Because I needed a variable power supply to add to the tools on my workbench, I first surveyed the commercial market to see what was available. Right off the bat I was put off by the cost of many of the manufactured units. I then began to look for a DIY version that I could build. After some major surfing around the web I cam across the data sheet for the LM-317T variable voltage regulator which had a suggested circuit that looked like it would do the job. I also came across a web site where someone had powered his version of this same circuit using the switching power supply of an old printer.
Looking around my shop I discovered a similar "brick" power supply from a long defunct laptop. I also had most of the components for the data sheet circuit in hand. I needed only a pair of binding posts and a heat sink for the LM-317. At least that is what I thought.
The initial build went very well. For an enclosure I used the case from an old data switch about 2" x 7" x 5" roughly in size. The circuit was built on a generic circuit board from Radio Shack, where I also obtained the binding posts and the heat sink. Once it was all put together I tried it out and it worked quite well except for one flaw. I had used an audio taper potentiometer instead of the linear taper called for in the data sheet. This meant that the adjustment of the voltage was pretty much all crammed into the first 90 degrees of the pots 300 degree swing. Another more esthetic "flaw" was that the laptop "brick" just sat on top of the VPS. This looked "goofy" to my eye so I set about making some changes. One of the changes was to add a second enclosure on top of the first one to house a panel meter that displayed volts and also I decided to wire the on off switch to the AC input to the laptop brick and add a fuse on the AC side to protect against a short circuit. Unfortunately, during the installation of the second enclosure and an attempt to place the laptop "brick" inside said enclosure, a short circuit occured and the "brick" let all of it's smoke out. I "firmly believe" that keeping the smoke in is what keeps the device running :).
For the next attempt, I purchased new components for everything except the binding posts, the enclosures, and a few other mechanical components. My son found me another laptop brick and reminded me that this particular model has three connections in the DC plug. The outer shield is ground, the inner shield is actually +19.5 VDC and the center conductor (very thin) is merely a sensor line that tells the laptop that it is OK to charge the battery. Hmm, maybe I should have thought of that on build one. (A similar power module for a printer probably would not have that third conductor so a printer brick might be the easier way to go.)
Then I did the smartest thing I did during the entire project. I called my friend and fellow ham, Steve, KG4LJB, who has done some repairs on my radios over the years and asked him to take a look at the circuit to try and figure where it could be improved. Steve agreed so I dropped it off at his place. He promised to get to it in a couple of days and let me know what he found. Next week, the device was back at my house fully functional. Steve had suggested that it might be better to use another similar circuit also in the data sheet for the LM-317. This circuit was the same in concept but added a couple of diodes to keep the circuit being powered by the VPS from dumping energy from its capacitors back through the regulator. An event, by the way, that ws sure to let the smoke out of the regulator.
As I had not included the diodes in my circuit initially, we suspect that the LM-317 was dead almost from the start of the third build. What Steve sent me back was really a whole new circuit. He replaced the 317 with an LM-350 which has a higher amperage rating. He added a pair of diodes to protect the new regulator and added several polar capacitors to further smooth the output from the regulator and even a few polar caps on the input side. Now I had a really good variable power supply.
When the new laptop "brick" arrived I installed it properly this time. I ran the AC cord through a toggle switch so that the entire circuit gets shut down, not just the variable part of it. I also added a 3 amp blade fuse and holder to the "hot" side of the AC feed. Instead of trying to force fit the "brick" inside the enclosure I wisely chose to simply adhere it to the rear of the enclosure using hot glue and keeping its plastic outer case intact.
In the final build I decided to add a second set of connections to the front panel of the VPS. It not only has the binding posts, it also has a set of "Powerpole connectors in parallel with the posts. The second identical enclosure on top of the first one houses a panel mounted meter that reads DC volts and the on off switch. If I had a larger enclosure to begin with all of the components could have been mounted on one level, but I didn't so this was the result. I had planned to mark the specific voltage points on the front panel around the knob for the potentiometer, but since I added the voltmeter I decided to just mark the minimum and maximum limits of the knob sweep and let the voltmeter read out the actual voltage. The VPS is capable of going to 18.1 volts but the meter only reads to 15. Since I doubt that I will go up to 18 volts anytime soon the lower range of this meter will be adequate. The meter will be easy to replace if I need a broader scale. I can always measure the exact voltage at the binding posts on the rare occasion that I might need a voltage between 15 and 18 VDC.
This variable voltage power supply circuit (as rebuilt by Steve) has a range of 1.5 volts to 18.1 volts and a capacity for about 3 amps with a good heatsink on the LM-350. You will find the final circuit below along with some photos of the completed project on this page.
A couple of advisories are in order. This project starts with 120 VAC at one end and puts out a wide range of DC voltages at the other. AC is serious and can be lethal. Be extremely careful around the AC side of the circuit. All I killed was an inexpensive switch mode power supply, you might not be so lucky. Another point that should be made is that sometimes a circuit has to have a little more complexity to be the best possible circuit for its intended purpose. Simple often works but with a little sophisticationit can work much better. And last but by no means least, when you run into a problem, do not be afraid to ask for some help. Steve is a professional with electronics. He easily discovered what went wrong and was able to put the project back on track and significantly improved its performance.