I have built a plus/minus 12 volt power supply using a center-tapped transformer (transforms 120 V AC to approximately +12.6/-12.6 RMS), a bridge rectifier, multiple filter capacitors, and two voltage regulators for the plus and minis ~12 volts. I have a switch and a fuse connected with the transformer and power cord.

What can I do to check before I plug in my power cord to the wall? I'm worried something may be connected wrong and kill the socket and circuit breaker.

Aside from making sure the wires are soldered well and no wires are connected together accidentally to make a short, is there another way to check before I test this on the wall?

Thanks!

Howie

 

A very old technique is to put a light bulb in series with the black lead. Worst thing that can happen is the light bulb turns on, and your system, depending on the current it is designed for, may or may not survive. At least, you won't take down the grid in your area. John

 

A very old technique is to put a light bulb in series with the black lead. Worst thing that can happen is the light bulb turns on, and your system, depending on the current it is designed for, may or may not survive. At least, you won't take down the grid in your area. John

Hmm...not sure...black lead, meaning the bulb in series with with the power cord and primary transformer wire? Or you mean the output?

And I think my transformer is rated 2 amps.

 

My comment was based on your comment about killing the socket and circuit breaker. A 2-amp transformer shouldn't do that, if properly connected to the mains. If not, then the light bulb will help. Frankly, I wouldn't worry about the circuit breaker given that situation, unless you have concerns about how the primary is wired.
John

 

Well, the first time I connected it, I put the fuse, switch, and power cord all in parallel. That would...do something bad if plugged in.

Someone suggested to me to use a digital multimeter, and measure all the short circuits and make sure it shows 0 volts...but this means I have to plug in the cord right?

 

At this point, posting a schematic is about the only way for us to be able to see if your stuff is connected properly. But using the light bulb in place of the fuse will limit ant damage from misconnections, and still let the voltages to come up is all is well.

 

Well, the first time I connected it, I put the fuse, switch, and power cord all in parallel. That would...do something bad if plugged in.

Yes, that would do exactly what you are afraid of...

The Switch and the Fuse should be in series in one of the leads on the primary side; the power cord in parallel, as you put it.

Except for obvious shorts, look out for the right polarity of the filter capacitors, and the right connection of the in/out legs on the regulators.

When you say your transformer are rated 2 amps, would that be total on the secondary? or between each secondary winding?
If it's 2x12V, 2x2A on the secondary(which is my guess), then the correct Fuse on the primary side should be about 500mA to protect your transformer.

 

Yes, that would do exactly what you are afraid of...

The Switch and the Fuse should be in series in one of the leads on the primary side; the power cord in parallel, as you put it.

Except for obvious shorts, look out for the right polarity of the filter capacitors, and the right connection of the in/out legs on the regulators.

When you say your transformer are rated 2 amps, would that be total on the secondary? or between each secondary winding?
If it's 2x12V, 2x2A on the secondary(which is my guess), then the correct Fuse on the primary side should be about 500mA to protect your transformer.

Alright, I think I am sure of the polarity for the filter capacitors. I used 4 electrolytic 1000 uF caps. 2 for the input and 2 for the output. 1 for the +12 input side, 1 for the -12 input side, 1 for the +12 output side, and one for the -12 output side. For the +12 input/output side, I put the negative side to ground. For the -12 input/output side, I put the positive side to ground. The other two capacitors on the output are ceramic capacitors I added for noise?

I have used a 1.5 amp fuse on the primary side..just over a little 1 amp...

Here's a crude..making of the schematic I used to make my little less than half of a masterpiece:

Comments greatly appreciated! This is my first...soldering try of a real project.

 

The only things safety related are:
1) The fuse should be the first and only component that is connected to the "hot" wire. As things are now, your fuse is AFTER the switch, which is not good!
2) Reduce the size of your fuse to 1A. You might wish to use a slow-blow type, as charging up the capacitors will put a heavy initial load on the supply.

What are the regulator ICs that you are using? I see "I", "G" and "O", are they fixed 7812/7912 regulators?

 

The only things safety related are:
1) The fuse should be the first and only component that is connected to the "hot" wire. As things are now, your fuse is AFTER the switch, which is not good!
2) Reduce the size of your fuse to 1A. You might wish to use a slow-blow type, as charging up the capacitors will put a heavy initial load on the supply.

What are the regulator ICs that you are using? I see "I", "G" and "O", are they fixed 7812/7912 regulators?

1) Hmmm...I don't see why it's not good. Could you explain?

2) I have used a slow blow-type. Why reduce to 1 amp?

3) I've used an LM1320MP-12 for the negative regulator and a LM7812 for the positive regulator. Are these okay? Also..for both...is it: INPUT-GROUND-OUTPUT start from left to right where left is the first letter of writing? Or are the pins different?

 

1) The fuse is there to protect the user of the device as well as the device itself. As you have things wired right now, the switch is connected directly to the mains, with no fuse protection. If something inside the switch breaks, it may be possible for the internals that are now rattling around inside the switch to create a short to the chassis/enclosure. If the chassis/enclosure is grounded, this will cause maximum current to flow until the breaker trips or the wiring burns up. If the chassis/enclosure is not grounded, this will put 115VAC on the enclosure, which of course is quite dangerous.

I forgot to mention in my prior post that you don't show a ground for your power cord. This is very important. Don't rely on the neutral line to provide a ground. Use a cord that has a 3-prong plug. The green wire is ground, the white wire is neutral, and the black wire is hot.

The green wire should be connected to your chassis/enclosure.
Connect the white (neutral) wire to one side of your transformer's primary winding.
The black (hot) wire connects to one side of your fuse, then from the fuse to your switch, then from the switch to the other side of your transformer's primary winding.

2) If the primary side of your transformer draws more than 1 Ampere, there is something wrong, and you want the fuse to blow instead of damaging something.

3) I can't find a datasheet for the LM1320MP-12. Are you sure that's the correct part number?

Something else - the capacitors on the input side of your regulators are quite small. Consider increasing them significantly. Use caps that are rated for a minimum of 35v.

 

I would have reduced the fuse further...

If the 'model number' of the transformer; 25.2CT are any indication, i would guess it's rated a total of 25V at 2A, which gives 50Watt.

50W at 120V = 0.417A, so a 500mA slow-burn fuse would protect your circuit better. Maybe go a little bit up, if the capacitors makes the fuse blow at startup...

 

Thanks for the great suggestions.

Here's the datasheet for the part I used:
http://www.digchip.com/datasheets/parts/datasheet/321/LM320MP12.php

Seems like it's I-GND-O, just like the 7812.

 

Thanks for the great suggestions.

Here's the datasheet for the part I used:
http://www.digchip.com/datasheets/parts/datasheet/321/LM320MP12.php

Seems like it's I-GND-O, just like the 7812.

OK, but look at the terminal connections on page 8 of that datasheet.

If it's a TO-220 case, then looking at it with the printing oriented so you can read it, the terminals are (L to R) GND, IN, OUT - with IN also being on the tab.

BTW, I hope you have good-sized heat sinks attached to the tabs, and you have used heat sink compound. Otherwise, your regulators will rapidly get hot when you put a load on them.

 

OK, but look at the terminal connections on page 8 of that datasheet.

If it's a TO-220 case, then looking at it with the printing oriented so you can read it, the terminals are (L to R) GND, IN, OUT - with IN also being on the tab.

BTW, I hope you have good-sized heat sinks attached to the tabs, and you have used heat sink compound. Otherwise, your regulators will rapidly get hot when you put a load on them.

Hey Wookie, I was just about to comment on the heatsinks! you read my mind.

I connected a 12V LED to the load and you're right, both regulators get rapidly hot...but at least that tells me I connected everything good together. I am getting at least 12 V.

What can I do? What is a heat sink compound exactly?? I know it's some type of paste, but how do you put it on there and where?

 

Well, actually, only the LM320 neg. regulator gets hot..I must've got the pins wrong. I will try to resolder.

 

If the LM320 is getting hot with just a standard LED (with current-limiting resistor) for a load, you definitely have something connected incorrectly.

Heat sink compound is a very white substance in a paste form. You just need a small amount between the back of the TO-220 case and the heat sink; basically just enough so that a little bit squeezes out all around the junction between the case and the heat sink.

It's a filler material. Both the surface of the component and the heat sink have (on a microscopic level) a rough surface. Much of the heat transfer is performed by direct mechanical/physical contact between the two surfaces. Heat sink compound is no substitute for a good solid physical connection, but it fills in the valleys and makes a better thermal junction.

You can get a small tube of it at your local Radio Shack, if noplace else.

 

Well, I tried using the GND-IN-OUT configuration for the LM320...seems like it still gets hot. I've also blown 2 fuses now. Is it normal for the fuses to seem foggy after running the circuit? Also, is it normal for the transformer to make a little, small buzzing, humming sound?

Not sure what's happpening. I think I'll just try using a LM7912...In the datasheet (page 1) it says the LM320 series needs a capacitor at the output...maybe this is causing me problems? Maybe causing the heat? I have an LM320MP-12 and i don't see this on the datasheet either...all I see are LM320H's..and LM320T's... =\

Well, I'm pretty frustrated.

 

The LM320MP is rated for 7.5W, and has an 0.5A output capability; it's in a TO-202 case which looks similar to a TO-220 case, but smaller. You would've been better off with an LM320T-12, which is 15W, 1.5A in a TO-220 case.

Beware of the heat sinks touching something. Remember that the TO-220 case has the input connected to both the tab and the middle pin. The datasheet does not indicate if the TO-202 case tab is connected the same way, but I suspect that it is.

Try disconnecting the input to the LM320, and see if you get around -18VDC between the negative side of the bridge rectifier and the transformer center tap. Verify that you are getting +18VDC between the positive side of the bridge rectifier and the transfomer center tap. Sometimes, bridge rectifiers are mis-marked (particularly on the surplus market), and it can be confusing trying to figure out what connection is what.

Are you sure you have the capacitors hooked up correctly? Electrolytic capacitors usually have a strip on one side pointing towards the negative lead (for radial caps; ie: both connecting wires coming out of one end) or the polarity indicated on the ends (for axial leads; ie: one wire on either end of the cap).
They need to be hooked up like you have them in your schematic.

If you have electrolytic capacitors connected backwards, they will conduct electricity, overheat, and blow up.

Your fuses should not be getting foggy.

You should have at least a 1uF capacitor on the output. Very often, you will see power supply schematics that have a 0.1uF (100nF) and a larger capacitor, say 10uF to 1000uF, in parallel on the output. The 0.1uF cap gets rid of high frequency noise, and the large cap takes care of low frequency transients.

 

1A * 120VAC = 120W

120W/24V (total PS voltage) = 5 Amps

So that fuse should feed appox 5 Amps output.

Dang, did it again, missed the second page before I postes.

Just my opinion, but you don't need that huge capacitance on the output of the regulators. A small one, that has been recommended, would work just as well.