How this ATX PSU works anyway?

[meliomelenios]

Hello,

i'm trying to decipher this schematic that i found for an ATX power supply:

https://www.smpspowersupply.com/ATX_power_supply_schematic.pdf

I have stuck on a capacitor... I'm talking about the capacitor C5, a 2.2uF capacitor in series with the power transformer. I have mark this capacitor in this screenshot:

Weird capacitor

This power transformer is the one that generates all the voltages. If C5 is in series with the transformer coil, how can this transformer generate the 12V 15A output??? And where is the switching transistor for this transformer?

What am i missing here?????

 

[MrAl]

Hi,

That capacitor is in series with the PRIMARY. That means it only has to handle maybe 1/10 the current. So with 15 amps out at 12 volts and a primary of 120v it might only be around 1.5 amps.

 

[ronsimpson]

Q2.Q4 are the switching transistors. They drive pin 6 of T1 from ground to HHV and back to ground.

C5 is high impedance at 60hz but low impedance at the switching frequency.

 

[Jaguarjoe]

If you have a copy or can borrow a copy, "Art of Electronics" totally tore apart a switcher and explained every everything about it.
If you look for a "Cambridge low price edition" on ebay or Amazon, it'll be pretty cheap. The book is pretty dated but the fundamentals in there will hardly change.

Art is a smart guy

 

[meliomelenios]

So, if 220 v is the supply, C3 and C4 perform a voltage divider, and C5 is just to limit the current through the primary. At 10KHz this capacitors has about 7 ohms impedance. This makes absolute sense.

Another ides: Suppose that we run the CPU at 110V. Then we bridge the input selector. The capacitor will not allow the 50Hz to pass since it will have 1.5KHz impedance. Could this be the real job of this capacitor????

 

[bountyhunter]

Hello,

i'm trying to decipher this schematic that i found for an ATX power supply:

https://www.smpspowersupply.com/ATX_power_supply_schematic.pdf

I have stuck on a capacitor... I'm talking about the capacitor C5, a 2.2uF capacitor in series with the power transformer. I have mark this capacitor in this screenshot

The configuration is an offline half-bridge, which requires the DC blocking cap in series with the transformer primary so there won't be any flux offset.

 

[bountyhunter]

Q2.Q4 are the switching transistors.

I believe they are Q3 and Q4.

 

[MrAl]

Hello again,


C5 is not there to limit current and it is not there to block any 50Hz as there is only ripple which the control circuit will help to reduce. It is there to block the DC current allowing only the AC from the switching transistors to reach the primary.

The two 680uf caps produce a dual plus and minus DC power supply, whether or not the 115v short is there or not. That DC is then chopped by the two transistor. When one transistor is on it provides a positive pulse to the lower terminal of the transformer primary, and when the other transistor is on it provides a negative pulse to the lower terminal. This way the lower terminal is getting a rectangular wave pulsing plus and minus, and the cap allows the other side of the transformer to be connected to the DC mid point voltage.
Without this cap the transformer would not get equal plus and minus voltages because of differences in the transistors. A difference in plus and minus voltages of even a little can cause large DC currents in the primary of the transformer which could quickly cause it to saturate. With a cap in series, there can be no DC current and so the primary does not saturate.

 

[meliomelenios]

Thank you for your replies people. Now i understand what this capacitor does... Thanks

On bottom left of the datasheet there is another switching topology, a flyback topology. It is a typical schematic with no series capacitor. So a question rises: What is the advantage of using the series capacitor? I understand that it blocks DC currents, so i suppose that this capacitor turns the specific inductor more efficient, right? I mean, with the same inductor we can get higher current at the secondary coils because it saturates at higher switching current, am i right?

 

[MrAl]

Hello again,


It's mostly about how the core is used. Some topologies only use 1/2 of the total flux capability of the core. What that means is that they can only use 1/2 of the voltage range too. Topologies that use the full flux capability of the core can use the full voltage range available for the number of turns on the core because the voltage goes plus and minus instead of just plus or just minus. If we had a 1:1 transformer that can only take plus 10 volts on the primary we could only get 10 volts out, but if that transformer can take plus and minus 10 volts on the primary then we can get plus and minus 10 volts out which is twice the voltage level using the same amount of steel for the core.

If a transformer is to be efficient we want low resistance in the windings so making the primary resistance low is good. But that also means that if we have any difference in plus and minus voltage swings that there will be what is usually called a "net DC current" in the primary. For example, if the resistance of the winding is 0.1 ohms and the voltage difference is only 1 volt, we have:
I=V/R
10=1/0.1
so we have 10 DC amps in the primary. That's totally unacceptable because the DC does not contribute to the power transfer at all.
To help the situation if a cap is inserted in series with the primary (topology must be right too) even a voltage difference of 2 volts doesnt matter anymore because the cap will average this out to the midpoint voltage. What this means is that the upper terminal of the transformer is at the exact center voltage between the plus and minus and this means no DC in the primary. So for a plus voltage of 11 volts and a minus voltage of -9v, the midpoint voltage is 1v. The cap will average the two voltages and come up with 1v and that will be applied to the upper terminal of the transformer with +11 or -9 applied to the bottom terminal, so the winding sees plus and minus 10 volts rather than plus 11 and minus 9 which could saturate the core and stop this power supply from working at all.

Each topology should be studied for it's advantages and disadvantages. There are good reasons for using different topologies for different purposes. These reasons may vary based on one or more of the components of the system and the desired end goal and cost of the system. Understanding the different topologies a little helps to understand why these differences came about. It's harder to understand the differences about something if you dont have any good information about it in the first place. If i told you i had a box of oranges and called them 'product A' and then told you to compare this to 'product B', how would you know how to compare them if you didnt know what product B was. Once you know what product B is you can then begin to compare the two boxes to see what the similarities and differences are. The point is if you've never seen product B it would be difficult to compare it to anything else.