• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Subscribers 63 subscribers
  • Views 8475 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

LM317 switching regulator

Robert Caruana
Prodigy 100 points
Other Parts Discussed in Thread: LM317, LM3150

Hi, any one ever tried this circuit before or has any experience regarding its operation. Does it function like a switch mode power supply and is it more stable and energy efficient compared to to other common linear schematics?  Any tricks/tips will really help out.

 

 

Input voltage 15v AC 5.6A

Output voltage 9v DC 4A

Thanks a lot for your help. Robert.

 

  • 0
    TI__Guru*** 139846 points

    Hello Robert,

    This is a linear regulator, as the LM317 draws current it turns on the PNP and NPN transistors providing more current for the output. The inductor and diode seem to be pointless. The input voltage needs to be DC , not AC.

    Where did you find this circuit?

    Regards,
    Ron Michallick

     

  • 0 in reply to Ron Michallick
    Prodigy 100 points

    I found the schematic in the National Semi LM317 datasheet. The input voltage is 15v AC, rectified to approx. 13v DC + charging capacitors so actually the input to the regulator will be DC. If you manage to take a look at the datasheet will you please recommend any other type of efficient and stable circuit if this one isn't that good. Also what ESR values and capacitor types (tantalum, ceramic) do you suggest with this or any other schematic to obtain the best results? One last thing, how can I calculate the charging capacitor value as I was going to use 2x 2200uF @ 35v but although the more capacitance the better, I think I don't need that much so I just want to make sure with a calculation. Thanks.

  • 0
    TI__Mastermind 22610 points

    Yes, this circuit works at the rated current, not so much at lower currents. Generic type might be close to  'Hysteretic Buck Regulator'

    The high-current NPN array feeds the switch node on the left of L1, while R4 and R5 provide some hysteresis, and LM317 should be either ON or OFF.

    Here's the best tip I can give you ... Personally, I fail to see any reason to use this circuit for anything. Compared to any modern SMPS Buck Regulator, this 30+ year old circuit is a non-starter: high parts count (especially compared to Simple Switchers); no apps support to get component values specific to your input and output requirements; obsolete components in the BOM; poor efficiency (terrible efficiency by present standards), etc, etc.

    WEBENCH shows more than a dozen SMPS that will work with your input/output requirements. I would suggest LM2678-ADJ or LM22678-ADJ Simple Switcher Regulators as two reasonable starting candidates. The LM22678-ADJ has a slightly wider input operating voltage range (4.5V to 42V) than the LM2678-ADJ (8V to 40V). Both are good to 5A out and have the switching MOSFET built in. WEBENCH (see the WEBENCH tool on right hand side of the Product Folder pages) can calculate the inductor, diode, and capacitors for your specific design, while the datasheets have the needed formulae if you want to do your own calculations.

    http://www.national.com/pf/LM/LM2678.html#Overview

     http://www.national.com/pf/LM/LM22678.html#Overview

  • 0 in reply to Donald Jones
    Prodigy 100 points

    The ICs you mentioned can be directly connected to a bridge rectifier? So the schematic would be a step down transformer + bridge rectifier + charging capacitors + LM2678-ADJ (for example)? Thanks.

  • 0 in reply to Robert Caruana
    TI__Mastermind 22610 points

    With full wave bridge and adequate ripple filtering capacitance you will have DC voltage. Just try to keep the input ripple voltage minimized. Input ripple voltage affects duty cycle, and there is typically a max duty cycle allowed for the device, and/or a point where efficiency begins to seriously suffer. You will need to size the ripple filter capacitors based on current demand and ripple frequency.

    Keep in mind that the SMPS input capacitors calculated by WEBENCH, or called out in the datasheet, for any SMPS are separate from the low frequency ripple capacitors at the rectifier.

    I put Vin(min)= 12V, Vin(max)= 13V, Vout=9V, Iout= 4A, Ambient Temp= 30C into WEBENCH.  From the list of 22 parts generated against these parameters, I would recommend the LM3150 (it's a synchronous controller, so there are two external MOSFETs) mostly because you can purchase a prototype of the design to try out (look for the shopping cart icon in the WEBENCH Tools column), or an assembled/tested evaluation board (pre-adjusted to Vout=3.3V/10A). The LM2678-ADJ is on the list, and you can purchase a prototype of the design, but with an estimated operating temperature near 100C, I would lean towards the controller.

     

  • 0 in reply to Donald Jones
    Prodigy 100 points

    What's the trick to keep the input ripple voltage minimized? Will just charging capacitors do the job? Also do you know of any tutorials (like pdf) on these circuits because I really don't have the knowledge on SMPS, regulators, rectifiers and controllers?   I only know how they work and how to repair them but on the other hand I am not that good in designing one although I'm eager to know. Thanks.

  • 0 in reply to Robert Caruana
    TI__Mastermind 22610 points

    It's the load current discharging the filter capacitor during the time between the voltage peaks (charging from the rectifier) that creates the ripple.

    The simple fix is adding more filter capacitance. Throwing more capacitance at the ripple issue is not usually the best fix, next up could be some type of low frequency pass filter (aka 'pi filter').

    I don't see any specific 'in-house' articles to walk you through this, could be my search terms aren't optimized.

    I'd suggest doing a web search for "Unregulated Power Supply Design" (with the quotes). Several good sites should show-up among the first five or ten hits (after the Sponsored Links).