SMPS input 110VDC to 12VDC

[vantusaonho]

Hello everybody!
 I am trying to design my own a SMPS with the requirement input 110VDC, output 12VDC 1A. I read buck converter but I don't find any IC for high voltage step down. Please guide me for researching.
Thanks

[Gallymimu]

http://www.onsemi.com/pub_link/Collateral/NCP1028-D.PDF
http://www.onsemi.com/pub_link/Collateral/NCP1072-D.PDF

http://www.linear.com/product/LT1105

http://www.powerint.com/products/tinyswitch-family/tinyswitch-4

what you are looking for is called an off-line switcher.  There are quite a few chips.  I have listed a few samples.

[vantusaonho]

If is there any way to design smps 110VDC to 12VDC without pulse transformer? if I can use MC34063 for this application?

[enzine]

Beware the dangers of project you want to achieve!
It is not possible to realize such a converter without a transformer for safety problems.
The 12V output would not be electrically insulated from the source to 110Vdc.
Is much more easy, safe and economical to use a commercial converter 80 ... 250Vac - 12Vdc, almost all converters have at input stage a diode bridge to transform the ac voltage  in dc and then do the conversion.

[vantusaonho]

I decided that I will use KA3842 IC controller for PWM, and a pulse transformer. I also need a KA431 for current feedback. I can imagine my project. But it is still a problem that I don't know design transformer. Turn, coil size, core size, Primary, secondary, auxilary coil., how to turn around coil. Please share me  a bit experiment about KA3842 and transformer. After finishing this project, I will post a detail tut for designing my smps.
Thanks all!

[Gallymimu]

This is a good book for transformer design.  I think a beginner could get the hang of it reasonably quickly.

http://www.sonsivri.to/forum/index.php?topic=48476.0

Posted on: November 04, 2012, 04:46:16 16:46 - Automerged

Beware the dangers of project you want to achieve!
It is not possible to realize such a converter without a transformer for safety problems.
The 12V output would not be electrically insulated from the source to 110Vdc.
Is much more easy, safe and economical to use a commercial converter 80 ... 250Vac - 12Vdc, almost all converters have at input stage a diode bridge to transform the ac voltage  in dc and then do the conversion.

A little bit of pity for my english !!!

Ciao

There isn't anything wrong with designing this without an isolation transformer. BUT you do have to take the proper safety design precautions. i.e. reinforced / double insulation.  As long as there is no human touchable access to any conductive parts it can be just fine.  Certainly isolation is better if possible.

Either way you should always make sure you are compliant with local safety standards.  IEC61010 and IEC60950 are good starting points for anything with dangerous voltages.

[vantusaonho]

Please explain for me what function of C4, R12, D7 in this schematic. Thanks

[chandra2sekhar2000]

hai,D7 is used to supress the Back Emf spikes,C4 ,R12 helps reducing the sharp spikes.
my suggestion is,if u want to go for Bulk production,then go with custom SMPS design.else if u want to go with just one unit,better go for a ready made unit <$2.

[solutions]

^ is correct.

If you want to understand its design, look up "snubber network" in a power supply context.

[FTL]

D7 is a flyback diode. It allows the current created in the 1-2 coil on the transformer to continue to flow when Q1 turns off. As mentioned, this stops a large voltage spike from occurring and killing Q1.

R12 is in the circuit to add resistance to the flyback loop so the energy from the collapsing magnetic field can be dissipated so the field will collapse faster.

I assume C4 is a small value, high voltage unit. I'm guessing it is there to help absorb any hi-frequency voltage spikes that could cause RF interference. There will be a bit of a spike in the flyback loop as D7 changes from reverse to forward biased. C4 should be able to store that energy to keep the switching a bit cleaner.

[Tech_n]

I would advice to consider use some integrated solutions like TOPSwitch. They are a little bit more expensive, but i have very good experince using them. Also you can buy transformers for use with this parts, so you dont have to wind your own. But there are not many turns, so winding it by hand is easy. I reccomend bought transformer, the isolation is guaranted and they arent so expensive.

For some introduction into this problematics, read "Power supply cookbook" by M. Brown.

[pickit2]

I would advice to consider use some integrated solutions like TOPSwitch. They are a little bit more expensive, but i have very good experince using them. Also you can buy transformers for use with this parts, so you dont have to wind your own. But there are not many turns, so winding it by hand is easy. I reccomend bought transformer, the isolation is guaranted and they arent so expensive.

For some introduction into this problematics, read "Power supply cookbook" by M. Brown.

If you want to use TOPSwitch ic's and need a jump in point, and have a panasonic workshop near to you, ask for scrap boards from microwave ovens, they have the transformers ready wound, if the topswitch ic is blown.
Panasonic circiut follows very close to data sheets for TOPSwitch.

[vantusaonho]

Oh, It is confusing to calculate the turns for primary winding.
First, select the core type belong to frequency operation, power output
Second, Calculate core cross sectional area Ae, determine flux density B from core type which is chosen
Third, calculate the turns of primary winding in the formular: Npri = ET/(B*Ae*10 exp-8)
I don't know how to choice ET (energy handling ability of transformer)and all right?
Please let me know the where is not exactly. It is seem that there are many way to calculate this.

[Gallymimu]

Oh, It is confusing to calculate the turns for primary winding.
First, select the core type belong to frequency operation, power output
Second, Calculate core cross sectional area Ae, determine flux density B from core type which is chosen
Third, calculate the turns of primary winding in the formular: Npri = ET/(B*Ae*10 exp-8)
I don't know how to choice ET (energy handling ability of transformer)and all right?
Please let me know the where is not exactly. It is seem that there are many way to calculate this.

What you are suggesting is a good place to start.  There are a lot of tradeoffs in transformer design.  I don't know how cost/space constrained you are but for energy handling if you go with less than a watt per cubic inch you should be okay but that is probably too conservative.

You could look at some basic equations for convection to determine how much power can be dumped to free standing air for a given exposed surface area.  That should get you reasonably close. 

[Tech_n]

For my designs with the integrated switchers i used this program

http://www.powerint.com/en/design-support/pi-expert-design-software

I used the core with the closest area. Because i cant get gapped cores, i used PET tape or paper for the air gap. But this makes a gap also in the outer E legs, which is undesirable, because this increases EMI radiation of the supply.

I think for the right frequency and power, you could choose an right design also for your UC3845 and the like supply. You need the sw only for winding calculations.

[solutions]

While it's nice to learn new things, you have to ask yourself if your time is better spent and then either buying a power supply or using a canned solution off one of the switcher IC manufacturer's websites...like webbench from National (now TI). The online tools use in-stock magnetics.

Winding your own is pretty hardcore

[vantusaonho]

I am a beginner in smps. I want to design smsp with input ACmin 110V and shutdown when the mains down to below 90V, please help me with the solution for it.

[The Conqueror]

Are you exactly looking for some transformer less models in that range

[vantusaonho]

Now I can design smps transformer for flyback topology using UC3842, but I don't know to shut down the power supply when the mains is downed below 90VAC

[solutions]

I don't understand why you want to shut it down. If it's a duty cycle limit on your controller stability, just monitor the PWM duty cycle. If not, ride it all the way down to 15V instead of 90

[vantusaonho]

My project is to design a smps with flyback topology which have input 110 -220VAC, output 12V 1A. I have already schematic for this but I want to shut down power supply when the mains below 90VAC to protect load. I use UC3842 to control PWM duty, and I wonder that Vfb pin of UC3842 have been used for monitoring output voltage. How to monitor input voltage to shut down smps and start again when the mains raise to 110VAC again.

[Gallymimu]

My project is to design a smps with flyback topology which have input 110 -220VAC, output 12V 1A. I have already schematic for this but I want to shut down power supply when the mains below 90VAC to protect load. I use UC3842 to control PWM duty, and I wonder that Vfb pin of UC3842 have been used for monitoring output voltage. How to monitor input voltage to shut down smps and start again when the mains raise to 110VAC again.

I agree with solutions and don't really understand why you would want to do this.  Parts like the UC3842 aren't really meant for that so anything you do will be a bit of a kludge.  One thought would be to use a threshold detector on the rectified input and feed it's output into the feedback pin to force the output high or low... whichever will force the voltage output down to zero, if it is even capable of going that low.  Sorry I can't be more clear but I don't have time to dig into the design implementation with the UC3842.

Now that I think about it you might also be able to set up a threshold detector on the rectified input and then use that to control a pass transistor to the regulator.  That might be a better option.

Solutions do you know of a simple BJT threshold detection circuit off the top of your head.  I don't do much discrete transistor work.  Would something like a discrete schmitt trigger with some hysteresis be the right way to do it?

Posted on: November 11, 2012, 03:33:13 03:33 - Automerged
maybe this:

http://www.daycounter.com/Circuits/Schmitt-Trigger/Transistor-Schmitt-Trigger.phtml  The instructions for this one aren't very clear.

http://www.johnhearfield.com/Eng/Schmitt.htm

you'll have to change things around so that the threshold voltage is set properly (probably with a zener feeding Q1 VP) and then use this circuit output to feed the pass transistor.

[vantusaonho]

Please remark to help me
My smps:
Input voltage: 110 - 220VAC
Output voltage: 12VDC
Output current: 1A
efficiency: 0.75
Dmax: (On cycle duty): 0.45
frequency smps (fs): 40KHz
Total power: Pout = (Vout +V_diode)* Iout = (12V +1.2V)*1=13.2W
Ipp=(2*Pout)/(Vdc_min*eff*Dmax) = 0.5A
L_primary: (Vdc_min*Dmax)/(Ipp*fs)= 0.00348 H
Ns= Te*Vout = 15 turns  - I choose Te=1.25 turns per volt
Np=Ns*(Vdc_min)/(Vout+Vd)*(Dmax/(1-Dmax)) = 175 turns
Nvcc = Ns*(Vcc+Vd)/(Vout+Vd) =22 turns  - Vcc for UC3842  I choose 18V
Core wire for Primary: Ap = Irms/J= 0.04 mm^2  => core #31 AWG 0.3mm diameter
Core for secondary: core #24 AWG 0.6mm diameter
Core for Vcc winding: #28 AWG 0.4mm diameter

But when I calculate snubber circuit with clamp mode at this site
 http://www.daycounter.com/Calculators/Snubbers/Snubber-Design-Calculator.phtml
Those value are really terrible. Please help me the right formular snubber circuit design
 

[solutions]

My project is to design a smps with flyback topology which have input 110 -220VAC, output 12V 1A. I have already schematic for this but I want to shut down power supply when the mains below 90VAC to protect load.

If you are intent on "protect the load", why are you not shutting the supply down when you get excessive dropout on the 12VDC itself? Who cares if it's 90VAC or 30VAC on the primary side?

[xenix]

There are two types of snubbers for flyback circuit:
•Snubber for transformer
•Snubber for transistor

In the above circuit, snubber is used just for transformer. Circuit parameters for such a snubber depends on transformer primary leakage inductance. Therefore without knowing the leakage inductance, you can not design the perfect snubber.

In theory, when the mosfet is turned off, the energy stored in magnetizing inductance of the transformer will be transferred to secondary winding. Since the leakage inductance of the primary winding is not coupled to secondary side, current on the leakage inductance will not be transferred to secondary side. This current has to continue its way, if there is no snubber, it will charge Vds capacitance of the mosfet to very high values and probably mosfet will be damaged. If there is snubber, the leakage inductance energy will charge the snubber capacitor. So we can formulate the following:

½×L×I²=½×C×V²

where L is the leakage inductance, C is the snubber capacitance, I is the current before the mosfet turned off and V is the voltage rise on snubber capacitor.

So a proper voltage rise which will not damage the mosfet should be chosen to decide C value. Also you should know, or at least estimate the leakage inductance. For a nicely designed transformer, it is 1% to 3% of the main inductance. But for a poor design, it can be 20% of the main inductance.

Snubber resistance value is not critical, it should discharge the capacitor befor mosfet is turned on. R×C=T/10 would be a practical formula to decide the resistor value (T is the switching period). The diode should be fast, never use 1N400x series !!!

[zab]

All you require is an smps ic. & high side fet driver. Then you can make buck circuit even  from 600 v to 12v dc easily.
But remember it  is risky.

[vantusaonho]

Please let me know how to calculate the voltage rise

[pickit2]

why don't you download the data sheets, do a bit of reading, first.
save us all from pain, why not buy one from ebay and dissect it and post your findings.

you don't even take advice, so why should we spoon feed you, do some research yourself...

[vantusaonho]

why don't you download the data sheets, do a bit of reading, first.
save us all from pain, why not buy one from ebay and dissect it and post your findings.

you don't even take advice, so why should we spoon feed you, do some research yourself...

I don't need anything spoon feed, I have already read datasheet, tut, ebook, paper thesis, journal. But they are diffirent from each others. It take me very confusing. Now I need a right thing of experimental person. If I do research myself, this forum is not mean. If  someone have experimental in smps, but don't share theirs. This forum what for?

[pickit2]

I don't need anything spoon feed, I have already read datasheet, tut, ebook, paper thesis, journal. But they are diffirent from each others. It take me very confusing. Now I need a right thing of experimental person. If I do research myself, this forum is not mean. If  someone have experimental in smps, but don't share theirs. This forum what for?

so why not find one that works, sounds as if you do need spoon feeding, "If  someone have experimental in smps, but don't share theirs. This forum what for?" your asking for a full tried and tested, solution, but don't take advice.

asking questions like "Please let me know how to calculate the voltage rise" don't make sense, voltage rise where? line or feedback.

Do you even know what dangers are involved, in such a product, your skill levels are not just there.

I have 10 plus such smps as part of products, placed in market place, by my employer sorry I can't give full details of products.
But they follow very close to the data sheet and Application Notes.

http://www.powerint.com/en/products/product-documents
http://www.powerint.com/sites/default/files/product-docs/miniapplianceweb_PSG.pdf
- - - - - (part of spoon feed) - - - - -

[vantusaonho]

Sigh!!!!!!!!!!!!!!!    I am a newbie in smps. And I am in Vietnam where can not buy anything in foreign. I can only buy some IC from China. How to let me try with Powerint product. I use UC3842 because I can get it. I almost have all design for myself but there are some problem with voltage rise when the switch off coil.

[pickit2]

the uc3842 is close to powerint devices, so circuit is about the same.
 
do you have a panasonic service agent, in your local, ask then for some scrap boards from microwave oven, not the inverter type.
they have powerint devices, have a 12-15Vdc at 1 to 2. plus 5v, get the apps notes: and you have all the parts to make a smps and no worry about transformer build.

[solutions]

What's it for and how many?

What's wrong with using these (not an endorsement, just an example)?



http://store.fundamentallogic.com/ecom/index.php?main_page=product_info&products_id=491

I would think you have better things to do than design a power supply and wind your own transformer (I find it hard to believe you can get your hands on the cores where you are), especially when you can pick one up CHEAPER than the parts you'll buy for your own design...cheaper, in fact, than the bare circuit board....

And for prototyping, you can just buy a portable hard drive, which uses a 12V 1A, 1.5A, or 2A wall wart.

[Ichan]

And I am in Vietnam where can not buy anything in foreign.

You have to find a way to break that boundary through, years ago i am on the same situation with you.

Anyway, learning from cheap power supply is worthy - i ever found one which only use several bipolar transistor.

-ichan