Dell H305P-01 Review

[Wester547]

It's not for the -12V - this unit has a 7912 for that, and the resistor isn't anywhere near the 7912 anyway. It's for the +12V.

How is it for the +12V? It's not near the -12V regulator but it is around the output capacitors for it. If it isn't for -12V, it's for +5V since the +5V rail is in between the -12V and +12V rails in that PSU.

It's the other way around - the FYP1010DN handles the pulses from the transformer and the STPS20S100CT does freewheeling. With a typical (?) duty cycle of 33% (one third), it would indeed give 30A on the +12V.

So they're both in parallel, essentially, with different duties?

Actually, forward doesn't work that way. It's the isolated counterpart of the "buck" topology used in motherboard VRMs along with the +12V to +3.3V and +5V converters in current high-end PSUs. The mag-amp in forward topology does the same thing as the coils in a motherboard VRM, albeit at a higher power level and lower frequency. Basically all you're changing from buck topology is inserting a transformer and (by necessity) another diode.

I would guess that in a forward converter outputting +12V, the pulses from the transformer would be somewhere from +24V to +40V. A higher duty cycle is better as it reduces conduction losses in the switcher(s) and the pulse voltage from the transformer (therefore allowing lower voltage diodes on the secondary).

A little confused.... what do you mean, forward doesn't work that way? In a half-wave rectified output (forward topology), doesn't one of the diodes in a two diode package have to be freewheeling and the other fully conducting, meaning one can only take 1/3rd of the load and the other 2/3rds of the load (derating the current ability since it has to take more of the load)? Just curious, how does the magamp circuit differ in forward topology by comparison to half bridge? Do you mean that the transformer in forward topology has transformer taps with voltages that are double that of the outputs or more?

Actually, because of how they're connected in this unit - which doesn't equally split the current between the two parts - it's 54A if the duty cycle at full load is a third...

I'm not agreeing with Hardware Secrets on this one - I think you can unevenly load the two diodes in a single part as long as the total dissipation of the two is within the safe limit, because they're on the same chip and will therefore stay at about the same temperature. But if you use multiple physical parts, you do need to be careful with the current split between them, as they can end up at very different temperatures. In this case, one diode out of the four is the pulse rectifier and the other three are freewheeling. With the one-third duty cycle:

(54*(1/3))+(18*(2/3)) = 30A average (the rectifier next to the transformer)
(18*(2/3))+(18*(2/3)) = 24A average (the rectifier on the other side of the heatsink, next to the mag-amp)
(18 being 54 / 3)

It's still 2.45x the label rating, though, and the one next to the transformer would have been more than enough all by itself.

Do you mean that in forward topology or even half bridge, two schottkys in parallel doesn't quite mean double the current, period? And are you trying to say that even if one of the two diodes is freewheeling and not fully conducting that the current ability really isn't derated, and that if it is derated, it's by no more than ~11%?

[LongRunner]

It's kinda hard to explain, but I'll try…

How is it for the +12V? It's not near the -12V regulator but it is around the output capacitors for it.

With a 7912 there's no need for more than a minuscule load on its output. I know you can't see where it is (it's not the one you can see in the secondary-side photo), but I can't point out its position without starting a new thread about it as ordinary members can't upload attachments in this area of the forum.

It's the other way around – the FYP1010DN handles the pulses from the transformer and the STPS20S100CT does freewheeling. With a typical (?) duty cycle of 33% (one third), it would indeed give 30A on the +12V.

So they're both in parallel, essentially, with different duties?

They're not in parallel. When one is conducting, the other is blocking.

...what do you mean, forward doesn't work that way?

Take a CPU VRM, which draws (for example) +12V and outputs +1.5V at a much higher current. The mag-amps together with the connected components similarly smooth high-ish voltage pulses into a lower constant voltage. Unfortunately, again it's impossible for me to post an illustration in this area of the forum.

In a half-wave rectified output (forward topology), doesn't one of the diodes in a two diode package have to be freewheeling and the other fully conducting, meaning one can only take 1/3rd of the load and the other 2/3rds of the load (derating the current ability since it has to take more of the load)?

In short, no, but it's too hard to fully explain without pictures.

Just curious, how does the magamp circuit differ in forward topology by comparison to half bridge?

There is probably little difference in the mag-amp itself.

Do you mean that the transformer in forward topology has transformer taps with voltages that are double that of the outputs?

That or more, in general. There's no law that says you can't have a duty cycle above half, though.
(LATE EDIT: You can only exceed 50% duty cycle safely, if using active clamps.)

Do you mean that in forward topology or even half bridge, two schottkys in parallel doesn't quite mean double the current, period?

In forward, it depends on how they're connected. The conduction duty cycle of the pulse rectifier is the same as the switching duty cycle and the duty cycle of the freewheeling rectifier is the inverse. I will start the thread when I'm ready. The only scenario in which you can confidently use the sum of their nominal ratings is if every dual-diode chip has one of its internal diodes used for the pulse rectifier and the other for freewheeling, which is not the case in this unit…
The bridge topologies on the other hand (whether traditional half-bridge, or resonant half-bridge or full-bridge) use the transformer and rectifiers symmetrically, so no sane PSU manufacturer would put an asymmetrical configuration as that would just uselessly "waste" the current capacity of the stronger side. So for those, you can indeed just multiply the current capacity of each part by the quantity.

[Wester547]

:huh: With a 7912 there's no need for more than a minuscule load on its output. I know you can't see where it is (it's not the one you can see in the secondary-side photo), but I can't point out its position without starting a new thread about it as ordinary members can't upload attachments in this area of the forum.

On that note, does a -12V and -5V rail need the freewheeling diode in forward topology too? I ask because I never see a second freewheeling diode in -12V and -5V rails except maybe in half bridge units for reasons you've mentioned before.

In short, no, but it's too hard to fully explain without pictures.

I thought one was conducting and the other was freewheeling (blocking and conducting, like you said)?

That or more, in general. There's no law that says you can't have a duty cycle above half, though.

But half bridge transformers do not have such high voltage transformer windings? Also, how is a -5V or -12V rail without a linear regulator (and for that matter with) in forward topology generated if the transformer windings are double the voltage or more? Also, since the magamp circuit in forward converter topology is used as something of a "buck converter", do the rectifiers before the magamp have a much higher voltage going through them than that of half bridge?

[LongRunner]

On that note, does a -12V and -5V rail need the freewheeling diode in forward topology too?

Usually forward converters use the mag-amp to generate −12V (with no winding on the power transformer*) from the +12V (and +5V if group regulated) rail(s), which indeed only requires one diode. (If you trace the circuit in those units, you'll find that one end of the −12V winding on the mag-amp leads to ground…) There is at least one exception though, the Macron MPT-xxx2 (the original MPT-xxx was half-bridge), which does have a transformer winding for −12V, along with two (unusually large) diodes (both the same size - most likely 3A), which I would guess makes the PSU more expensive but significantly reduces stress on the (220µF 25V) cap before the −12V regulator. (Yes, you do need a 25V cap before the −12V regulator, if you use one.)

I don't know why anyone would want to produce the −5V in a forward converter from a separate source - they'd take the easy way out, using a 7905 with its input connected to the −12V rail.

*Strictly speaking any inductive component with two or more windings sharing the same core is technically a transformer, but only those across the isolation barrier are usually referred to as such.

I ask because I never see a second freewheeling diode in -12V and -5V rails except maybe in half bridge units for reasons you've mentioned before.

In (all) bridge topologies, both diodes rectify opposing pulses from the transformer (which unlike in a forward converter is driven both ways, remember) and neither of them is called a "freewheeling" diode. But you're right in that only bridge topologies usually use two diodes for the −12V (and sometimes also for −5V but alternately they just use a 7905, as with most forward converters).

I thought one was conducting and the other was freewheeling (blocking and conducting, like you said)?

When the switcher is on, transferring energy through the transformer, what I refer to as the "pulse" rectifier is conducting, and the freewheeling rectifier is blocking. When the switcher turns off, the pulse rectifier blocks and the freewheeling rectifier conducts. Inductors in operation can best be compared to a spring in that the spring stores energy when you compress it, and when you release the pressure, it releases its energy by bouncing back. If there is little force acting against re-expansion of the spring (like a lightly loaded inductor), it will bounce back a large amount, quickly releasing its energy. If there is much force acting against its expansion (like a low circuit impedance), it will expand and release its energy less violently and over a longer period. It's better than anything else for explaining the inductive "kick-back" behaviour, without which these switching converters would never work.

But half bridge transformers do not have such high voltage transformer windings?

Actually they're twice as bad as forward converters (setting aside resonant bridges, which have no mag-amps at all). goodpsusearch (on Badcaps.net) has burned out an ESAD83-004 by putting it on a +12V rail in such a unit. If the voltage pulses from the transformer were not much higher than the output, I expect the rectifier would have been fine. It would also likely have been OK on the +12V rail of a forward converter. I actually think even 60PRV is marginal for the +12V in an old half-bridge design, and would recommend at least an 80PRV rectifier there.

Also, how is a -5V or -12V rail without a linear regulator in forward topology generated if the transformer windings are double the voltage or more?

With their own mag-amp windings.

Feel free to start a thread in the main PSU forum asking how this stuff works.

[Wester547]

Why does the 10A Fairchild part and 20A STMicroelectronics part form to make 30A in forward topology, in this PSU? Just curious... the 20A schottky is on the magamp side and the 10A rectifier on the transformer side so wouldn't one diode be freewheeling in the 10A rectifier, the other conducting, and both of the diodes freewheeling in the 20A part, making for 26A total (or does it work another way)?

[LongRunner]

...the 20A schottky is on the magamp side and the 10A rectifier on the transformer side so wouldn't one diode be freewheeling in the 10A rectifier, the other conducting, and both of the diodes freewheeling in the 20A part, making for 26A total (or does it work another way)?

The "1:3" arrangement is only for +5V. The entire FYP1010DN is for the "pulses" from the transformer and the entire STPS20S100CT for freewheeling. And it wouldn't be sensible design to connect the diodes from two different parts in parallel.

[Wester547]

If the STPS20S100CT is only freewheeling (both diodes in the package), it wouldn't be capable of 20A, though, right (more like 16A?)? Unless one is conducting and the other freewheeling.

[LongRunner]

If the STPS20S100CT is only freewheeling (both diodes in the package), it wouldn't be capable of 20A, though, right (more like 16A?)?

As far as I know, (in half-bridge or forward topology) they can't be subjected to more than the load current at any point in time, and neither of the two parts are continuously loaded - the FYP1010DN for the one-third or so of on time and the STPS20S100CT for the other two thirds or so. And hell, if an FR1003G can survive 19.2A (albeit with a full speed fan), I see no problem with 20A continuous through a "20A" Schottky rectifier (with adequate cooling). And while I don't like STMicroelectronics, it's not Hipro's fault if it fails within the rated limit.

Referring back to this thread, the circuit for +12V in this unit is like my example "A", with the FYP1010DN being D1 and the STPS20S100CT being D2.

(Note the statement "with identical parts" in the explanation. As the parts here are different, the optimal duty cycle is 33% in this case.)