the gutless, bloated, and fried power supply hall of shame

[Wester547]

Re: the gutless, bloated, and fried power supply hall of shame

Well, I was asking since in single/double forward converter topology (with exception to APFC PSUs and valley-fill circuit PSUs, both of which use full-wave rectification), because they use half-wave rectification (as has been stated before) that means that one of the two diodes (in a schottky package of sorts) acts as a freewheeling diode in order to preclude the output voltage from spiking negative (so as to discharge the coils/inductor), which means it only takes a 3rd of the load (only one diode, the first one, has half of the load) and that the overall rating of the rectifier's current is derated by maybe 40% by comparison to full-wave rectification (half-bridge, push-pull, full-bridge). I was wondering if schottky barrier rectifiers, on their datasheets, are in fact rated through half-wave rectification (50/60Hz) or full-wave (100/120Hz), or even square-wave (20KHz), as stated before me.

[tom66]

Re: the gutless, bloated, and fried power supply hall of shame

The current rating is determined by four main factors:
- Current density (excessive current density damages the device, even if transient, and this also determines the maximum permitted surge current)
- Thermal limits (how hot the die gets when current passes through it due to the voltage drop.) Bigger heatsink not necessarily any better -- an infinite heatsink cannot dissipate infinite power with a bolted semiconductor!
- Reverse charge limits (how much power is dissipated by the reversing action during each conduction cycle)
- Package limits (lead size etc.) Usually unimportant.

Usually the surge rating is much larger than the continuous rating. I'd say that as long as you can maintain the power dissipation, you can conduct any current up to the surge current, at any duty cycle. The maximum average current you can conduct will be determined by the thermal limits, not the current density.

[Behemot]

Re: the gutless, bloated, and fried power supply hall of shame

wester: By half-vawe and full-wave in regards of secondary rectification, you mean that in half-wave, only in one half of cycle there is power being transfered, and by full-vawe, there is power in whole cycle (=in both halfs) power being transfered?

tom66: as for the surge, I don't think so. There is usually notice the surge is to comply with some JEDEC standard and though I haven't looked for it yet, I think there is stated how long the surge can be and how often it can happen. Because usually the surge is like five times higher than "peak average rectied current" or "peak repetitive current".

[tom66]

Re: the gutless, bloated, and fried power supply hall of shame

I am doing a semiconductor course, so am learning a few things about this. But basically current density limits the surge current. You can't have 5x more surge even if it lasts a fifth of the time... it destroys the device.

[Behemot]

Re: the gutless, bloated, and fried power supply hall of shame

Look at pretty much ANY shottky barier rectifier and you will found out you can.

For example, here the comon MBR2045CT:

Maximum average forward rectified current
at TC=135°C I(AV) 20.0 Amps

Peak repetitive forward current per leg at TC=135°C
(rated VR, sq. wave 2.0 KHZ) IFRM 20.0 Amps

Peak forward surge current 8.3ms single half sine-
wave superimposed on rated load (JEDEC Method) IFSM 150.0 Amps

[Wester547]

Re: the gutless, bloated, and fried power supply hall of shame

Thermal limits (how hot the die gets when current passes through it due to the voltage drop.) Bigger heatsink not necessarily any better -- an infinite heatsink cannot dissipate infinite power with a bolted semiconductor!
- Reverse charge limits (how much power is dissipated by the reversing action during each conduction cycle)
- Package limits (lead size etc.) Usually unimportant.

Well, I understand how all that affects schottkys in any topology, I was asking because in half-wave rectification, one of the two diodes in a schottky package is used as a freewheeling diode so to prevent the output voltage from going too high which means that diode can only do 1/3rd of its full load, that is why I ask. I think TO-3P/TO-247 packages have an advantage because they cover more surface area of the heatsink and that is quite an advantage, though when you see some TO-220 parts rated at higher temperatures your point definitely isn't moot. I do think thicker heatsinks matter: it means less stress for the rectifiers which in turn means less stress for the capacitors, and it also means a cooler running PSU. Where that will make the biggest difference though is how well the PSU will hold up if the fan seizes. I agree though that in an ideal design, the lion share of your cooling will come from the airflow of the fan.

Usually the surge rating is much larger than the continuous rating. I'd say that as long as you can maintain the power dissipation, you can conduct any current up to the surge current, at any duty cycle. The maximum average current you can conduct will be determined by the thermal limits, not the current density.

Well, I'm not sure about that... surge currents can only be sustained for so long without damage. On the datasheets they always list the ratings as absolute maximums continuously. I think it would be way too much to ask even twice the rating of a 30A rectifier continuously, for an example, even in full wave rectification, with the best design. The rectifier would probably burn, short out, and die a silent death.

wester: By half-vawe and full-wave in regards of secondary rectification, you mean that in half-wave, only in one half of cycle there is power being transfered, and by full-vawe, there is power in whole cycle (=in both halfs) power being transfered?

Well, power conversion is another difference between full wave and half wave.... I have read that the power conversion quality is much poorer in half wave but most papers do not take into the consideration the real world implications of switch mode power supplies or even linear regulators, such as voltage drops, duty cycle variations, the fan, heatsinks, and thermal compounds, the input and output filtering, the greater stress of higher ripple frequencies (even with a much lower turns ratio), etc.

[Behemot]

Re: the gutless, bloated, and fried power supply hall of shame

In forward converter the second diode also completes circuit, where the infuctor there behaves as power source which discharges itself into the load. The power goes back throguh ground and the second diode. In one half, it blocks negative, in other, it conducts as a part of cirtcuit.

Here is how it looks like:

[Wester547]

Re: the gutless, bloated, and fried power supply hall of shame

I don't want to derail other threads so I might as well ask here:

In forward converter topology, in a SMPS, where does half-wave rectification start taking effect? At the output (the secondary), after the primary section? After the voltage doubler (with the main switchers) (in other words, which components are affected by it? ^^; )?

[Behemot]

Re: the gutless, bloated, and fried power supply hall of shame

Secondary, of course. There's a bridge rectifier on output, but before the filtering caps / doubler / PFC. Than switchers chop the current into pulses which you have to rectify back to DC after secondary side of the transformer.

[Wester547]

Re: the gutless, bloated, and fried power supply hall of shame

One last question...

Is it only the secondary side of the main transformer that half-wave rectification takes place after in forward topology, in a SMPS, or is it the smaller transformer, the flyback, auxiliary transformer for +5VSB, that half-wave rectification takes place after too (does it affect linear regulation/the flyback transformer since it's separate from the main one?)?

[Behemot]

Re: the gutless, bloated, and fried power supply hall of shame

Flyback usually uses optocouplers. Because of low power, there usually is half-wave on power supply of driving ICs, and because of price, there is usually half-wave on +5 V SB and -12 V.

On better supplies you can find half-wave but for two windings, using rectifier instead of single diode.

[ben7]

Re: the gutless, bloated, and fried power supply hall of shame

Howcome there are no posts of crappy psus in here lately?

[everell]

Re: the gutless, bloated, and fried power supply hall of shame

They all went to the recycling center



I haven't seen a Bestec ATX-250 12E in a couple of years now. RIP!

[ben7]

Re: the gutless, bloated, and fried power supply hall of shame

They all went to the recycling center



I haven't seen a Bestec ATX-250 12E in a couple of years now. RIP!

Time to find different ones to install the 5vsb mod in!

[everell]

Re: the gutless, bloated, and fried power supply hall of shame

Time to find different ones to install the 5vsb mod in!

But when the Bestec ATX-250 12E blows up, it does it in style!

[Gabriel]

Re: the gutless, bloated, and fried power supply hall of shame

These pictures come from a Norwegian forum, this is a AOpen psu - Fuhjyyu and Yec caps.

[Behemot]

Re: the gutless, bloated, and fried power supply hall of shame

Is it FSP design or something else? AOpen seems to actually be a daughter company under FSP Group so it should be theirs, but hey may have had outsourced this one…

[Pentium4]

Re: the gutless, bloated, and fried power supply hall of shame

Yeah, looks like an old FSP one from 2000 (judging by that Teapo cap) I had one like that and the caps died in it too cause they're all trash worth a recap though, although not a very efficient PSU at all

[Behemot]

Re: the gutless, bloated, and fried power supply hall of shame

It's very old design so you can't look for efficiency in that thing. But it can deliver it's rated power.

[Gabriel]

Re: the gutless, bloated, and fried power supply hall of shame

I don't have any more info on the AOpen psu other then that the resistor on the left gets really warm - you can see it has melted the shrink tube.

The psu doesen't belong to me, just happen to see the pictures on a forum yesterday - the owner did not post any specs on it.