Using Buck Converter to regulate from 120V to low DC

[zmoose]

I have been wondering why people don't use a buck converter (switched regulator) to go straight from 120V to a lower DC value, like 12V or 5V. Most conventional power supplies use a transform to go to a lower intermediate voltage and then use a switched regulator to get the desired regulated output.

It would seem that a two stage buck design can easily do the same without the need for a transformer. For example, if a 120VAC source is first rectified (at 120V) to DC, then it can be down converted to a lower intermiate voltage like 8V and a second buck stage can then be used to provide the final regulated output like 5V or 3V. I know AC isolation is one reason why a transformer might be needed. Are there other reasons why a buck (swithched) converter is NOT used in this manner. If would seem that a two stage buck design could be more cost effective and take less space than requiring a transformer.

Curious Moose

 

[Skogsgurra]

The one reason you mention, isolation, is good enough. Especially since many grids are vague on where phase and ground finally end up in an installation.

Also, using a non-transformer technique would exclude the use of a bridge rectifier. That would cause a problem with DC load on the grid.

Rest assured, if it isn't being done - then there are good reasons for it.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[MacGyverS2000]

That, and using two converters adds to the inefficiency of the design. And converting from 8V to 5V is inefficient (the closer the input and output voltages the higher the inefficiency).

Dan - Owner

http://www.Hi-TecDesigns.com

 

[IRstuff]

I might dispute the "cost effective" argument as well. Components that handle a couple hundred volts are not particularly cheap, compared to components that handle 12 V or so. Near as I can tell, there's at least factor 2x cost for higher voltage components. When all the extra component costs are summed, they'll be more than the cost of a 12V transformer

TTFN

FAQ731-376

 

[zmoose]

Thanks for all your replies. I really appreciate it.

Regarding isolation, I know it is an important consideration. However, I've seen some AC charger circuits that use a transformerless design, although I'm not sure if they are actually used in commercial AC chargers. Is isolation an absolute requirement.

Regarding not being able to use a bridge rectifier with a non-transformer design. I don't understand why that would be the case since there will be floating ground in either case. I know I'm missing something fundamental here. Please pardon my ignorance.

Regarding efficiency, I don't think having a two stage buck would be less effiencient than a transformer and a buck. The efficiency of a single stage buck should be comparable to that of a transform. So the efficiency of the two should be comparable.

Regarding the component cost of a buck vs a transformer. You are correct in that higher voltage component would cost more and therefore may end up costing more than a transformer. However, a conventional supply need a large capacitor at the output to store enough energy to sustain it for at least half a 60Hz cycle. This is espepcially true to higher power applications. I think a two stage would require a much smaller output cap since the first stage of the buck can be continuously maintained from the AC source. Just a thought.

Anyway, thank you all very much for indulging me in this exercise.

 

[itsmoked]

Isolation.. Required.

Let me put it this way.. Do you want to describe to a jury that it's a shame that the child died, but you were able to save a couple of bucks by dropping isolation?

Think Ford Pinto..

Just not worth entertaining the thought.



Keith Cress
kcress -

http://www.flaminsystems.com

 

[benta]

I don't see where your savings come in using a two-stage design.
There are two "normal" ways of providing isolation, and both are cheap:
1: normal 50/60 Hz transformer.
2: flyback converter with high frequency transformer.

Benta.

 

[Comcokid]

Actually, there are buck converters available to take 120 VAC and buck convert down to the 5 to 24 volt range.

Check out the Power Integrations Linkswitch-TN family. Designs can be made that can be UL and CE approved

http://www.powerint.com/en/products/linkswitch-family/linkswitch-tn

Not everything that connects to AC needs isolation - it depends upon the product and application.

 

[Skogsgurra]

zmoose says: "Regarding not being able to use a bridge rectifier with a non-transformer design. I don't understand why that would be the case since there will be floating ground in either case. I know I'm missing something fundamental here. Please pardon my ignorance"

Using a single diode rectifier will make the Neutral available to the user. Using a bridge rectifier will not. It depends a lot on what you will be using your power supply for. If it is for embedded control in a closed apparatus, it probably doesn't matter if your neutral is available to the user. If it is in a general application with sensors and actuators connected, then it matters a lot.

Is your question about embedded power supplies or general use supplies? In the former case, even I have used simpler "direct drop" power supplies. The authorities are rather strict about component choice and circuit topologies, though.

Gunnar Englund

http://www.gke.org

--------------------------------------
100 % recycled posting: Electrons, ideas, finger-tips have been used over and over again...

 

[butterworth]

Comcokid i checked out those products in your post. They look like the cheapest, simplest, lowest component count SMPS for eg a few Watts to eg switch out the inrush resistor relay in a Power factor corrected Switch Mode Power Supply.

Hope any reader can tell me if this is wrong.

And i didn't know there was such thing as offline buck smps control IC's.

I reckon the little LinkSwitch is just right......it would be better though if those 400V bus caps could be done away with in its datasheet. (obviouly a diode would have to be put in to mitigate the mains zero crossing situation) I presume it would still work like this but just give a bit poorer output regulation(?)

Till now i have been using a start-up linear regulator -but now i reckon these products are better -it doesn't even need a Vcc power supply -pulls its own from the mains, and no high-side drive for the FET.

 

[carnage1]

the other issue with a buck system is the output voltage tends to be a little less clean than a cheap linear running off a rectified step down transformer.

Luck is a difficult thing to verify and therefore should be tested often. - Me

 

[sreid]

I'm a little confused about what's being discussed here. Virtually all "Off The Line" power supplies have no 60/50 Hz transformer. There is a transformer (which provides isolation) but the transformer is (almost) an intregal part of the power supply design. Consider any PC power supply.

 

[butterworth]

carnage1.. I hear that a sub-miniature mains (50Hz) transformer can be a good idea for an auxiliary supply within an SMPS.....especially in large SMPS's.

If it was only for a few Watts then i suppose the fact that it probably uses a diode bridge followed by a smoothing cap probably doesn't make too much difference for EMI.

-Though these days, with regs being so tight , its tempting to go for the more power factor corrected solution of say a Linkswitch, or the like.

I would probably run the fans off an auxiliary coil, or another SMPS switching off the main output.

At lower power, the problem i find with flybacks with auxiliary windings is that when the load is removed, the auxiliary coil voltage drops sharply.

(of course the load is never absolutely removed because the divider (feedback) resistors are there, and any cap bleed resistors.

 

[Caledin]

probably it matters that AC voltage sometimes jumps to I don't know what value, and with a transformer it looks not that bad.