Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

DC-DC computer power supplies

Status
Not open for further replies.

Llamarama

Member
Hello everyone, just something that's been bugging me for quite a while now;

How do the tiny dc-dc computer power supplies that usually come with tiny cases actually work whilst being so small?

It's astounding that a board nopt much larger than a standard atx connector (in the case of a Pico PSU) can deliver enough current to drive a board and drives and the likes.

I've built a small atx style supply to run one such board and still can't figure out how they use such small inductors. I understand they're switching supplies, and know roughly how they work, but without a schematic to work from i'm stumped.

Has anyone got any schematics or even block diagrams on how these devices work to deliver not inconsiderable current whilst being small enough to fit in the tightest of spaces?

Any info would be greatly appreciated, i'd quite like to learn how it's done so I can build my own rather than waste a day waiting for the delivery guy to deliver :) Many thanks
 
Off line switching supplies run 1000x faster so the transformer can be much much smaller. 60hz compared to 60khz.
This Pico PSU supply probably runs at 2mhz. Because the input voltage is 12 volts it can run much faster than 110/220 VAC.

Transformers can be rated in volt*seconds. 220V*(1/60hz) is a big number compared to 12/(1/2,000,000)
At 2mhz inductors only have to hold energy for a very short time. Capacitors only need to store energy for a very short time.
 
Last edited:
By increasing the switching frequency, you can use smaller inductors and capacitors.

That Pico PSU, although claiming 120W, requires a 12V input from a 120W psu to run. This likely means that the 12V input is connected directly to the 12V output (thus not requiring any conversion or components), and the converter then only needs to generate the 5V/3V3, and the low power -5V/-12V. This means that the actual converter doesn't convert 120W of power, but some fraction (perhaps half at most) of it.

Some ATX power info here https://www.electro-tech-online.com...eloper5Cspecs5CATX12V_PSDG_2_2_public_br2.pdf
 
This likely means that the 12V input is connected directly to the 12V output (thus not requiring any conversion or components), and the converter then only needs to generate the 5V/3V3, and the low power -5V/-12V. This means that the actual converter doesn't convert 120W of power, but some fraction (perhaps half at most) of it.

Maybe but I disagree about the 12V passing through. The 12V is rated at 8A, 15A pk. The input is rated at 10.5 to 13.5V.

There are many different versions of little supplies like this.
 
Maybe but I disagree about the 12V passing through. The 12V is rated at 8A, 15A pk. The input is rated at 10.5 to 13.5V.

There are many different versions of little supplies like this.
Kindly read the manual you have attached.
 
I see it now. "picoPSU-160XT has a high performance 4mR Mosfet Switch for the 12V rail,
capable of handling large currents with peak currents up to 15A."
 
In answer to the original post, have a look at Power Integrations. They do a lot of the mains power supplies used in this stuff and have lots of reference designs.
 
So they get around the large inductors by using very high frequencys, how do they draw such high currents at the same time? Does the high current flow through the inductors or just the mosfets?

So far this thread has helped clear up many mysteries to me, thanks very much :)
 
The converters would be buck converters and the current has to pass through both the FET and the inductor. An example of the basic components in such a converter is shown here:

View attachment 66644

Of course, a controller is required to turn the FET on and off to regulate the output voltage.

The high efficiency of the picoPSU converter is possible through the use of a synchronous rectifier (ie the diode shown in the attached image would actually be another FET, and just turned on at the appropriate time). Using a FET in this manner avoids the voltage drop of a diode and the associated power loss.
 
Power Integrations mostly works with off line switchers. 50khz to 120khz.

The 12V to 5V and 3.3V is a DC to DC switcher and runs at xmhz. Switching frequency is related to the speed of the mosfets. The mosfet can switch at 2000V in a uS. (just made up a number) A off line supply (110/220 supply) will switch 400 to 500 volts. A 12V DC to DC supply will only have 15 to 20 volts on the MOSFETs. In some configurations only 12 volts. It takes much less time to 'turn on' 12 volts compared to 500 volts.

The current runs through the mosfets and inductors.
 
So the current can safely be passed through the inductors as they are only on for a fraction of a second? Does this also apply to the high speed rectifier diode, or does that also have to be a high current component?
 
So the current can safely be passed through the inductors as they are only on for a fraction of a second? Does this also apply to the high speed rectifier diode, or does that also have to be a high current component?

With a buck converter, the inductor passes the current continuously for any decent load. There is not necessarily any off time. The rectifier has to be capable of carrying the output current. As mentioned, the 'diode' is actually a FET in the picoPSU to increase the efficiency.
 
Ahh, right. So how are the inductors in a picoPSU so small? They can't be much more than an inch square, so how do they cope with such a current? Are they just large low value inductors that have 2 or 3 turns of high gauge wire?
 
Here's an example inductor: https://au.element14.com/vishay-dale/ihlp6767gzer220m01/inductor-22uh-23a-20/dp/1741422

It has a DC max current of 11A (limited due to the heat generated by the winding resistance) and 23A saturation current (limited due to the number of turns as well as the type and quantity of the core material). It's 17mm square.

Regarding size reduction as frequency increases... Cores are (or at least can be) rated in terms of their amp-turns, which is simply the allowable current multiplied by the number of turns. For lower inductances, there are less turns and therefore more current is allowed for a given core size. Amp-turns relates directly to the magnetic flux through the core.
 
Last edited:
That makes a lot more sense now, all this time i've been using low frequency power inductors. This has certainly helped me alot, thank you very much :) Do they use an off the shelf PWM ic, or is it usually a propriatory microcontroller?
 
It's an off the shelf synchronous controller IC for the 5 and 3.3V lines. The same IC handles both rails.

**broken link removed**
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top