Electronic Design

Bob's Mailbox

Hello Bob: I recently saw a question in Nuts & Volts magazine asking why the three-pin linear regulator package pinouts differ between the positive and negative regulators. The answer was that it prevented the input voltage from going to the regulator if a wrong polarity regulator was inserted in the circuit.

(That is not true. In a few cases, that might happen, but there is no cause or effect or intent. /rap)

I initially thought it might have something to do with what part of the regulator chip is the substrate and ends up as the tab or case of the IC. (Exactly! Most TO-5 or TO-39 \\[hermetic\\] cans have the case tied to the third pin \\[as viewed from below\\], which would be the collector of a transistor. It's hard to buy a TO-5 can where the case isn't tied to the third pin. /rap)

Then I got out the catalog and looked at all the various fixed and adjustable regulator packages. The negative regulators (top views) are arranged GND-IN-OUT or ADJ-IN-OUT, except for the TO-39 "H" package, which is IN-OUT-GND or IN-OUT-ADJ. So much for my substrate theory. (On a TO-220 or TO-202 plastic package, the negative input voltage is the middle pin, which ties to the paddle. The die has to sit on the paddle. This is usually true for TO-92s too. The "negative-input" supply is the "collector" pin, which is often called the collector on a transistor. The middle pin is the "collector" on most plastic transistors. /rap)

The positive regulators are even more varied. The fixed regulators are IN-GND-OUT, while the adjustables are ADJ-OUT-IN. (Correct: The negative voltage substrate of a positive fixed-voltage regulator is the middle pin of a TO-220. The negative voltage substrate of a positive adjustable regulator is the middle pin of a TO-220. /rap) Once again, the "H" packages are an exception with GND-OUT-IN or OUT-ADJ-IN. (Correct: The negative voltage substrate of a positive fixed regulator is at VOUT, and that is the middle pin of a TO-220. /rap)

I once updated a dual-rail linear regulator using TO-220 transistors to one using adjustable LM317T and LM337T regulators. The LM337T dropped right into the B-C-E pins of the pnp negative transistor, while I had to cut and jump tracks to use the LM317T in the npn positive-regulator spot. What is the real reason for the varying pinouts of linear-regulator ICs?


  • Charles Hansen (via e-mail)

  • Pease: Hello, Charles. In general, you are right. The substrate is the most negative bus, which is the pins as I listed them. The thing that looks strange is that on an adjustable regulator, such as an LM317, the ADJUST pin is slightly more negative than the substrate. But that is okay.

    See associated figure


    Dear Bob: I just read your column on "Theme Park Stuff" (electronic design, June 7, p. 18), and I commend you. It has always been a source of sorrow to me that working families spend all their time and money visiting a sanitized Disney "global village" just to buy their kids cheap imported figurines. And it's at 10 times the original price, when they could have actually gone to those places and experienced something--for the same money. (Airfares are reasonable these days, but not that cheap!! Sigh. /rap) But I guess if you insist on safe, protected environments not linked to reality, then that is Disney. But I'd pick the Indian theme park any day.

  • Jeannine Neagle (via e-mail)

  • Pease:I'm with you, Jeannine. NSC's security guys cautioned me that I should avoid any personal travel when I went to India and South Africa. I specifically ignored that and travelled around a lot--but carefully. When I climbed up Table Mountain, which overlooks Capetown, I hiked like a fool--but left my suitcases at my hotel, rather than in my parked car. I experienced no break-ins and no muggings. In fact, all day, I saw no other hikers!! Similar to my hiking around Lady Grey and Barkly East. I met lots of nice people in South Africa (as well as in India), where I also spent several hours hiking around Mysore. But I was nearly mugged (of my camcorder) by an elephant. They forgot to tell her the script!!
  • Comments invited!
    [email protected] --or:

    Mail Stop D2597A, National Semiconductor
    P.O. Box 58090, Santa Clara, CA 95052-8090

    See associated figure

    Hide comments


    • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

    Plain text

    • No HTML tags allowed.
    • Web page addresses and e-mail addresses turn into links automatically.
    • Lines and paragraphs break automatically.