5

Is it possible to connect 3 LM317s in parallel to get upto 4A of current output ? I need 4A but I'm not able to figure out any other possibility. Is there any other possibility to achieve this ?

Thameem Abbas
  • 53
  • 1
  • 4
  • Have you considered using a regulator that can provide 4A by itself? This is somewhat of an XY question, instead of asking whether you can connect 317s in parallel, you might want to ask (yourself, first) how you can effectively regulate 4A of current. – uint128_t Apr 06 '16 at 17:12
  • 4
    You might even want to read the datasheet. There's an example in it for getting high current with regulated voltage. – JRE Apr 06 '16 at 17:16
  • Many of the answers here apply to your question as well. – The Photon Apr 06 '16 at 17:16
  • 2
    @JRE, you might want to specify which LM317 datasheet. Not all of them have the same application circuits. – The Photon Apr 06 '16 at 17:18
  • 1
    OP, TI's LM317 datasheet has an application circuit specifically for 4 A output, in figure 23. – The Photon Apr 06 '16 at 17:22
  • 4
    Yes it possible. But is it a good idea ? No For this much current, just don't use LM317 or any other linear regulator. Switched converters have been developed, they come on cheap modules, cheaper than the heatsinks you need for all those LM317s anyway. – Bimpelrekkie Apr 06 '16 at 17:28
  • 1
    Use lm338 it's like lm317 ,but deliver 5amp – Ragesh Sep 12 '17 at 02:13
  • @Bimpelrekkie I mostly agree, unless the OP needs (a) something producing lower noise and (b) something dirty cheap (3 LM317 will cost less than 1 LM317+additional BJTs to deliver the same amperage with the same level of thermal+short circuit protection, and the circuit is simpler to set-up). – LorenzoDonati4Ukraine-OnStrike May 04 '18 at 15:12

2 Answers2

10

Yes, it is possible. Do not connect the outputs directly together but rather put a small resistor (0.5 Ohm or so) between each output and the summing node.

Check: http://www.ti.com/lit/ds/symlink/lm317.pdf 8.3.11 Adjustable 4-A Regulator Circuit on page 15/16.

You can also look at 9.3.13 which generates higher current with a single LM317, using instead NPN's to deliver the extra current.

LorenzoDonati4Ukraine-OnStrike
  • 24,608
  • 5
  • 64
  • 108
jbord39
  • 4,350
  • 1
  • 18
  • 25
  • You meant 8.3.12 on page 17, didn't you? – try-catch-finally Sep 12 '17 at 04:52
  • The circuit you're referring to is now 9.3.11, and seems to be a very different thing than what's being asked. It's a 4A regulated current source, but the OP wants a regulated voltage source. I also came here searching for whether I can build a higher current capacity regulated voltage source by using a number of LM317s in parallel. Can you or someone else please comment on whether I misunderstood something about the "4A" device from the datasheet? – Violet Giraffe Oct 31 '21 at 17:26
  • As an aside, could the same effect be achieved by using a 2A diode instead of 0.5 ohm resistance? The forward drop of the diode could serve the purpose of the 0.5 resistance, and as a corollary - the diode would provide high impedance to any attempt by one 317 to drive the other. – Everyone Mar 27 '23 at 18:28
0

You really need a switcher solution. TI/National has chips that will do what you want and software to make the design a snap. I've used them.

Robert Endl
  • 2,207
  • 12
  • 12