POWER SUPPLY DESIGN - part 4

SIMPLE VARIABLE IC VOLTAGE REGULATION

This type of regulation is ideal for having a simple variable bench power supply. Actually I think this is quite important because one of the first projects a hobbyist should undertake is the construction of a bench supply. While a dedicated supply is quite handy e.g. 12V, it's much handier to have a variable supply on hand.

There have been many times I have had to "smoke test" a project. This means I have started out at the minimum voltage on my supply (about 3V), seen nothing untoward happening and then slowly winding up the voltage to say 15V (this would only be where the project design called for 15V).

By the way, this is the recommended way to test a project after having double checked parts placement against circuit drawings and parts placement guide (if any). Always start out at minimum voltage and look for signs of distress among components (usually resistors) THEN progressively wind up the voltage to your project voltage. Again always look for distress among your components.

Now one of the heartaches I have here is that inevitably, to derive the necessary voltage it must originate from the mains supply, either 240V or 120V A.C. I have to be mindful of the fact that many people who read my pages are VERY inexperienced. AND

ELECTRICITY KILLS.

So, if only to potentially save one life, I'm going to wimp out and stick with my original idea of using "plug packs" or also known in the US as "wall warts".

Because you are unlikely to want much in excess of 15V D.C. I have restricted my search to types of plug packs of around about 15V A.C. at something in the region of 1.25 amps. Checking through some suppliers (see below) I have found available what is available at the moment * (see below) for plug packs, meters, regulators etc.

For the purposes of this project I have nominally selected a 16V A.C. @ 1.25 Amp plug pack which comes with bare wire ends. Now you can choose to hardwire the bare ends into the case or use a plug and socket connection. You might buy something different which comes with a plug on the end - in which case buy a suitable panel mounting socket and discard my recommendation of plug and socket below.

The choice is yours but if you discount the initial bit of extra work involved, as well as about $A 1.50 extra expense ($US 1.00), I think it's nice to be able to store plug pack and case away seperately. Of course if your plug pack has it's own plug you will need a mating socket to suit.

Here is the basic circuit:

Fig 4 variable bench supply schematic

HOW IT WORKS

Now the basic theory is this. U1 is a TO-220 style variable voltage regulator (LM317T). It is a complete integrated circuit. Quoting, in part, from my old National Semiconductor handbook:

"The LM317T is an adjustable 3 terminal voltage regulator capable of supplying in excess of 1.5A over a 1.2V to 37V output range. It is exceptionally easy to use and only requires two external resistors to set the output voltage. Further, both line and load regulation are better than standard fixed regulators."

"In addition to higher performance than fixed regulators, the series offerss full overload protection available only in IC's."

FEATURES

* Adjustable output down to 1.2V
* Guaranteed 1.5A output current
* Line regulation typically 0.01%/V
* Load regulation typically 0.1%
* 80 dB ripple rejection

"An optional output capacitor can be added to improve transient response. The adjustment terminal can be bypassed to achieve very high ripple rejection ratios which are difficult to achieve with standard 3 terminal regulators".

- source National Semiconductor Linear Databook 1982 P 1-23

So there it is. A genuine "all singing, all dancing variable voltage regulator IC". All depending upon two resistors, which in our case are R1 and, the parallel combination of R2a and R2b. The maths are:

Figure 4a - Voltage out formula

Assuming we use a 5K linear potentiometer (never log or audio type) as the basis of R2 and call it R2a (real clever stuff this) AND put another resistor in parallel called R2b. For this resistor we will allow a value of 15K or 15,000 ohms. This parallel combination would at the maximum resistance setting of R2a yield a net resistance for our R2 combination (do some sums along here with me folks - do follow the leader on the calculator) of 3750 ohms.

Now for sure and certain, someone will email me and say "howfor this?" Every class has it's dill. If you don't know the answer then either you took short cuts and didn't do the basics or you just plain forgot.

Either way when you do email me about this, I can virtually guarantee I will email myself back to you in return, grab you around the throat and, speak quite severely to you. Enuf said?

Worse still, I will send you a .jpg photo of myself and that has been known to terrorise small children, cause little old ladies to faint with fright and for cows' milk to sour. Don't cross me. Remember, "when all else fails read the directions"

Now back to our regular scheduled program. If we allow R1 to become 240 ohms and substituting the values into the equation above we get:

Which of course is what we set out to do. Other maximum voltages can be accomodated by simply fiddling with value of resistor R2b. There are limits and, R1 remains at 240 ohms.

Other points of interest:

1. If you want all the technical nitty gritty then download the current data sheet. I've already done the searching for you. The .pdf file is 645K. Take that or look at the online version. DON"T ask for the email version - remember we're not manufacturers' and we are imposing a little on National's goodwill - show a little netiquette.

http://www.national.com/pf/LM/LM317.html#Datasheet

2. An input bypass capacitor C2 (0.1 uf) is recommended as close as possible to the "input" terminal. Similarly, R1 should be connected as close as possible to the respective "out" and "adjust" terminals. Likewise C3 and C4 (both solid or tag tantalum) are recommended for improved performance and as close as possible to the respective terminals.

3. C1 is a recommended 2200uf / 35V electrolytic filter capacitor.

4. Diodes D1 to D4 form the bridge rectifier and are rated at 3A. Any voltage rating of 50V and up is OK but 3A is the minimum current rating. You could substitute a dedicated bridge rectifier but once over the 1A rating the next size up is !0A and becoming expensive.

5. Diodes D5 and D6 are for protection in the event either the input or output are shorted. The resistor R1 connects across the anodes of both D5 and D6 and nowhere else. Look closely at diagram. R1 does NOT connect to the "out" line. Similarly C4 connects from "output" to ground and NOWHERE else.

6. The inline ammeter and the voltmeter are optional. You can do away with either or both and substitute your digital (assuming you have one) multimeter onto or through the output terminals outside the case. Largely a question of convenience versus budget. If you use either remember to observe polarities.

7. You still have to buy a few bits and pieces not shown on the list below e.g. R1 240 ohm 1/2 watt resistor, R2b (whatever value you decide) or my 15K, and the 0.1 uF ceramic capacitor.

8. I have no connection whatsoever with the firm Jaycar I have recommended below excepting I have known the owner for over 20 years and have always personally found them very satisfactory to do business with. They have a wide number of outlets throughout Australia as well as New Zealand. As in all cases, use your own judgement. I can only comment as a satisfactory buyer.

CONSTRUCTION OF POWER SUPPLY

Obviously the power supply must have a suitable case. Size is governed by whether you include metering or not. Two side by side meters of the type recommended below, each has a face measuring 58mm(W) X 52mm (H). Then space must be allowed for the on / off switch (I recommend a DPDT type) and the red / black output binding posts. The fuse holder (essential) is mounted on the back along with the input socket. To accomodate all that you need a case of sufficient width and depth. I recommend the Jaycar Cat. HB-5446 or similar, 184(D) X 70(H) X 160(W). All dimensions in mm (inches divide by 25.4) There is a lot of drilling - so take your time.

TIP - when cutting the shaft of the potentiometer (pot) to suit the correct length, insert the shaft into a bench vise and cut with a hacksaw where you have marked it. Never wedge the pot in anything!

As I said before largely a question of convenience versus budget. If you can, try and find a suitable "U" shaped handle to fix to the top of your case in the centre, a real convenience.

Internally, mount your components on tag strips. Haunt your local parts supplier for all parts. You will also need assorted nuts, bolts and washers.

To identify the pins on the LM317T place the component on the table with the pins facing toward you. Any markings will be uppermost. The pin to your left is "adjust", Vout is the centre pin and Vin is the right most pin. The electrolytic capacitor and Tag Tantalums should have a + sign on them. See schematic diagram above - this is important. Some types of electrolytics only indicate the minus or negative pin - be careful!.