Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Question about single and dual supplies

  1. Aug 17, 2013 #1

    Whenever reading about op amps, I have come across two terms, 'dual supply' and 'single supply'. Whenever referring to dual supply, I have seen +/- next to the indicated value of the supply, such as +/- 5 volts. Though, whenever single supplies are discussed, there is no negative sign and simply a value such as +5 volts.

    This is a pretty basic question, but does the +/- refer to the fact that dual supply means AC and single supply means DC? And if so, would it mean an op amp that calls for a single supply should be powered at the supply terminals with DC, and dual supply op amps should be powered with AC?
  2. jcsd
  3. Aug 17, 2013 #2


    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    Hi puppetstring
    no, it doesnt. a dual rail supply is one where you have say +5V, 0V and -5V DC rails

    eg.... here's a 12V dual rail supply.... the Ground rail is the 0V rail


    a single rail supply is where you just have say +5V and 0V rails ( this one shows +12VDC) (in this diagram read the - rail as 0V)



    Attached Files:

  4. Aug 17, 2013 #3


    User Avatar
    Gold Member

    Forget AC. It's all DC.

    A dual rail opamp is connected to +5VDC and - 5VDC. It's output can swing from +5 tp - 5 centered on 0. A single rail op amp is connected to +5VDC and 0VDC and its output can swing from 0 to +5V centered on 2.5V.
  5. Aug 17, 2013 #4
    Thank you for your replies. I want to use a 30 volt power supply and regulate it down to various voltages to run six op amps, so in order to use this single supply op amp ( http://ww1.microchip.com/downloads/en/DeviceDoc/21881e.pdf ) I need a voltage regulator that outputs a positive voltage and to run this to Vdd, then run the ground rail to the ground of the power supply?

    And, if I wanted to use this dual supply op amp ( http://www.fairchildsemi.com/ds/L2/L272.pdf ) I would need to get a voltage regulator that puts out positive voltage and run that to Vcc, a regulator that puts out negative voltage and run that to Vee, then what about ground?

    Additionally, because it can only swing to 0 volts, a single supply op amp couldn't be used to amplify an audio signal?
    Last edited: Aug 17, 2013
  6. Aug 17, 2013 #5


    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    yes thats correct

    see my first circuit ... and how a centre tapped transformer is used to create a split rail supply

    not sure if there's a way around that ... some one else may be able to confirm
    But any audio gear I have worked on always used a split rail supply for the pre-amplifier op-amps

  7. Aug 18, 2013 #6
    Using the circuit attached, but instead receiving the initial 30 volts from a variable power supply for prototyping reasons, and using the op amp in this link (L272; http://www.fairchildsemi.com/ds/L2/L272.pdf ) what pin of the op amp would be connected to the ground created by the two zener diodes? I see the pin labeled Vee/GND, so the negative voltage is applied there (in this case -5V for the 3 op amps I will be using) and the positive (+5V)will be wired to Vcc. This is referring to the 8-DIP version

    Also as a side note, what did you do to make the attachments in your post appear in your message?

    Attached Files:

    Last edited: Aug 18, 2013
  8. Aug 18, 2013 #7


    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    hi again :)

    yes you could use a couple of zeners like that but you need to keep in mind that the circuit is
    much more limited in the current that it can supply. Its not really an idea way and I wouldnt
    The other problem with this way is that its a stable form of voltage regulation ... that is...
    if the input voltage changes so will the output. Using voltage regulators as in the above circuits,
    you output voltage is fix and with current limit and thermal protection

    as far as the L272 goes ... with its pin identification ... it appears it can be used for single or dual rail supply

    attachments in post....

    when you open the manage attachments box and upload you file
    then rightclick on the filename and select save shortcut.
    then close the upload box in the row of icons across the top of your text box
    ( visible when you click the Go Advanced Button below the text box)

    put the cursor in the text box to where you want the image to be placed

    select the insert image image ... its below the blue undo arrow
    a small window opens where you can paste the shortcut you saved earlier

    click on OK and you should see the html text appear in your text window

    have a try see how you go :)

  9. Aug 18, 2013 #8


    User Avatar
    Gold Member

    In order to amplify audio with a single rail opamp (it's done all the time) you can use input and output capacitors to block the DC. (look up single rail capacitively coupled op amps) This becomes a problem when you are driving a low impedance since a huge capacitor is required to provide low frequency response. You can use two single supply opamps to drive a speaker if you connect them in bridge mode (each amp driving one terminal of the speaker).
  10. Aug 18, 2013 #9
    It worked, thanks. :)

    I am using a single rail 30 volt variable power supply to test out my circuit, and the power supply does not have a ground, so I am going to create a virtual ground in order to test. The aim of the circuit is to create a color organ, so that LEDs blink and vary in brightness in response to an audio signal. I want to build a circuit so that the voltage from the 30 volt variable power supply becomes +15V, -15V, and 0V, because that is the voltage requirement to run the op amps I am going to use. I need at most 1 amp, so with these specifications the circuit below seems like an ideal candidate. More details on it are in this link. ---> http://powersupply88.com/unipolar-to-bipolar-dc-converter-30v-to-±15v.html


    I would use this circuit to run three op amps, each of which will use +5 and -5 volts to run (each op amp would be provided +5 and -5 volts from two fixed value voltage regulators that have the -15 or +15 split power supply voltage as their input), in total requiring +15 and -15 volts, which could be derived from the 30 volt power supply.

    Each op amp powers a string of 5 LEDs wired in parallel, with each LED having a current limiting resistor in front of it. Each string of LEDs will need at minimum 150mA to run, and can sustain a maximum of 375 mA.

    Because the circuit shown above can supply 1A of current at most, then by my math 3 strings of 5 LEDs (the 5 LEDs in each string wired in parallel to each other) that require at most 375mA of current each should mean that in order to provide each with 375 mA the power supply needs to output 1.125 A of current.

    Here is the main portion of my circuit:

    The input jacks take in the audio signal from a TRS connector, the are filtered into three bands;

    Red = high pass filter

    Green = band pass filter

    Blue = low pass filter

    After that, the op amp amplifies the audio signal so that it has enough voltage and current to power the LEDs

    The diodes attached to the op amps are voltage protection so that the op amps are not driven to saturation.

    From there, the AC audio signal is rectifier into DC so that it can power the op amps.


    And here is how the op amps will recieve power from the voltage that has been changed from single supply to dual supply from the circuit that I first included in this post.


    The voltage regulators are fixed to output 5 volts.

    As for the specifics of what values of the components to use I believe I have worked those out, so I am asking for a critique of the design of the overall circuit.

    My questions then are:

    1.) Does the circuit provided above that splits 30V into -15V, 15V and 0V have any obvious flaws?

    2.) Does the manner that I supply power to the three op amps make sense, and if not what are its flaws?

    3.) Are there any flaws to the over all design of this circuit that would make it extremely wasteful, dangerous or not function?

    I greatly appreciate any information or critiques, because this is my first electronics project.

    Attached Files:

    Last edited: Aug 19, 2013
  11. Aug 19, 2013 #10


    User Avatar
    Gold Member

    Personally I would use a SMPS (switched mode power supply) to create +-5V from 30V and not waste 2/3 the power.

    I'll assume you will choose the proper components for the linear regulators.

    If you trace the path from the opamp to the LEDs, there is no way for current to light up the LEDs. Not sure what I may be missing.

    But, looking at the problem from the top level, there is no need for bipolar drive. You can drive just as well with half wave rectified audio as with full wave. The voltage still tracks the sound level just fine and you don't care about the fidelity.
  12. Aug 19, 2013 #11
    Thank you for the suggestions.

    So instead of splitting the 30V into a rail of -15V and +15V turn the 30V it into -5V and +5V? Ah, I see what you mean, because the same voltage is available to all of the op amps, so only -5 and +5 is necessary. Would it make sense to use the regulators, but instead of have them step down the voltage from +/-15 to +/-5, have them insure that +/- 5V is always available to the op amps?

    Also, by SMPS you mean use one plugged into a wall to create the +/- 5V instead of the one that goes from mains AC to 30V, or build a circuit that uses the 30V from this variable power supply and creates +/- 5V?

    What makes you say the current cannot reach the LEDs, the output of the op amp is sent to the LEDs?

    If the LED is off for half of the period and receives zero volts while the voltage goes negative, how could that accurately track the voltage? Unless taking into account how short the period is for an audio signal and negating the negative half because it has a short duration and it missing would not be noticed.
  13. Aug 19, 2013 #12


    User Avatar
    Science Advisor
    Gold Member
    2017 Award

    Hi puppetstring

    lets sort a few things out :)

    unless you specifically need +15 0 -15 for any of the op amps start with a lower voltage than 30V
    as commented by mebigguy ... its just a waste of power
    of you only need + and - 5V, start with a 15V supply and create + and - rails which will be ~ 7.5V each. This should be OK to feed into the 5V regulators as they need ~ a minimum of 2V difference between the input and output voltages

    you don't need to use 3 other regulators as you show 1 for each op-amp
    you just need 1 x 7805 and 1 x 7905 to supply ALL 3 of the op-amps off the same + and - rails

    look at the path from the output of any of the op-amps to the bridge rectifier ... it goes through the lower left diode of the bridge to ground ... ie. you are basically grounding the output of the op-amp. I don't think the bridge rectifiers are really needed anyway.

    A SMPS ( Switch Mode Power Supply) they are just more efficient than the standard linear PSU using a large transformer.

    I cant comment on the reverse zeners and the other 2 back to back diodes you are showing between the outputs and inputs of the op-amps. I have no idea what you are trying to achieve there ???

  14. Aug 19, 2013 #13


    User Avatar
    Gold Member

    Let's first speak to the LED's flickering. That's actually another short-coming in your design that I did not think of. You actually want to drive the LED's with a true rms detector. Now, the human eye can do that for you but at frequencies below 100 HZ you can perceive flicker. So, if you don't want the bass notes to flicker you need to generate a filtered Average or RMS voltage that drives the LED's. That can be done from full wave or half wave rectification.

    With respect to the SMPS, I was referring to regulating down from 30V. I have no idea why you chose 30V. I would tend to do the whole thing with a 12V or lower wall wart.

    If you want to get fancy, you would use PWM to drive the LED's. Not sure of the best way to convert analog audio to PWM.

    But, getting back to a simple design, if you google "color organ", and look at images, there are a few single ended designs that will work, or at least give you new ideas.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook