I've always appreciated the LM317 linear power supply regulator for being a sturdy workhorse. In particular, it's easy to build a simple regulator circuit which "just works" using an handful of external parts: a couple of resistors and caps, maybe a couple of cheap diodes for protection, and you are done with it. I've used it also as a basic building block for lab experiments with my students at school (I'm an high school teacher). It is really difficult to kill one of those IC, even in the hands of the most careless student.

Since I'd like to enhance my lab experiments with some switched mode power supply circuits, I wondered whether there exists a switching regulator IC "comparable" with the LM317 among the metric ton of regulators out there. In particular, my question is whether there is a regulator which has, more or less, the following characteristics:

  • cheap (like the LM317 - ideally less than about 1 EUR retail);
  • through-hole package: it must be suitable for breadboarding;
  • sturdy: it should survive being handled carelessly by students without particular anti-ESD contermeasures during circuit construction and should resist a fair amount of abuse when operated (so it should have short-circuit protections, overtemp protection, etc.);
  • buck, boost and, hopefully, buck/boost mode capable (actually I don't care about the internal topology, as long as the IC can be used both to increase and decrease the input voltage);
  • reasonably wide input voltage range (~1V-20V); it would be nice if it could be used in boost mode using a single AA cell as input;
  • easily adjustable output voltage;
  • 1-2A max continuous load current, so that it could be also used for lower currents (say up to 200-300mA) without heat sinking;
  • reasonable efficiency (>60%) even when used in non optimized designs;
  • few external additional components needed; in particular it should already contain the switching transistor and, if possible, any switching diode (ideally it should contain any component which can reasonably be integrated, leaving only inductors, big caps and settings resistors outside);
  • non-critical typical application circuits: it should be easy to select external parts and achieve stability;
  • external parts should not be critical, especially the inductor (or transformer): I'd like to be able to wind it up myself around a, say, 3C90 large (~30mm) ferrite toroid with (not too many turns of) enameled wire without much hassle (a beginner student should be able to do it without too much trouble);
  • datasheet with clear instructions/formulas on how to select external parts values to achieve desired specs;

Any hint, suggestion and pointer is appreciated!


It turns out the suggestion from Dwayne Reid about the LM2576 family of ICs was spot-on. I happened to come across an inexpensive batch of LM2595 (from ONsemi) and tried it. Almost painless! I bread-boarded a prototype not even selecting a proper inductor: I just wound a dozen of turns on a ferrite toroid I had lying around until I got about the inductance required by the application examples. And I threw in the first electrolytic caps I had at hand (just a bit oversized, to keep ESR down). It worked without problems. Just a bit of oscillations in some load conditions. Once put on a veroboard with proper star-ground and no wires flapping around in the breeze even the oscillations went away.

I meddled with that IC again and also designed a tracking preregulator for an LM317 regulator. Nothing fancy, but just to test the part (and my skills). No problem whatsoever!

Most of my requirements were met. The only downside is it cannot act as a boost converter.

  • 1
    \$\begingroup\$ This sounds like a shopping question, and you're asking a lot of a single part. For any particular switching topology, there are plenty of parts that meet the rest of your requirements. However, the trend is away from through-hole chips for the most part. If you want something that's breadboard-friendly, there are plenty of pre-built "modules" with 0.1" pin spacing that include both the IC and the passive components. \$\endgroup\$
    – Dave Tweed
    Apr 25 '15 at 15:48
  • \$\begingroup\$ @DaveTweed As I said in the question, I'd like to arrange some experiments on switching PS for my students. I'm not an EE designer so I needed an informed advice for selecting possible suitable parts. I don't see why you deem this question is unacceptable whereas there are other questions on this site which ask similar advice and are upvoted, like this \$\endgroup\$ Apr 25 '15 at 16:11
  • \$\begingroup\$ @DaveTweed BTW, I don't need to build a product and a simple DC-DC converter module won't be didactically useful for me, since it hides every detail from the student. \$\endgroup\$ Apr 25 '15 at 16:14
  • \$\begingroup\$ OK, so if you're not an EE, what exactly are you trying to teach your students with respect to these converters? Any simple-to-use IC is also going to hide most of the operating details and design principles. \$\endgroup\$
    – Dave Tweed
    Apr 25 '15 at 16:21
  • 1
    \$\begingroup\$ @DaveTweed I'm an engineer, but not an EE designer, so I've no expertise in selecting a part for a product, especially for an SMPS. Teaching teenagers about electronics in a cost-constrained public school system you have to strike a balance among different "parameters". Explaining buck or boost topologies with simple "inductors and switches" theoretical schemes may do the job up to a point. Then you have to show them some practical circuits, which must be meaningful but not too complex. LM317-based linear regulators are a nice compromise. I need a comparable regulator circuit for SMPS. \$\endgroup\$ Apr 25 '15 at 16:40

Here is a quite short list of TI's Buck/Boost converters which have PDIP packaging.
There are only 2 of 33 DC-DC converters which are available with though-hole package , and the only difference of the listed items is the Operating Temperature Range (C°).
So as you can see and as @Dave Tweed said, through-hole design is not popular.

Let's have a look on the MC33063A's features:

  • Wide Input Voltage Range: 3V to 40V
  • High Output Switch Current: Up to 1.5A
  • Adjustable Output Voltage
  • Oscillator Frequency Up to 100kHz
  • Precision Internal Reference: 2%
  • Short-Circuit Current Limiting
  • Low Standby Current
  • Worst efficiency is 62.2% when using it as a voltage-inverting converter. As a step-up or a step-down converter, it has efficiency over 80%.

The datasheet has a quite good Detailed Design Procedure guide including component selection and typical application schematics.

As for clear instructions, here is a snippet for step-up converter: enter image description here

There are a couple mismatch between the features and your criteria, but I think there always be since there isn't a wide selection of through-hole converters.

  • 1
    \$\begingroup\$ Thanks for the tip! Yep, I knew my "specs" could have been difficult to meet (that's why I said "more or less"): the through-hole constraint is not in line with industry development, but at school breadboarding is essential and there is no viable alternative to it, AFAIK. P.s: I'll wait the canonical 24h before accepting an answer. \$\endgroup\$ Apr 25 '15 at 17:22
  • \$\begingroup\$ I understand your point of the through-hole package, and you can wait as much as you want. I did not do a big research there can be plenty of better answers and components. :) \$\endgroup\$ Apr 25 '15 at 17:35
  • \$\begingroup\$ If you have extra time and enthusiastic students (and some other resources) you can make little boards with pin headers for SMD parts which can be placed into breadboards. Though, it can be a little messy and you have to solder a little too. \$\endgroup\$ Apr 25 '15 at 17:45
  • \$\begingroup\$ Unfortunately PCB manufacture has been almost been eliminated from school and soldering is going the way of the dodo too! This is due to cuts to class hours and lab equipment. Less than 10 years ago I had often 4-6 hours a week available for lab activities (and 3-5 for the theory). Now I'm lucky if lab time reaches 3 hours and 2 for the theory! That's also why Arduino (although on EE.SE seems to be considered almost a hate word by some) here in Italy has saved many electronics courses in schools from extinction (we no longer have the time to build 10+ ICs PCBs). :-( \$\endgroup\$ Apr 25 '15 at 17:58
  • \$\begingroup\$ Then through-hole it is. I guess they do not like Arduino here because it has an independent site. \$\endgroup\$ Apr 25 '15 at 18:06

I would take a close look at the National Semiconductor (now TI) LM2575 family. Very, Very easy to use.

LM2575: 1 Amp buck converter LM2576: 3 Amp buck converter

I would suggest a couple of things, though.

  1. Have a bunch of simple PC boards made up that the students can assemble. Although this SMPS family is very forgiving, grounding DOES matter if you want to have good results.

This then lets your students have a small SMPS regulator that can be used for a variety of projects.

  1. Let the students try breadboarding the regulator and its components on a breadboard. This can be the basis of beginning to understand how important ground paths can be.
  • \$\begingroup\$ Nice part that LM2575! Unfortunately PCBs must be ruled out. A couple of years ago we had to ditch our "school-made" PCB chemical etching system, since it wasn't in line with current Italian safety regulations (no one was ever hurt, but inspectors came in and said it wasn't compliant). We purchased a PCB CNC milling machine but it failed after a year and we are still waiting for it to come back from warranty repair (it's been away for 8 month!). So we have to stick to just breadboarding things (matrix boards are impractical, since they require too much soldering expertise from students). \$\endgroup\$ Apr 26 '15 at 10:53
  • \$\begingroup\$ Why not order boards from one of the low-cost Asian PCB suppliers? You can probably get 1000 boards or more for a few hundred dollars. FWIW - our Chinese PCB supplier reaches extremely low cost when the aggregate total area of all of the boards in the run reaches 2 square meters. For a single-sided board without plated-through holes, that's less than US $200. \$\endgroup\$ May 15 '15 at 18:12
  • 1
    \$\begingroup\$ Sadly any substantial purchase (>50EUR) in public schools must be planned with almost a year in advance and must undergo a rather cumbersome auditing procedure which involves asking at least three independent vendors for a cost estimation for the needed stuff. Then the one who gave the least estimation must be selected. Vendors must comply with some strict regulation which involves (among other things) providing a certificate where it is stated that they have no previous or pending crime issues regarding fiscal frauds or mafia-related crimes. Ugh! It's bureaucracy at the n-th power. \$\endgroup\$ May 15 '15 at 18:43
  • \$\begingroup\$ Most probably (I'm not well versed in that procedure) foreign vendors are excluded altogether (if not by law, by the practical impossibility to comply with all the paperwork, which must be in Italian). I have a colleague that purchased various electronic stuff on eBay with his own money just two months ago (about 100 EUR worth) in order to avoid all the paperwork! The sadder thing is that all this procedures prevents us teachers to strike very good deals with "last-minute" offers on the internet (even from well-known eCommerce sites!). :-( \$\endgroup\$ May 15 '15 at 18:50
  • \$\begingroup\$ I tried your suggestion lately. I edited my question to acknowledge the fact. Thanks! I wish I could accept both answers. \$\endgroup\$ Jun 22 '16 at 19:46

You can start by looking at the MC34063. It is older, but very commonly second-sourced, and there are lots examples and calculators available.. http://www.ti.com/lit/ds/symlink/mc34063a.pdf

  • \$\begingroup\$ And basically what every cheap ""1 Amp"" usb car charger uses. \$\endgroup\$
    – Passerby
    Apr 26 '15 at 1:17

Not the answer you're looking for? Browse other questions tagged or ask your own question.