home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: sci.electronics
- Path: sparky!uunet!usc!zaphod.mps.ohio-state.edu!pacific.mps.ohio-state.edu!linac!uchinews!rover.uchicago.edu!frank
- From: frank@rover.uchicago.edu (Frank - Hardware Hacker - Borger)
- Subject: Re: Voltage regulators in parallel?
- Message-ID: <17DEC199210170717@rover.uchicago.edu>
- News-Software: VAX/VMS VNEWS 1.41
- Sender: news@uchinews.uchicago.edu (News System)
- Reply-To: frank@rover.uchicago.edu
- Organization: Joint Center for Radiation Therapy
- References: <9212170609.AA07808@gaak.LCS.MIT.EDU>
- Date: Thu, 17 Dec 1992 16:17:00 GMT
- Lines: 37
-
- In article <9212170609.AA07808@gaak.LCS.MIT.EDU>, jpexg@gaak.LCS.MIT.EDU (John Purbrick) writes...
- >
- >I want to set up a power supply delivering more current than an integrated
- >voltage regulator (LM78xx if fixed, or LM317 if variable) will provide. It
- >seems valid to parallel as many of the relevant component as I need to get
- >the current; certainly they won't all pass the same current, and the voltage
- >out will drop if the initially-highest regulator starts to overload, but it
- >should settle to some lower voltage as other regulators start to conduct
- >more. It seems that this would apply whether the overload was caused by
- >overcurrent or overheating. And when all the regulators are conducting, the
- >output should be no lower than the lowest output would have been if they were
- >wired individually.
- >
- The technique works well, but you need two things additional.
-
- A/ You need what are called ballast resistors to balance the current
- delivered by each regulator. Typically you should have something
- on the order of 0.5 volts or so drop in each resistor at full
- current. (Note that this requires a regulator that lets you use
- a remote voltage sensing point.)
-
- B/ It's a good idea to add a fuse in the output of each regulator,
- so that if one regulator shorts, it doesn't take the rest out with
- it.
-
- For the case where load was reasonably constant, I would just figure
- each resistor for a 0.5 volt drop, then put a diode between the
- normally grounded pin of the 3-pin regulator. The nominal 0.5 volt
- drop in the diode would make up for the 0.5 volt drop in the ballast
- resistors. For highly varying loads, use an appropriate zener and
- resistor to bias the normally grounded pin of the regulator.
-
- Frank R. Borger - Physicist __ Internet: Frank@rover.uchicago.edu
- Michael Reese - Univ. of Chicago |___ Phone : 312-791-8075 fax : 567-7455
- Center for Radiation Therapy | |_) _ In the first place God made idiots.
- | \|_) That was for practice. Then he made
- "Birthplace of Softball" |_) school boards. - Mark Twain
-