Connect with us

Problems with linear voltage regulators

Discussion in 'General Electronics Discussion' started by neilw, Nov 1, 2012.

Scroll to continue with content
  1. neilw

    neilw

    15
    0
    Oct 26, 2012
    I have a circuit powered from a 12v car battery which uses a 5v fixed regulator on board.

    I'm having trouble with low voltage output of the regulator...

    Can someone explain why an LM2940S-5 is only giving an output of 4.8-4.85v whereas an LM340T5 is bang on 5v in the same circuit. The input and output bypass caps remain the same and are within the specified ESR range for the LM2940. I've tried 3 different 2940regs and they all are the same, whereas any 340T5 is rock solid at 5.0v.
     
  2. (*steve*)

    (*steve*) ¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd Moderator

    25,496
    2,837
    Jan 21, 2010
    If you read the spec for the LM2940S-5 you'll find that your output voltage is within spec.

    However it is close to the lower limit, and that tends to be a little suspicious.

    What is the input voltage to the regulator?

    What are the sizes of your input and output caps?

    Are these measurements done under load? If so, what is the load?
     
  3. neilw

    neilw

    15
    0
    Oct 26, 2012
    Input voltage makes no difference - If I run it from a bench supply at between 8 and 20 volts the output stays the same at about 4.85.

    Input capacitor is 33uF and total capacitance on output is probably close to 200uF scattered around the board. However, I've even tried removing capacitors but nothing makes the slightest difference.

    Normal load is about 300mA but it's the same with no load. I even pulled the micro out of the circuit but nothing changes. What I find odd is that I've tried 3 different 2940's (from 2 different manufacturers) and they are all the same, but when I put a good old 340 or 7805 in the circuit it's fine ???

    I should add that I only noticed a problem when the CANBUS went flakey at 1mbps. The transceivers (MCP2551) don't seem to like anything less than 4.9v unless I slow down the bit rate, even though they are spec'd at 4.5v. When the CAN acted up, the last thing I looked for was low supply volts, given that the PIC micro was happy and the CAN would work 100% at 500kbps.
     
    Last edited: Nov 2, 2012
  4. (*steve*)

    (*steve*) ¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd Moderator

    25,496
    2,837
    Jan 21, 2010
    Do you have an output capacitor close t the output of the LM2940?

    Low dropout regulators are notorious for their relative instability and intolerance for people not following the manufacturers recommendations.

    I would place a 0.47uF as close as possible to the regulator's input, and a 22uF (or larger) as close as possible to the regulator's output.

    Is there any reason you're using an LDO regulator here?
     
  5. CocaCola

    CocaCola

    3,635
    5
    Apr 7, 2012
    The LM2940S-5 datasheet suggest a .33uf on the input at at minimum a 22uf as close as possible to the output, the .33uf should be as close to the regulators input as possible too... Try it this way with no other items attached, just the basic circuit, also make sure the ground is solid...

    But, as Steve said it's operating within it's specs according to the datasheet, so it might very well not go up anything...

    *** Steve jumped in front again :)
     

    Attached Files:

  6. neilw

    neilw

    15
    0
    Oct 26, 2012
    I have input and output bypass caps right next to the reg (within 3mm) and the board has a full ground plane. I was careful to use caps that are within the spec as the datasheet uses the word "critical".

    I wanted to use an LDO because I need the circuit to stay alive when the engine is cranking with a low charged battery.
     
  7. CocaCola

    CocaCola

    3,635
    5
    Apr 7, 2012
    Chances are REAL good that if the battery level falls below 7ish (normal 5V regulator dropout) you won't be 'cranking' any longer ;)
     
  8. (*steve*)

    (*steve*) ¡sǝpodᴉʇuɐ ǝɥʇ ɹɐǝɥd Moderator

    25,496
    2,837
    Jan 21, 2010
    If this was a real concern, I'd have a series Shottky diode and a large reservoir capacitor before the regulator. This will maintain the voltage to the regulator through short negative-going spikes. It would probably also provide protection against positive going spikes too.
     
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

-