LM317 power supply issue

Discussion in 'Power Electronics' started by jm-a, Jul 2, 2016.

  1. jm-a

    Thread Starter Active Member

    Oct 20, 2010
    105
    2
    Hi all,

    I'm building a power supply with LM317 and 2N3055.
    Schematic is safe, because build this pcb for friends of mine, and it works; not the same for my example :(:(!!!!!!

    So, in short:

    Bridge output is 16,6V Correct up to now.

    About Vref Measure points A B

    Must always be 1,25V, if i understand LM317 datasheet.

    For Pot 4.7KOhm value = 0 read 1,23 V and Output voltage - Measure points CD 14,1 decreasing to 2.1V

    Pot 4.7KOhm value = 4.7KOhm read 0,68 V and Output voltage - Measure points CD 16,2V


    See Measures-Points.pdf.


    Can i replace LM317 with LM338, in this circuit?

    Any idea for anything to replace, and how to know right one?

    Thanks a lot guys and nice day.

    jm
     
  2. AlbertHall

    Well-Known Member

    Jun 4, 2014
    1,956
    387
    Nothing wrong with those measurements :)
    With the pot set to 4.7K the circuit is trying to get an output voltage of around 26V which it can't do with a 16V input. Set the pot around 2k and you should get about 12V
     
    jm-a likes this.
  3. jm-a

    Thread Starter Active Member

    Oct 20, 2010
    105
    2
    Hi Albert,

    Thanks a lot for your fast reply.

    So, Vref must be 1,25 V with Pot 4.7KOhm set to 0 KOhm, and lower like 0,7V with increasing Pot value ????

    I'm wrong or right?

    Thanks a lot.

    jm
     
  4. dannyf

    Well-Known Member

    Sep 13, 2015
    1,825
    364
    It is not exactly a power supply as the output voltage will vary with load. To fix that, take the top of R1 and connect it to the V+ output pin (top of C3).

    Yes, but you may have to resize the resistors.
     
    jm-a likes this.
  5. AlbertHall

    Well-Known Member

    Jun 4, 2014
    1,956
    387
    Wrong and right!. The pot will vary the output voltage and the ref voltage will remain fixed at 1.25V until the requested output voltage is getting close to the input voltage. At that point the regulator will no longer be able to regulate, the output will increase no further and the ref voltage will fall.
     
  6. jm-a

    Thread Starter Active Member

    Oct 20, 2010
    105
    2
    Hi ,

    Thanks to dannyf and AlbertHall, for their replies.

    jm
     
  7. bertus

    Administrator

    Apr 5, 2008
    15,648
    2,347
    Hello,

    Your circuit is not ideal.
    The transistors will amplify the changes on the output of the regulator.
    Normaly they use a PNP transistor for boosting the LM317.
    In the picture below (taken from the german magazine FUNKAMATEUR 3/2016) , you will see what the PNP transistors does in the circuit:

    Boosting_a_LM317.png

    Bertus
     
    jm-a likes this.
  8. crutschow

    Expert

    Mar 14, 2008
    13,028
    3,237
    If you want to use the power NPN 2N3055 in bertus's circuit (which is indeed the proper way to boost the current of an LM317) you can use it with a small PNP (such as a 2N2907) in a Sziklai pair (sort of a complementary Darlington) to emulate the PNP power transistor.
     
    Last edited: Jul 2, 2016
    jm-a and bertus like this.
  9. ScottWang

    Moderator

    Aug 23, 2012
    4,855
    767
    I'm not sure how much current of 7812 that you will draw, if the draw current of 7812 >= 0.5A then the AC input better using 12Vac-0V-12Vac.

    Vout = 15Vac * 1.414 = 21.21Vdc
    The voltage drop for input/output of 7812 about 3V,
    Vi_7812 = 12V+3V =15V.
    Vi_over = 21.21V - 15V = 6.21V
    W_over = Vi_over * 0.5A = 3.105 Watts
    So the 7812 will waste 3.105 watts on heat.
     
    jm-a likes this.
  10. jm-a

    Thread Starter Active Member

    Oct 20, 2010
    105
    2

    Hi Scott,

    7812 part of this " high-tech " circuit is only to power lcd dmm display and fans ....

    Works perfectly for such a power.

    Thanks for your reply.

    jm
     
    ScottWang likes this.
  11. ScottWang

    Moderator

    Aug 23, 2012
    4,855
    767
    That would be nice.
     
  12. jm-a

    Thread Starter Active Member

    Oct 20, 2010
    105
    2
    Hi all,

    I'm still stuck with this basic circuit :(

    Replaced 12 x 1,1414 -2 V= 15V by 2x12x1,414-2 = 32V for input voltage . 34V displayed ????????????????
    If i read LM317 Datasheet input max is 37-40V . Up to now, right or wrong?

    Thought LM317 was faulty, so desoldered it , put it in a breadbord, a 220 Ohm resistor, and 4.7 KOhm potar, and power supply.
    Works perfectly. Output voltage varied with pot.

    Checked 1N4004 and 220 Ohm resistor: Good

    Checked also 3x 2N3055 with meter and collector insulated : Good


    Now , when i turn 4.7KOhm pot - 0 or 4.7KOhm - my display says 34.0 V !!!!!!!!!!!!!!!!!!!!!!!!!!

    Vref measured on 220 Ohm resistor is 0.00V .

    Now, have you any idea ????


    Faulty capacitors, and if yes, how to check them, without buying LCR meter for in circuit measure?
    Or others components?



    Thanks alot for your help.

    jm
     
  13. AlbertHall

    Well-Known Member

    Jun 4, 2014
    1,956
    387
    Make sure that pot is actually connected from LM317 adj to 0V. Check the resistance between those two points.
     
    jm-a likes this.
  14. jm-a

    Thread Starter Active Member

    Oct 20, 2010
    105
    2
    Thanks albert , i'm going to check this.

    jm
     
  15. jm-a

    Thread Starter Active Member

    Oct 20, 2010
    105
    2
    Check continuity:

    - From adjust LM317 lead and 0 KOhm value for pot 4.7KOhm
    Resistance about 0 Ohm.
    And now ??????

    Thanks a lot.

    jm
     
  16. AlbertHall

    Well-Known Member

    Jun 4, 2014
    1,956
    387
    So, voltage on adj pin = 0V and voltage across R1 is zero, therefore the voltage on the 2N3055 base connections is zero. But the output voltage is 34V? So D2 or one of the 2N3055 transistors is short circuit.
    (all of this assumes everything wired as per schematic)
     
    jm-a likes this.
  17. jm-a

    Thread Starter Active Member

    Oct 20, 2010
    105
    2
    I made a bad measure yesterday, in fact V adj / ground is 34,1 V.:(:(

    What component to check ?????

    Everything wired as per schematic, going to recheck it also.

    Thanks a lot and nice day.
     
    Last edited: Jul 3, 2016
  18. AlbertHall

    Well-Known Member

    Jun 4, 2014
    1,956
    387
    As the pot is connected between adj and ground then we have 34V across the pot. What was the setting of the pot at that time?
    Set the pot around mid-way. Measure all three pins of the LM317 and the output voltage of the supply (i.e. voltage across C3).
     
  19. jm-a

    Thread Starter Active Member

    Oct 20, 2010
    105
    2

    For 0 KOhm :

    V adj/ground = 34.8V
    V output/ground = 35 V
    V input/ground = 35V


    For mid-range - about 2.1 KOhm:

    V adj/ground = 34.60 V
    V output/ground = 35.2 V
    V input/ground = 35.2 V

    Output voltage - measured from an output socket - 34.55V

    Not taken from capacitor, cooper tracks are on bottom side.

    Thanks for your help.
     
  20. AlbertHall

    Well-Known Member

    Jun 4, 2014
    1,956
    387
    For 0 kOhm from adj to ground, you measure 34.8V. One of those two is wrong as you cannot get 34.8V across a short circuit :confused:
     
Loading...