LDO/Transformer Problem

Discussion in 'General Electronics Chat' started by edd345, Jul 17, 2010.

  1. edd345

    Thread Starter New Member

    Jul 17, 2010
    5
    0
    Hi,

    I'm working on a project for a high power LED controller which is powered by 12V.
    The LEDs run at 12V and the microcontroller system works at 5V so I'm using a 5V LDO to convert the 12V input down for the microcontroller.

    Ok, so I plugged in my circuit using a bench power supply @12V and the circuit works fine, which I'm happy about. Then I switched that power supply for a transformer (like laptop charger or external HDD supply) and my LDO fried.
    I tested the transformer with my multimeter and it's giving out 12V. Is there something I don't know about mixing voltage converters?

    I should probably point out that the transformer is pretty cheap.
    Also, I replaced the LDO and went back to the bench PSU and the circuit is working fine again, amazing unaffected by the fact that the LDO was clearly drawing too much current.
     
  2. hgmjr

    Moderator

    Jan 28, 2005
    9,030
    214
    Please attach a schematic so that we can see your circuit.

    hgmjr
     
  3. edd345

    Thread Starter New Member

    Jul 17, 2010
    5
    0
    Hand drawn is all I've got.

    It's pretty simple, although 2 things did stand out as I drew it:
    The 100nF decoupling capacitor and also the fact that I haven't added the pull-up resistor on the microcontroller's reset pin. I guess a cheap fluctuating power supply could cause problems. Plus I don't really want to try other supplies as I'm low on ICs...


    Thanks for your help.
    [​IMG]
     
  4. hgmjr

    Moderator

    Jan 28, 2005
    9,030
    214
    You appear to be missing a series current limiting resistor in your LED circuit. That is going to probably damage your LED and cause unpredictable behaviour by your circuit.

    hgmjr
     
  5. pragsmike

    New Member

    Jul 17, 2010
    2
    2
    I know this is an "is it plugged in?" question, but... Are you sure the transformer is really producing DC? Some wallwarts have rectifiers (sometimes even filters and regulators) and will produce DC. Some literally are just transformers and produce AC.
     
    edd345 likes this.
  6. edd345

    Thread Starter New Member

    Jul 17, 2010
    5
    0
    Thanks for the replies.

    @hgmjr
    Yes, you're right the LEDs do have current limiting resistors.

    @pragsmike
    I don't have an oscilloscope to look, but I've just done a test with a capacitor and an ammeter. With my bench supply; nothing unusual. With the wall wart transformer the ammeter showed current flow and the capacitor started getting hot, so you're right there must be some AC component.


    Would a simple diode solve this or will -12V cause it to reverse breakdown?

    Actually, if the transformer output is still AC, this will probably stop my microcontroller from functioning?
     
    Last edited: Jul 17, 2010
  7. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    It sounds like the transformer output is AC, which will cause your LDO to melt down and your uC to stop functioning.

    You can use a bridge rectifier to covert the 12VAC to DC. You will need a filter capacitor on the output of the bridge rectifier; several thousand uF's.

    12Vac will become roughly 16V DC once rectified and filtered.
     
    edd345 likes this.
  8. eblc1388

    Senior Member

    Nov 28, 2008
    1,542
    102
    Power up your "transformer"(laptop adapter) separately without connection to your LED and microcontroller circuit.

    Use your voltmeter, with its black lead grounded to a nearby ground connection, try measure the voltage on the outputs, both +12V and 0V connection of the adapter to see if you have excessive voltage present.

    While the adapter voltage output is +12V, sometime there could be voltage as high as 70v riding on output of these adapters with respect to ground.
     
  9. hgmjr

    Moderator

    Jan 28, 2005
    9,030
    214
    So you are saying that the LED assembly that you are using already have a current limiting resistor built into it?

    hgmjr
     
  10. edd345

    Thread Starter New Member

    Jul 17, 2010
    5
    0
    Thanks for the help, it's clearly a problem with the power supply. I just didn't realise that when the box says "Output: 12Vdc" it doesn't neccessarily mean Output: 12Vdc.

    My system functions correctly with a suitable 12Vdc supply, so I'll just find a proper 12Vdc supply for it as opposed to adding an AC/DC convertersimply because of a poor power supply.

    I tried measuring the output 0V and 12V of the supply with a voltmeter and initially they were both just floating relative to ground, so I grounded the 0V ouput and the 12V output showed a constant 12V. Maybe it's not constant, I wish I could see what's happening but I don't have an oscilloscope :(

    And yes my LED array has built in resistors which is why I forgot to draw them in the schematic.
     
  11. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    Keep in mind that if you are trying to use a linear regulator to step the voltage down from 12v (or higher) to 5v, you will get absolutely no benefit from an LDO regulator.

    The only reason to use an LDO regulator is if you have just a volt or so to drop. Going from 12v to 5v will cause a lot of power dissipation in whatever linear regulator that you might choose, and LDO then has no significance. You'll still dissipate far more power in the regulator than you will in the load, which nowadays is a sin.

    Learn about DC-DC converters and switching regulators. They might add some noise on the power rails, but they can be extremely efficient.
     
  12. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    Every LDO regulator I have seen must have a suitable capacitor on the output or they oscillate then melt.
    You forgot to tell us which LDO regulator you have or forgot to look at and post its datasheet.
     
  13. edd345

    Thread Starter New Member

    Jul 17, 2010
    5
    0
    Sorry, I've been busy.
    Also, I apologise, I thought all voltage regulators were LDOs.

    The voltage regulator datasheet is here and I currently only have 0.1uF on it's ouput, not massive but it was supposed to act as a decoupling capacitor for the IC to reduce noise on the power rails.
    http://docs-europe.origin.electrocomponents.com/webdocs/0dbc/0900766b80dbce33.pdf

    But either way the system works, and I know my dc power supply is giving me some ac.
     
  14. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    The 78xx series of regulators require an 0.33uF (330nF) bypass cap from the input terminal to the ground terminal, and a 0.1uF (100nF) bypass capacitor from the output terminal to ground. They must be located as close to the regulator as possible. They should be ceramic or metal poly film. You can add more capacitance if you wish, but you must have these two caps installed. If you do not have both caps in place, the regulator may oscillate at frequencies up in the MHz range, and overheat. A couple of years ago, I did an informal bench test of a variety of regulators, and both Motorola 7805 regulators that I tested without the bypass caps oscillated in the MHz range; the output would not have been usable.

    The 78xx series are not a LDO type. LDO implies that the input to output voltage differential is <1V at near full load. The 78xx series will drop around 2v at light load, and more than that with a heavy load.

    Be aware that the center terminal (pin 2) is also connected to the tab, and they are both ground. This is different from the LM317 positive adjustable or LM337 negative adjustable regulators; with those, the tab is connected to the output.
     
    edd345 likes this.
  15. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    The 7805 is an ordinary regulator with a dropout voltage of about 2V.
    It is not a low dropout (LDO) regulator that has a dropout voltage of 0.1V to 0.5V and needs a certain value of capacitor on its output.
     
  16. pragsmike

    New Member

    Jul 17, 2010
    2
    2
    Actually, you probably do have an oscilloscope, sort of, that will work well enough for this purpose. Google "soundcard oscilloscope".
     
    edd345 likes this.
Loading...