. part 2

They look like 2-4A diodes. Two of those conduct at the same time, so they share the current.

At 100-200w average usage, that would be 200 / 110v = ~ 2A, so each diode has about 1-1.5a max on it. They'll be fine for the usage, especially since they're kept cool by the fan.

Bridge rectifiers are better, but without a heatsink they're not MUCH better, they get derated with heat (a 6A bridge rectifier for example is derated to 2A without heatsink) but depends from model to model.

[Pentium4]

Re: Power supply build quality pictorial. part 2

Bridge rectifiers are better, but without a heatsink they're not MUCH better, they get derated with heat (a 6A bridge rectifier for example is derated to 2A without heatsink) but depends from model to model.

Are they really derated that much? That seems pretty low, they seem to run fine in PSU's with a 300-350W load with a 6A bridge no heatsink, at below 80% efficiency. But they're usually close to the fan

[mariushm]

Re: Power supply build quality pictorial. part 2

Have a look at this:

http://uk.farnell.com/fairchild-semi...gbu/dp/1700185

FAIRCHILD SEMICONDUCTOR - GBU4G - BRIDGE RECTIFIER, 4A, 400V, GBU

Datasheet from that page:

And you have the graphs on the second page:



With heatsink, it will do 4A up to 100c body temperature.

Without heatsink, dissipating only through package and leads to the pcb board, it can do 3A as long as it stays under 40c.
It gradually drops down to 2A at about 75c and down to 1.5A at around 100c

Note thought that this assumes no forced convection (air flow from a fan over the rectifier, just air naturally moving around). On a fan cooled power supply, the bridge rectifier probably stays at around 45-50c or less as the air from the fan cools the rectifier a bit.

But it also depends on how much current goes through the rectifier, if there's little current it won't heat as much.

[Pentium4]

Re: Power supply build quality pictorial. part 2

Oh wow, yeah it drops pretty quickly. Thanks for posting that info. Even in lower end PSU's i try to at least have a 6A bridge

[Wester547]

Re: Power supply build quality pictorial. part 2

The "without a heatsink" measurements in those datasheets are with no fans at all and at 25*C ambient... so that's totally convection cooling, as mariushm said... I wouldn't be surprised if the bridge rectifier was at 100*C at that point (since they can operate up to 125-150*C). Bridge rectifiers are likely not to get so hot being so close to the fan so I think they can do at least 75% of their rating even without a heatsink, so long as they receive decent airflow. Note that with a heatsink and fan within close proximity a bridge rectifier may be able to do slightly more than the rating. It does also depend on efficiency though, that might derate them further.

[tleu8472]

Re: Power supply build quality pictorial. part 2

Anyone know if the Delta Greenpower M300W is any good?

[DJduck]

Re: Power supply build quality pictorial. part 2

Probably good, Delta makes really high end stuff.