LM2940 Voltage Regulator Question

Discussion in 'General Electronics Chat' started by mikewashere05, Nov 13, 2009.

  1. mikewashere05

    Thread Starter New Member

    Oct 26, 2009
    I have a very simple circuit, a 12v 500mA wall wart going into a LM2940 voltage regulator connected in series with a 130Ω resistor and an LED. When I specified the voltage regulator to order, I gave the part # for the 5v. Case markings indicate 5v (LM2490T-5.0). All signs point to this being a 5v regulator.

    HOWEVER, in my circuit, when voltage is measured from Vout of the regulator to ground of the regulator approximately 10.7v is measured. Current draw should be 5v / 130 ~ 38mA. The datasheet says for 5mA < Io < 1A there is a typical output of 5v (being a 5v regulator and all).

    Why am I seeing 10.7v? I tried this on two regulators, both are the same way. I'm afraid to go further in the project (attaching microcontroller, etc.) without making sure it will only see 5v.

    EDIT: Without the LED I measure 9.6v...
    Last edited: Nov 13, 2009
  2. k7elp60

    Distinguished Member

    Nov 4, 2008
    Do you have the recommend external capacitors connected to the regulator? The 0.47uF in the input and the 22uF on the output. The data sheet says the 22uF is required for stability. Generally the input capacitor(.47uF)is required if the regulator is greater than 6" away from the filter capacitor. The 6" generally refers to the length of the input + wire.
  3. mikewashere05

    Thread Starter New Member

    Oct 26, 2009

    So what exactly was I seeing without the capacitor? Does the DMM read Vp, Vpp, Vrms? Was it just very noisy and the capacitor helped smooth it out? Why is the capacitor necessary and why isn't it just integrated into the IC?

    Just some questions so I know the why behind the how.
  4. SgtWookie


    Jul 17, 2007
    Hard to say, without looking at your output using an oscilloscope.

    You tell us - what did you have the DMM set to read?

    Without input/output capacitors, many monolithic voltage regulators will oscillate. I have a few old Motorola 5v linear regulators in my parts collection; if there is no capacitance on the output, they oscillate at between 3MHz-8MHz; the output voltage varies wildly.

    To ensure stability that it doesn't oscillate. It lowers the resonant frequency to below where it would oscillate.
    It wouldn't fit!

    It is simply not practical to include large values of capacitance on the substrate of an integrated circuit such as a regulator.
    Last edited: Nov 14, 2009
  5. THE_RB

    AAC Fanatic!

    Feb 11, 2008
    LM2940 will give all sorts of problems if you dont have input and output caps both connected, and pretty close to the pins of the regulator. Sometimes they refuse to power up too depending on the nature of your load, or have source voltage issues.

    I have a bag of LM2940 5v here that I just stopped using because I got sick of their fussy operation.
  6. Audioguru


    Dec 20, 2007
    The LM2940 is a low dropout regulator. Most low dropout regulators use a PNP transistor as its adjustable series resistance. This PNP has voltage gain and phase-shift unlike the NPN transistor that has no gain and a small phase-shift that is used in ordinary regulators.
    So low dropout regulators must use a pretty big capacitor at their output to prevent oscillation.
  7. bountyhunter

    Well-Known Member

    Sep 7, 2009
    Meter was peak reading the output oscillations.
  8. bountyhunter

    Well-Known Member

    Sep 7, 2009
    The LM2940 is an LDO (low dropout regulator) which has a PNP pass device. It is a two pole control loop because it drives the load from the collector of the PNP which models as a current source, unlike the NPN regs which drive off an emitter which acts like a pure voltage source. The output cap is required to reduce the bandwidth and the output cap's ESR is required to cancel out one of the loop's poles.

    A cap that large can not possibly be fabricated into an IC.