SMPS chip but not for 240V?

[FriedMule]

I have tried to learn about buck converters, with the goal to learn how you convert from 240V to 5V but I keep running into chips for low voltage except a few for 600V-1000V. Even Davids "jelly bean"  MC34063 is for max 40 voltage. Where do the 240V go in, is 40V fine, I seem not to understand how it all can be.

Hope thay uoy please will help? :-)

[james_s]

While high voltage buck converters do exist, that topology is not particularly common for such applications. There are lots of power conversion ICs out there intended for converting mains voltage down to low DC voltages.

[FriedMule]

While high voltage buck converters do exist, that topology is not particularly common for such applications. There are lots of power conversion ICs out there intended for converting mains voltage down to low DC voltages.

Thanks, but is that the way a standard SMPS do it?

[oPossum]

Typically flyback or forward converter.

[FriedMule]

Typically flyback or forward converter.

I have hard of these terms but do not understand how it all works, how does Dave's jelly bean at max 40V handle 240V?

[brabus]

You usually have an external MOSFET or a BJT doing the hard lifting and dealing with the high voltages and currents. The controller only provides the driving commands for them.

[oPossum]

There are offline flyback controllers with integrated switch for low power applications.

A random example... https://www.onsemi.com/pub/Collateral/FSL137H-D.pdf

Using a MC34063 with an external switch may be possible, but there are may other better choices.

[magic]

In addition, they usually use an auxiliary transformer winding which produces 5~20V for the SMPS chip.
Startup is provided by a pullup resistor which slowly charges the controller's power supply capacitors when the PSU is connected to mains. This is possible because the controller draws little power before it becomes active.

[FriedMule]

You usually have an external MOSFET or a BJT doing the hard lifting and dealing with the high voltages and currents. The controller only provides the driving commands for them.

Am I guessing right in that the Mosfet is better because of less voltage drop?

There are offline flyback controllers with integrated switch for low power applications.

A random example... https://www.onsemi.com/pub/Collateral/FSL137H-D.pdf

Using a MC34063 with an external switch may be possible, but there are may other better choices.

I'll take a good look at that! :-)

In addition, they usually use an auxiliary transformer winding which produces 5~20V for the SMPS chip.
Startup is provided by a pullup resistor which slowly charges the controller's power supply capacitors when the PSU is connected to mains. This is possible because the controller draws little power before it becomes active.

Do you mean a transformer before the SMPS circuit?

[MosherIV]

Hi.
Typically mains voltage smps rectify the mains AC into a high voltage cap.
This HV DC is pulsed into an isolation transformer at high freq.
The freq controls the output voltage/current.
The output V needs low ESR caps to smooth the output of the isolation transformer.
The output voltage is monitored and fed back via an opto isolator.
This feedback is used to determine the freq to pulse the isolation transformer.
This is all controlled by a ASIC which is powered striaght from the HV DC.

I think (but I am not sure) this is the 'forward converter' topology.

[oPossum]

If it has one switch it is probably flyback. If it has two switches (or a multiple of 2) it is probably forward.

Switch refers to whatever is doing the switching - BJT, MOSFET, IGBT

[FriedMule]

Hi.
Typically mains voltage smps rectify the mains AC into a high voltage cap.
This HV DC is pulsed into an isolation transformer at high freq.
The freq controls the output voltage/current.
The output V needs low ESR caps to smooth the output of the isolation transformer.
The output voltage is monitored and fed back via an opto isolator.
This feedback is used to determine the freq to pulse the isolation transformer.
This is all controlled by a ASIC which is powered striaght from the HV DC.

I think (but I am not sure) this is the 'forward converter' topology.

Yes thats sounds logical, and like the attached drawing. But In my book it looks like that 240V is connected to the chip?

[oPossum]

Do you mean a transformer before the SMPS circuit?

See figure 1 in the FSL137 data sheet.

[oPossum]

But In my book it looks like that 240V is connected to the chip?

That is for a bootstrap circuit that gets the chip running. Operational power comes from the transformer winding.

[FriedMule]

Do you mean a transformer before the SMPS circuit?

See figure 1 in the FSL137 data sheet.

So the 240V does not go undisturbed down to the HV-pin at the chip?

[mikeselectricstuff]

But In my book it looks like that 240V is connected to the chip?

That is rare - the most common arrangement is a resistive dropper from HV to get it started, then use a winding on the transformer for normal running. 

There are some pure buck converters designed for mains voltages, look at Power Integrations' LinkSwitch series.
The reason that simple buck converters aren't common for mains is that mains PSUs usually need isolation, so a flyback with a transformer makes the most sense.

[tunk]

I guess it's a bit beside the point, but the 240V is an average and
after rectification and smoothing it's more like 340V (240*2^0.5).

You could also take a look at some of DiodeGoneWild's youtube videos.

[FriedMule]

Oh, that makes a  lot of sense. What are the difference between the two schematics? Safety, efficiency, pros / cons?

[FriedMule]

Bump:
I am not asking for a scientific explanation, just a simple one, since one has a lot more components then the other but both should buck converters. :-)

[james_s]

Buck converters do not provide isolation between line and load. Typically if you are going to go to the effort of making a off-line SMPS you may as well go with an isolated topology.

[FriedMule]

Buck converters do not provide isolation between line and load. Typically if you are going to go to the effort of making a off-line SMPS you may as well go with an isolated topology.

Great advice, Both of these do look like isolated, but are one better the the other?

[David Hess]

While high voltage buck converters do exist, that topology is not particularly common for such applications.

Typically flyback or forward converter.

The reason for this, beyond galvanic isolation from input to output, is that the transformer adjusts the ratio of voltage to current in the power switch to a more advantageous value.  If a buck converter is used, then the power switch has to handle the maximum output current and the maximum input voltage making it significantly more expensive and lowering efficiency as well.  The transformer in a flyback or forward converter considerably reduces the current which the power switch must handle.

[FriedMule]

While high voltage buck converters do exist, that topology is not particularly common for such applications.

Typically flyback or forward converter.

The reason for this, beyond galvanic isolation from input to output, is that the transformer adjusts the ratio of voltage to current in the power switch to a more advantageous value.  If a buck converter is used, then the power switch has to handle the maximum output current and the maximum input voltage making it significantly more expensive and lowering efficiency as well.  The transformer in a flyback or forward converter considerably reduces the current which the power switch must handle.

Yes I see the logic in that and you do also, as written earlier" get added safety". What I am out to achieve is a 240V to 5V isolated smps. Since I have about no idea on what to do, I hope that some of you could point me to a schematic that is universally okay. I know that higher frequency is better but do also demand a lot more skills in designing the circuit board.

[MosherIV]

There is no one circuit that is universally ok, it depends on the parameters of the circuit.
Easiest is to find app notes for particular devices.

Building your own 
It is highly dangerous - there is rectified mains, HV DC, with all the power of your national grid behind it !
It is very difficult to find the isolation transformers. They are not easily available off  the shelf.

If you are doing this for learning, be very careful 

[BravoV]

Just because you could do & probably understand "buck converter" from say 12 V down to 5 V, it doesn't mean you can still to stick with buck converter ONLY to do mains 240V AC down to 5 V DC.

Go learn and study further other topologies as pointed by many posters above, that is not smart safe method of practicing electronics, especially related to mains voltage.

[FriedMule]

Just because you could do & probably understand "buck converter" from say 12 V down to 5 V, it doesn't mean you can still to stick with buck converter ONLY to do mains 240V AC down to 5 V DC.

Go learn and study further other topologies as pointed by many posters above, that is not smart safe method of practicing electronics, especially related to mains voltage.

There is no one circuit that is universally ok, it depends on the parameters of the circuit.
Easiest is to find app notes for particular devices.

Building your own 
It is highly dangerous - there is rectified mains, HV DC, with all the power of your national grid behind it !
It is very difficult to find the isolation transformers. They are not easily available off  the shelf.

If you are doing this for learning, be very careful 

I am glad for your warning, but the mains side of the project am I not unsure about, it is the part I know most about and have done most work on, i.e. at worksites and home, I am not an electrician but can do most installations. I am also planning on using a variac for a minimum of safety.
I keep reading, look at videos and know that there is a lot to learn. All I have previously done is bould "old" school power supplies.

Do you have any advices, beside YouTuber DiodeGoneWild on who to wach, where to read and so on?

[james_s]

You are in way over your head, I don't intend to sound rude however if you had the knowledge to do this safely and successfully you would not be asking many of these questions. SMPS design is a rather advanced topic and you will need to be proficient at all the underlying concepts. If you don't get everything just right, they have a tendency to blow up instantly when power is applied, or if they work they may be spewing copious amounts of RF hash and knocking out radio or TV reception. You are asking someone to explain to you how to fly a 787 when you have never flown a Cessna or drive a Ferrari when you have never driven a basic passenger car.

Start reading every application note and textbook you can find on the subject, study existing designs, eventually you will get there but it's not easy.

[soldar]

You are in way over your head, I don't intend to sound rude however if you had the knowledge to do this safely and successfully you would not be asking many of these questions. SMPS design is a rather advanced topic and you will need to be proficient at all the underlying concepts.

+1

I have repaired dozens of this type of SMPS and have a pretty good understanding but I would not dream of building one. I would still need to study and learn a lot more. A lot. This is a very specialized field and no one is going to be designing and building this in a few hours.

[FriedMule]

You are in way over your head, I don't intend to sound rude however if you had the knowledge to do this safely and successfully you would not be asking many of these questions. SMPS design is a rather advanced topic and you will need to be proficient at all the underlying concepts.

+1

I have repaired dozens of this type of SMPS and have a pretty good understanding but I would not dream of building one. I would still need to study and learn a lot more. A lot. This is a very specialized field and no one is going to be designing and building this in a few hours.

You are in way over your head, I don't intend to sound rude however if you had the knowledge to do this safely and successfully you would not be asking many of these questions. SMPS design is a rather advanced topic and you will need to be proficient at all the underlying concepts. If you don't get everything just right, they have a tendency to blow up instantly when power is applied, or if they work they may be spewing copious amounts of RF hash and knocking out radio or TV reception. You are asking someone to explain to you how to fly a 787 when you have never flown a Cessna or drive a Ferrari when you have never driven a basic passenger car.

Start reading every application note and textbook you can find on the subject, study existing designs, eventually you will get there but it's not easy.

Sorry that I was not clear. What I meant was that the safety and handling mains, is what I know most about. I can understand that you could compare me to knowing the sea and how to navigate it safe, but what I am trying to do is to sail around the world in a DIY atomic u-boad:-)
My error in thinking that I could build a SMPS was that Dave, in a video is saying that it is almost the same, that the one is not much harder then the other. And since I know how to make a decent linear power supply, what was then holding me back? :-)

[soldar]

Great advice, Both of these do look like isolated, but are one better the the other?

They are basically the same thing. One is a conceptual explanation of how the circuit works and the other is an actual application. The first one omits many minor things not needed to understand the working of the circuit.

You need to read, study and understand the document in its entirety. You are asking questions which are answered in there. Reading manufacturers' application notes is a very good way of learning.

You need to read up about the different topologies and what they are good for. Nobody can give you here the basic knowledge needed to design SMPS. That requires hundreds of hours of study and learning.

For a start:
http://ww1.microchip.com/downloads/en/appnotes/01114a.pdf

BTW, it says

A  buck  converter,  as  its  name  implies,  can  only produce  lower  average  output  voltage  than  the  input voltage.

"As the name implies"? What is the meaning of "buck" in this context? And why does it imply it con only produce lower voltage?

Another: https://www.ti.com/seclit/ug/slyu036/slyu036.pdf

[FriedMule]

Great advice, Both of these do look like isolated, but are one better the the other?

They are basically the same thing. One is a conceptual explanation of how the circuit works and the other is an actual application. The first one omits many minor things not needed to understand the working of the circuit.

You need to read, study and understand the document in its entirety. You are asking questions which are answered in there. Reading manufacturers' application notes is a very good way of learning.

You need to read up about the different topologies and what they are good for. Nobody can give you here the basic knowledge needed to design SMPS. That requires hundreds of hours of study and learning.

For a start:
http://ww1.microchip.com/downloads/en/appnotes/01114a.pdf

BTW, it says

A  buck  converter,  as  its  name  implies,  can  only produce  lower  average  output  voltage  than  the  input voltage.

"As the name implies"? What is the meaning of "buck" in this context? And why does it imply it con only produce lower voltage?

Another: https://www.ti.com/seclit/ug/slyu036/slyu036.pdf

Thank you a lot, great links and helpful comment, I will do as you and other suggests. :-) I can also start by learning ad trying out low voltage bucks (12V to 5V) and so on, if I destroys some components on the way, no harm:-)