SMPS design

Thread Starter

mahmoud shendy

Joined Dec 23, 2007
21
Dear All, really I want to know about Switching Mode Power Supply (SMPS) but I don't know anything about. Would you
guide me by recommended Sites, recommended free books, Schematics, tutorials....etc
again i don't know any thing about.. My target from this future study is to build my Hardware fed by 110V mains and
has no the HEAVY TRANSFORMER...
Thanx for all.. M.SHENDY
 

Papabravo

Joined Feb 24, 2006
21,159
Your goal is probably not a rational one. In addition to providing isolation and safety, a transformer is about to only way to avoid substantial losses that disappear as heat. There are several SMPS designs that go from one DC voltage to another. A common thread among all of these designs is that input power is ALWAYS strictly greater than output power. A typical design goal is to make this efficiency factor land in the 80% to 90% range. To have a successful design articulating the requirements is essential.
 

scubasteve_911

Joined Dec 27, 2007
1,203
I'm not sure how it is irrational to think you can bring your 120V mains down to DC voltages without using bulky transformers.. Sure, they do use transformers in a lot of different switchers like that, but they aren't nearly the size of the strictly linear transformers. They do their switching at a much higher frequency, so core material is an important consideration for functionality and losses.

Check out www.national.com , they have a ton of old appnotes. It is fairly challenging to bring the rectified 120Vac mains down to reasonable levels, but people do it all the time. Just be safe!!!! That's a lot of power to deal with, respect its ability to kill you.

Steve
 

mrmeval

Joined Jun 30, 2006
833
I have several sitting here now. One is out of a computer, one out of a PTV, one from a Cisco router. They use the smaller high frequency transformers which I'd not call heavy or bulky.

On the PTV one the transformer weighs slightly less than the switching control IC and it's heat sink.
 

Papabravo

Joined Feb 24, 2006
21,159
I have several sitting here now. One is out of a computer, one out of a PTV, one from a Cisco router. They use the smaller high frequency transformers which I'd not call heavy or bulky.

On the PTV one the transformer weighs slightly less than the switching control IC and it's heat sink.
Are you saying they are using a small high frequency transformer with the primary connected to the 110V 60 Hz. mains? That sounds like a prescription for disaster to me. You're not sounding very credible on this issue or I've missed something. I understand that in an SMPS design for DC to DC conversion you can pick the switching frequency. With the mains you have to take the 50/60 Hz. you're given you have no choice.
 

beenthere

Joined Apr 20, 2004
15,819
The switchers I've been in to all rectify the mains voltage and use a FET to pulse current through the input winding of a high frequency transformer just like a DC to DC converter.

That makes them lethally dangerous to poke around in while running, but the output/s are isolated from the line.
 

scubasteve_911

Joined Dec 27, 2007
1,203
Yes, That is the topology I was referring to in my previous post. "switching from mains" pretty much means you rectify the AC input and switch it down to an appropriate voltage. An incredibly low duty cycle would be needed for anything relatively low-voltage, so high frequency transformers are necessary for this.

I don't know why some people seem so reluctant about this, except for the fact that we're dealing with high power and mostly custom windings on a core.

Steve
 

thingmaker3

Joined May 16, 2005
5,083
I don't know why some people seem so reluctant about this, except for the fact that we're dealing with high power and mostly custom windings on a core.
A shorted component in an isolated power supply can be cause for inconvenience. A shorted component in a non-isolated power supply can be cause for a funeral. Just don't do it.
 

Papabravo

Joined Feb 24, 2006
21,159
Yes I see it now. This is actually similar in concept to power factor correction schemes where you get maybe 160VDC in NA or 310VDC in Europe. Then you chop it at high frequency and step the voltage down. I think this ends up being more headache and components than it is worth. Especially the FET with the large breakdown voltage. The other thing is that since the duty cycle is so small you don't have much dynamic range to adjust to load transients since small changes in duty cycle produce large changes in output voltage. It's kinda like having a tiger by the tail. Let us know how it works out for you, but I predict that the results will be disappointing assuming you don't win a Darwin award along the way.
 
Top