Connect with us

convert switching-mode PS to stabilized PS?

Discussion in 'General Electronics Discussion' started by arg733, Feb 1, 2013.

Scroll to continue with content
  1. arg733

    arg733

    89
    0
    Dec 14, 2010
    Hi.
    Is it possible to convert a 15A switching mode power supply to stabilized power supply?
    And if yes why there is no such internal circuit?
    Thank you.
     
  2. BobK

    BobK

    7,682
    1,686
    Jan 5, 2010
    What is a stablized power supply?

    Bob
     
  3. arg733

    arg733

    89
    0
    Dec 14, 2010
    Well... it has no output noise no Hz. When i measure the "switched power supply" (220ac to 12dc) it has more than 20kHz out (my multimeter can't show more) but when i measure an "stable power supply" (regular 220ac to 12dc adapter) I get no 0Hz just like a battery.
    The "Switched-mode power supply" is noticeably lighter and smaller.

    HTML:
    http://en.wikipedia.org/wiki/Switched-mode_power_supply
    Also should that >20Khz have any negative effects on ic's and other electronics?

    Thank you.
     
    Last edited: Feb 1, 2013
  4. BobK

    BobK

    7,682
    1,686
    Jan 5, 2010
    Yes, switching supplies might have high frequency noise. But I use them all the time. Are you experiencing a problem that you think is due to power supply noise?

    Bob
     
  5. arg733

    arg733

    89
    0
    Dec 14, 2010
    Never tried to use one with circuits I only used it as a supply for loads with no electronics circuitry involved.
    So if i want to make let's say a 555 oscillator , would that high frequency noise cause problems like altering the 555 output? When should I use switching supplies and when should I not? :confused:
     
  6. (*steve*)

    (*steve*) ¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd Moderator

    25,387
    2,772
    Jan 21, 2010
    What you call a "stable power supply" is more than likely one using a linear regulator. What you call a "switched power supply" is a switchmode power supply. (or a switchmode regulator).

    There are times you would definitely use one or the other.

    For example, you'd never use an SMPS with a high quality audio preamplifier.

    You'd never use a linear power supply with a high power computer.

    Inbetween these there are tradeoffs you need to consider.

    In general, if the power supply needs to dissipate more than a couple of watts (and it must be regulated) then you would be looking at an SMPS.
     
  7. arg733

    arg733

    89
    0
    Dec 14, 2010
    But if i make 555 oscillator would it function properly? I mean that it will be getting 100khz input from the switched mode power supply , right? Is that ok?

    Also, can you explain this a bit more?

    Thanks
     
  8. Raven Luni

    Raven Luni

    798
    7
    Oct 15, 2011
    Yes it will function properly. A switching supply simply gives an average voltage. Noise is practically nonexistent. In fact it they are probably less noisy than a linear regulator operating on a mains supply.

    The only way to get perfectly smooth DC is to use a battery.
     
  9. (*steve*)

    (*steve*) ¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd Moderator

    25,387
    2,772
    Jan 21, 2010
    SMPSs tend to radiate more noise than linear PSUs. However the ripple is generally very small.

    For a 555, I would be completely happy.

    In answer to the second question, SMPS's tens to be smaller and more efficient than linear power supplies. You can get a 1000W SMPS to fit into a small box in a computer case. An equivalent linear power supply would require a massive heatsink and it would dissipate *at least* a couple of hundred watts. (And if it weighted less than 5 kG I'd be surprised.

    Another factor is that SMPSs can be made to operate from (say) 90 to 265 volts, making them usable from mains anywhere. A linear power supply is probably going to want the mains to be within +/- 10 to 15%, and if you want the same device to run from 110V as well as 220V, it will need a switch to select voltages.
     
  10. Raven Luni

    Raven Luni

    798
    7
    Oct 15, 2011
    So whats the difference between noise and ripple (as seen by the circuit)
     
  11. arg733

    arg733

    89
    0
    Dec 14, 2010
    So always use SMPS except audio applications?
    I used a cap parallel to the smps and earthed the negative node, to see what would happen, and the frequency dropped to zero does this means that now the output is the same with a linear ps?
     
    Last edited: Feb 3, 2013
  12. (*steve*)

    (*steve*) ¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd Moderator

    25,387
    2,772
    Jan 21, 2010
    Well, ripple is *A* form of noise, but not all noise is ripple.

    As an example, you could look at the output of a SMPS and see 100kHz ripple (from the switching frequency), with a small 100/120Hz component from the DC rectification that occurs before the SMPS, as well as small variations caused by changes in load and line voltage.

    On top of that, when someone switches a fridge on (say) you may get impulse noise on the mains that may make it through to the SMPS output. Or a nearby SMPS may radiate (magnetically) a signal that is picked up by this SMPS.

    Then there's random noise from resistors and semiconductor junctions, etc...

    About the only one of these you *won't* see from a linear regulator is the ripple from the switching regulator (because it doesn't have one).

    The intensities of all of these will differ between regulator types but also between different designs of each type.

    It's also important to note that while it is pretty clear the a SMPS is prone to radiating noise (both electromagnetically and via the power leads -- especially the input leads), that linear regulators are not immune from doing this.

    A simple bridge rectifier/capacitor combination will draw spikes of current at twice the mains frequency and this can cause the shape of the mains waveform to become distorted. This caues a complex form of "power factor" problem.

    Simple SMPS's have *exactly* this problem, but it is made worse by the fact that there is no transformer (which tends to limit the maximum current). There is a so-called "green" technology which places what is essentially a boost SMPS between the rectifier and the capacitor to charge the capacitor during the full cycle rather than just at it's peaks.
     
  13. KrisBlueNZ

    KrisBlueNZ Sadly passed away in 2015

    8,393
    1,268
    Nov 28, 2011
    A few things to add to Steve's explanation.

    Any DC power source ideally provides a perfectly clean DC voltage. If you connect an oscilloscope across it, you should see a perfectly flat trace, at a vertical position that reflects the DC voltage. Any imperfection on that line is noise.

    Common types of noise are continuous wideband (random) noise, spikes due to interference such as mains voltage noise, short- and long-term voltage fluctuations, and ripple.

    Ripple is a specific type of noise that originates with an AC signal in the power supply. A battery produces no ripple because is no such AC signal. A linear regulator may produce ripple at the AC mains frequency (if it uses a half-wave rectifier) or at twice the AC mains frequency (if it uses a full-wave rectifier).

    A switching supply produces ripple at its switching frequency, which is usually between 30 kHz and a few hundred kHz for high-voltage (mains-powered) supplies, and may be up to a few MHz for low-voltage buck converters like those found on computer motherboards.

    If the switching supply is running into a heavy and stable load and is properly designed, the power supply operates in "continuous" mode and ripple may only be at the switching frequency. In other circumstances, ripple may also appear at other, lower frequencies that are related to the switching frequency. This can produce wideband noise which is very difficult to filter out.

    Also, some switching supplies vary their switching frequency according to loading, so the ripple frequency may not be constant.

    Noise may also be generated by the load and injected back into the power source. For example, a laptop contains several switching converters to provide low voltages for the logic circuitry. These converters produce ripple at their inputs, as well as their outputs.

    Mains-powered switching supplies also generate ripple which would be fed back onto the mains supply if it wasn't for the heavy interference suppression components in the mains input path.

    Switching supplies typically produce several types of noise. As well as ripple at the switching frequency, you will usually see noise at frequencies lower than (but related to) the switching frequency (due to imperfections in the regulation and control method), wideband noise (ditto), and ripple at twice the AC mains frequency.

    This noise is mostly removed by heavy filtering inside the power supply (except on budget units) and its peak-to-peak voltage is normally less than a few percent of the DC voltage (ditto).

    External filtering can be added, typically capacitors across the voltage rails and inductors in series with them. This adds to the filtering in the power supply. This type of filtering has a low-pass filter characteristic, and attenuates higher frequencies the most.

    There are two types of noise that can be produced by a power supply. Differential-mode noise is noise voltage measured ACROSS the power supply's output, and is what I have been talking about so far. Common-mode noise is noise that is COMMON to both wires, i.e. the +12V output and the 0V wire.

    If the noise is on both wires, then what is it measured relative to? It's measured relative to mains earth, or the dirt under your house, which is (more or less) coupled to the floor you stand on, and to you, through all the stray capacitance and resistance in the materials of your house and your clothing.

    Common-mode noise can be responsible for the behaviour you sometimes notice where moving your hand near a circuit or device seems to cause noise to be coupled into it. The noise is actually coming from the power supply. Any capacitance between parts of the circuitry and the environment then effectively injects noise into those parts of the circuitry.

    Switching supplies, especially budget ones, generate huge amounts of common-mode noise. It is insidious and it can't be eliminated in any practical way. The cylindrical blobs you see on laptop charger cables are there to reduce it, and they're effective over a range of frequencies, but there are some applications that switching supplies just are NOT suitable for, such as audio preamplifiers (as was already mentioned), AM radios (medium wave broadcast band, and short wave), sensitive medical equipment, and probably others.

    General electronic applications such as calculators, cellphones, computers, routers, and most hobbyist circuits, are not too bothered by common-mode noise, nor differential-mode noise.

    If you're concerned about your circuit being disturbed by differential-mode noise from a switching supply, add more L-C (inductor-capacitor) filtering, using inductors in the range 1 uH to 1 mH (make sure they're rated to carry the DC current you need) and both ceramic capacitors (1 uF or so) and low-ESR electrolytic capacitors (1000 uF or so). And/or add a linear regulator between the power supply and your circuit.
     
    Last edited: Feb 3, 2013
  14. arg733

    arg733

    89
    0
    Dec 14, 2010
    So the smps won't make a mosfet switch at it's output frequency right?
     
  15. KrisBlueNZ

    KrisBlueNZ Sadly passed away in 2015

    8,393
    1,268
    Nov 28, 2011
    You mean, if you use a switching supply to power a circuit that uses a MOSFET, will it affect the switching of the MOSFET? No; the MOSFET's switching is controlled by its gate-source voltage.

    The noise and ripple coming from the switching supply is relatively small and is just superimposed on the power supply voltage. It doesn't have much effect on normal analogue or digital circuitry. Only sensitive circuitry such as audio circuitry will be affected by it.
     
  16. arg733

    arg733

    89
    0
    Dec 14, 2010
    I mean if i have a push-button that gives 15v from an smps to the GATE of the mosfet, will the noise cause it to switch witch 100khz while i push the button?
    what if it is a logic level gate?
     
  17. BobK

    BobK

    7,682
    1,686
    Jan 5, 2010
    The noise will be in millivolts. So the 15V might go from 14.095 to 15.005. This willl not switch the MOSFET on and off. It would have to swing from 15V down to mayby 3V or less.

    Bob
     
  18. (*steve*)

    (*steve*) ¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd Moderator

    25,387
    2,772
    Jan 21, 2010
    Think of a power source like a road.


    • AC is a road that hills and dips in it, and the dips are all underwater.
    • Rectified AC is like the worst possible corrugations you can imagine.
    • AC that is Rectified and filtered is like driving on the ripple strip.
    • The output of an SMPS is like a roughened road surface. It feels smooth, but you get noticable noise from the tyres.
    • AC that is Rectified, filtered, and regulated is like driving on a smooth concrete road. You can barely hear the tyres on it.
    • A battery is a road made of polished glass. The tires are silent on it.
    If you have an all-terrain vehicle (a light globe) then AC will work just fine.

    If you're driving a tank (a DC motor) then rectified AC will be fine

    If you want to hold a conversation (drive a mosfet), you'll need rectified, and filtered.

    If the other person is a bit hard of hearing (a computer), an SMPS or better is needed.

    If you want to enjoy the full dynamic range of classical music (an audio preamp), perhaps rectified, filtered, and regulated.

    If you want total silence, then perhaps you need to go the glass road (battery)

    On second thoughts, the categories are not quite as helpful as I imagined.
     
    Last edited: Feb 5, 2013
  19. arg733

    arg733

    89
    0
    Dec 14, 2010
    Thanks for the explanation steve i mistakenly thought that the smps goes from 15v to 0v at 100khz like rectified AC :D .
    Thank you.
     
Ask a Question
Want to reply to this thread or ask your own question?
You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.
Electronics Point Logo
Continue to site
Quote of the day

-