How to limit current flowing through LM317

Discussion in 'General Electronics Chat' started by iinself, Feb 9, 2015.

  1. iinself

    Thread Starter Member

    Jan 18, 2013
    69
    0
    Hi,
    I am to reduce Vin to Vout but my current requirements are higher than the 1.5A that LM317 can supply. Hence I am using a transistor to provide this high current (8A) based on the high current circuit in the datasheet. Since the current is being provided by the transistor I do not want to put a large heat sink on the LM317 also and thus would like to reduce the current through it to a bare minimum enough t drive the base of T1. Will the resistor RL do the job?

    Thanks
     
  2. WBahn

    Moderator

    Mar 31, 2012
    17,726
    4,788
    That topology isn't going to give you very good voltage regulation. Your circuit is simply going to hold the voltage at the input to the base resistor at a set level and as your circuit draws more current the amount of base current will increase and the output voltage will drop by any additional drop across the base resistor (plus the small amount of increase in Vbe of the transistor).

    You want a circuit in which changes in output voltage are servo'ed to increase or decrease the base drive voltage of the bypass transistor.
     
  3. Veracohr

    Well-Known Member

    Jan 3, 2011
    550
    75
    I don't know what datasheet you got that from, it's not the high-current circuit shown in the TI datasheet.
     
  4. ScottWang

    Moderator

    Aug 23, 2012
    4,853
    767
  5. ian field

    Distinguished Member

    Oct 27, 2012
    4,413
    782
    You've got the external transistor in the wrong place!

    You need to develop the Vbe bias across a resistor in series with the 317 input, the collector is in parallel with the 317 output. The orientation of the transistor means it needs a PNP to boost a positive regulator.

    Its fully described in various appnotes and application examples in some of the datasheets.
     
  6. iinself

    Thread Starter Member

    Jan 18, 2013
    69
    0
    I was hoping that since T1 is forward biased, it was going to act as closed switch and all current to the load will be drawn through it. So will this circuit work. Is it right to assume that most of the current is coming from T1 now ?
     
  7. ian field

    Distinguished Member

    Oct 27, 2012
    4,413
    782
    Do it like in the appnote or the regulation will be useless.
     
  8. iinself

    Thread Starter Member

    Jan 18, 2013
    69
    0
    Currently I have a circuit that this that is working to supply in excess of 1.5 A, I found this circuit here - http://www.reuk.co.uk/LM317-High-Current-Voltage-Regulator.htm. The 220R is in the wrong place but with that correction it is working fine. I don't understand the need for 6a4 diode though?
     
  9. iinself

    Thread Starter Member

    Jan 18, 2013
    69
    0
    Thanks.
     
  10. WBahn

    Moderator

    Mar 31, 2012
    17,726
    4,788
    The voltage at the load will be the voltage at the LM317 output minus the IR drop across the base resistor minus the Vbe of the pass transistor. The base current will be the load current divided by the transistor beta. So you have:

    <br />
V_{load} \; = \; V_{317} - \frac{I_{load}}{\beta}R_b - V_{be}<br />

    Hence as the load current changes, so will the load voltage. If beta is 100 and Rb is 1kΩ, then a 10mA change in load current will result in a 0.1V change in load voltage.
     
    iinself likes this.
  11. iinself

    Thread Starter Member

    Jan 18, 2013
    69
    0
  12. WBahn

    Moderator

    Mar 31, 2012
    17,726
    4,788
    Paralleling output transistors like they show is asking for problems with thermal runaway.
     
  13. ian field

    Distinguished Member

    Oct 27, 2012
    4,413
    782
    There is a way to do load sharing in multiple transistors - but not all the appnotes show it.
     
  14. WBahn

    Moderator

    Mar 31, 2012
    17,726
    4,788
    There are lots of ways to do it. The simplest is to just use ballast resistors in the emitter paths before they join up with the other transistors.
     
  15. WBahn

    Moderator

    Mar 31, 2012
    17,726
    4,788
    I'm not sure. I don't see any explanation. My first guess is that it is there for temperature compensation, but that's just a guess. I'd have to look at the tempos for the diode and the transistors to see if they are roughly equal and opposite.
     
  16. ronv

    AAC Fanatic!

    Nov 12, 2008
    3,282
    1,241
    You know how it is.. One bad internet circuit leads to another. You may notice there is nothing to turn on T1. :rolleyes:

    I think the diodes are a misguided attempt to discharge the input cap and the adjust pin cap if the input or output is shorted to ground.
     
  17. WBahn

    Moderator

    Mar 31, 2012
    17,726
    4,788
    I didn't even pay attention to the schematic with the problem of not being able to turn on T1 since I was looking for a schematic that had that 6A4 diode in it and that one doesn't.

    I don't think it's for temperature compensation. I can't find the data, but I don't see anything that indicates that that 6A4 diode has a positive tempco.
     
  18. crutschow

    Expert

    Mar 14, 2008
    13,000
    3,229
    Any scheme with a NPN follower at the LM317 output will have poor regulation and temperature stability. There's no good reason to even consider them.
    Use one of the schemes shown in the data sheet using a PNP transistor booster for a stable output voltage.
     
  19. iinself

    Thread Starter Member

    Jan 18, 2013
    69
    0
    Ok, thanks.
     
  20. iinself

    Thread Starter Member

    Jan 18, 2013
    69
    0
    Thanks that helps a lot.
     
Loading...