Calibrating linear 13.8V powersupply

Status
Not open for further replies.

systemloc

New Member
Hi, I'm trying to troubleshoot a small 13.8V/2.5A regulated power supply I picked up from Radio Shack awhile back. I don't think much of it, and I'm planning on snagging an Astron when I get the opportunity, but I thought playing with this peice would be a nice learning experience.

The problem is that this 13.8V regulated power supply puts out a very stable 15VDC! I cracked it open and mapped out the circuit, which I included below. V-in is a standard chunky transformer, bridge rectifier and large filtering capacitor. These components put out 22VDC. The regulator circuit is a linear regulator, this isn't a switching PSU or anything. I read the handy "Build your own power supplies" book published by Radio Shack, and this circuit doesn't seem to make much sense to me.

I understand that the Zener diode is designed to provide some manner of reference voltage with which to compare the regulated output and produce an error voltage. In the book I read, a secondary transistor is used to amplify this signal and feed it to the base of the primary regulating transistor. In this power supply, though, I see no way in which the regulated voltage is sampled at all! I could really use a brief run down of how regulation is produced in this circuit, and perhaps a suggestion of how to tune the output voltage to 13.8VDC. My first reaction is simply to replace the 2200 ohm resistor with a potentiometer and decrease the resistance. I believe this should leak more voltage to ground and away from the base plate of the regulating transistor and therefore decrease its output voltage. In practice, though, this didn't seem to work having only the voltmeter on the V-out poles (no load).

The regulating transistor is a Motorola branded TIP3055. The secondary transistor is a TIP31C. I'm guessing that the capacitors here are for filtering; correct me if I'm wrong. Also, I have no idea why they put a 27 ohm ceramic power resistor on V-in.

Thanks for any help/insight!

**broken link removed**
 
Last edited:
The circuit you have drawn is a very simple power supply, there is no feedback as in the more usual regulated PSU.

How your circuit works:
The zener diode provides a fixed voltage to feed into the two transistors which are connected as a "darlington pair". The output voltage will be the zener voltage plus the base-emmiter voltages of the two transistors, in this case 0.6 + 0.6 volts.
So the output voltege will be (Vzener + 1.2).

The only way to adjust the output of this thing is to change the zener diode.

JimB
 
I can't see the schematic you posted but to add to Jim B's comments, the 27 Ohm power resistor is probably to discharge the smoothing capacitors when the unit is switched off. This is a common feature as it helps the power supply turn off quicker especially with light loading on it.
 
I updated the schematic with measured voltage drops across most of the components. The Zener had a voltage drop of 16.36V, and the base-emitter voltage of both transistors was as Jim B said, approximately .6 V each.

One question raised is why the V out is not Vzener + 1.2 ?

I can easily change the Zener, there's a Radio Shack down the street that still sells components. I had a bit of enlightenment strike me as to why changing the 2200 ohm resistor wouldn't affect Vout. If the two transistors form a darlington pair, what is the purpose of this resistor, anyway?

I switched formats of my pic as well; can everyone see it ok, now?

**broken link removed**
 
I would say the 27ohm power resistor is more to do with a crude way to limit the current.

I would rip out this crappy old circuit and replace it with an LM317 regulator with a series pass transistor and if you want add current limiting too, or you could save some space and go for the switching regulator; schematic for all of these can be foun on the **broken link removed**. Or if you want to save even more time and money use the **broken link removed** 3A regulator.
 
systemloc said:
One question raised is why the V out is not Vzener + 1.2 ?

Simple answer: Stupidity and lack of thought on my part!

I should have said Vout = Vzener -1.2

systemloc said:
If the two transistors form a darlington pair, what is the purpose of this resistor, anyway?

I dont think that resistor is necessary at all.

As for the 27ohm resistor, I think it is there as a short circuit protector to limit the current through the transistor when the output is overloaded or just plain short circuited.

JimB
 
Jim B: What you have said has proven true. I went ahead and swapped the zener for a new one rated at 12V, and the resultant output was indeed approximately 11V. Unfortunately, Radio Shack did not have a Zener rated for 15V, which should give roughly 13.8V.

I also picked up a board blank and an LM317. I'm thinking of building a regulator using parts from this one and the LM317. The power transistor could be reused and the output of the LM317 could be fed to it in place of a zener. I am unsure if I would still require both of the two transistors in the existing circuit, or if I could replace the existing secondary transistor with the LM317.
 
The 2200 Ohm resistor is kind of necessary. The bases of the darlington pair need conduction paths to ground or they wont turn off easily when needed.

The left hand transistor has the zener, but the right needs the 2k resistor, otherwise reverse leakage current would cause it to turn on. Without it the base is staring into the emitter of the first transistor. Not exactly known for gobbling up stray electrons.

edit- I think you may be able to replace the whole circuit, zener + transistors with the LM317 in "voltage reg" mode. It's a pretty capable chip as long as it has the required heat sinking. The 27 Ohm resistor in series makes me think this circuit doesn't need to supply a lot of power.
 
Last edited:
StopGo said:
The 27 Ohm resistor in series makes me think this circuit doesn't need to supply a lot of power.

The supply is rated for 2.5A. Not sure what you consider a lot of power.
 
If the supply is rated at 2.5A then the "27 ohms" resistor is actually 2.7 ohms. Or else it will begin to drop dead with a load of only 260mA.
 
Measured it; my mistake. It is .27 ohms.

I drew out a quick schematic of what I intend to build. Anyone care to comment? The secondary transistor here is the LM317. Is there any problem re-using these caps, and is there a problem using a 220 ohm resistor between Vout and Adj instead of 240?

**broken link removed**
 
Last edited:
I built the above, but for some reason, I can't get the voltage out of the LM317 that it claims to provide. Even when replacing the potentiometer with a 10K resistor, I can still only get 8 volts out of the LM317's Vout.

edit: my mistake... 2200 ohms is not 220 ohms... least I didn't burn anything up
 
Last edited:
Check the datasheet for the LM317 to make sure you have its pins correct.
Double check that the "27 ohms" is 2.7 ohms because only 1A through 27 ohms is a loss of 27V which is not available.

The LM317 with 220 ohm and 5k resistors should try to make an output voltage of 29.7V but with the 22V input its max output will be about 20V.

If the collector of the output transistor is shorted to ground instead of being connected to the positive supply then it becomes an 8V zener diode.
 
The OP's original problem is that the zener voltage is too high for his purpose and he wants adjustability.

For this, he should keep the original circuit but replace the zener using TL-431 and some resistors which then acts as a zener with adjustable zener voltage.
 
Ahh, yes, I see. The TL-431 would have been perfectly acceptable to replace the zener and tune the voltage. Had I been made aware of it, and it been available, I probably would have used it instead.

As it stands now, I rebuilt the entire circuit already around an LM317, and the circuit performs quite well under load.

audioguru: That resistor was in fact a .27 ohm resistor, which shouldn't cause me any sort of problem. I haven't bothered to update all of the pictures. The source of my initial trouble with the new circuit built around the LM317 was my inability to tell a 220 ohm resistor from a 2200 ohm resistor. On using the correct resistor to set the LM317's voltage, everything worked quite well!

Thanks everybody for the help
 
In my opinion, this thread nicely represents what might be expected of the original poster and contributors. The problem/project was not terribly complex and the poster described it well. Suggestions were made and the original poster respectfully provided status reports and shared mistakes. It is a nice illustration of a walk-before-you-run effort. Good lessons here, especially for newcomers.
 
The LM317 isn't a transistor, it's an Intergrated Circuit voltage regulator.

The R27 resistor is probably a fusible resistor.

What sort of current are you planning to draw?

You'll only be able to get 1A with the LM317 on its own, but with other componants 3A is easilly possible (see datasheet).

Which circuit are you planning to build?
 
Hero999 said:
You'll only be able to get 1A with the LM317 on its own, but with other components 3A is easilly possible (see datasheet).
Yeah, look at its datasheet. It givs a lot more than only 1A.
You must be thinking of the 78xx regulators.
 

Attachments

  • LM317.PNG
    9.5 KB · Views: 173
If RadioShack rated it at 2.5A then it can probably give an output of about 1A continuously.
 
Status
Not open for further replies.
Cookies are required to use this site. You must accept them to continue using the site. Learn more…