LM317 Variable power supply

[ffrige]

Hi everyone! I saw there are several different projects for power supplies and I'm not sure what the real differences between them are. In particular, what is the difference between using a high current regulator (say lm317 or lm350) and using a low current one (lm723) with external transistors to boost the output current? Which design is cheaper?
Also, what's the easiest way of getting +/- output voltage with the same current requirements? Do I need to double the trafo voltage or can I just flip the input to the regulator with a switch?

Thanks in advance!
ffrige

 

[MP1]

You cannot just flip the input to get a negative voltage if you are using a regulator. You will need a positive regulator for the positive voltage and a negative regulator for the negative voltage. I hope I did not misunderstand what you meant. If you mean a switch to use the same AC voltage off the secondary side of the transformer to go to either of the two regulators, then yes this could be done.
The power supplies which use a transistor to boost current will give you more current output but at a little more expense obviously because you need more parts.

MP

 

[ffrige]

MP, thanks for the answer and sorry if I was confusing. I guess I didn't want to use two different regulators since I don't need positive and negative voltage at the same time. But I need a switch to be able to flip the output somehow. If I can't do that with the ac voltage using only one regulator, can I maybe do that directly on the output? What kind of relay would I need to do this automatically?

Thanks,
ffrige

 

[ante1]

Ffrige,

If you don

 

[MP1]

Yes, as ante has pointed out, if you have a floating supply, you can simply reverse the leads of the output.

MP

 

[trigger]

The higher the power dissipated by the regulator, more the ripples at the output.

It is what I learn from real product design.

So my experience is the input voltage usually 20% more than the specification required in order to give that 20% safety margin. It will not cause much heat generated.

 

[MP1]

Use better filtering and do not put more voltage through the regulator than needed.

MP

 

[theox]

The 317 can also current limit.

This feature could be added to this simple yet elegant circuit.

http://www.electronics-lab.com/projects/power/010/index.html

The math is R=E/I

"The Key feature of the LM317, and its family, is that the regulation voltage between the

View attachment 36414

 

[ffrige]

So... I finally managed to finish my variable power supply, but unfortunately it didn't last long. I used it to drive a DC motor and, as you guys suggested, I simply reversed the output wires to change direction. After just a few switchings the LM317 croaked. I imagine some inductive spikes from the motor were too strong for the poor regulator. (although I thought it had an internal protection...)
What can I do to prevent this from happening again? The thing is that the output from the LM317 does not go to 0V ever. Which means that even reducing to minimum the voltage applied to the motor, this will always go nuts with a sudden input reversion. Do I need to disconnect the motor from the regulator before switching the polarity?

 

[ante1]

Hi ffrige,

You must at least have a diode across the output terminals to protect the regulator from reverse voltage.

 

[deadhead]

i agree,you should use diodes regardless of the application. use a diode with a peak inverse voltage (PIV) of at least twice the max output voltage. 1N4002 or hgher, if you plan on alot of transients and spikes, a schotty diode (SB180) may be a better choice. also the lm317 series and most adjustable regulators do not drop to 0V

 

[ffrige]

Thanks a lot guys. I now put two diodes in parallel, just to make sure that nothing bad happens again...

 

[EnigmaOne]

Thanks a lot guys. I now put two diodes in parallel, just to make sure that nothing bad happens again...

Uh...oops.

...and you're going to be reversing the voltage?

How about output fusing that power supply first?

 

[ante1]

Hi Enigma,

The regulator should also be backwards bypassed by a diode to play it safe. If the voltage is set to 5volts and you connect a cap that is charged to 15volts for example you can blow the regulator. Another scenario could be while running a small motor at high rpm suddenly turn down the output voltage the emf from the motor can kill the regulator. But I am sure you knew this! ;D

 

[EnigmaOne]

I certainly wasn't thinking on that one.

In my mind's eye, I saw the diode being implemented at the motor in the same way as a clamping diode, instead of in the usual manner for regulator input shorting protection.

More coffee is required.

 

[TechnoBoy]

Hey, i'm building the variable power supply using an LM317T as well. I only need the output to be 1.5 - 12V. I've got a couple of questions.

1. Using the 2200uF capacitor in the circuit, would the inrush current blow the 2A rectifier?

2. If i used a bigger capactior, say 4700uF, what would i have to change the rectifier because of a bigger inrush current?

3. What secondary voltage should i get for my transformer, since my output is not going to go above 12v.

4. What size of fuse would i need for the primary side of the transformer?

Cheers people.

 

[audioguru2]

Hi TechnoBoy,
Welcome to our forum.  ;D
Don't worry about inrush current. It is limited by the resistance of the transformer. Even a little 1N400x 1A rectifier has a surge current rating of 30A!

To get a 12V output, the LM317 needs a minimum input voltage of 15V. A 12V transformer would just barely make it, a little more maybe 15V if you could get one would be best.
A 15V transformer would have a peak voltage of 21.2V, which is reduced to about 17.7V by rectifier loss and filter cap ripple. If the rectifiers and cap are really good and your mains voltage is high, the input voltage to the LM317 could be as high as 21.4V
If you short the output or try to power a 1.5V load at 1.5A, the LM317 will heat with up to 29.9W, so will need a huge heatsink with thermal grease, and a fan.

Since it will draw a maximum power of about 40W, use a 1/2A primary fuse for 120VAC, or 1/4A for 240VAC.  ;D

 

[audioguru2]

Hi Um...Me,
A 24V transformer will allow this LM317 project to produce up to 28V.
About 1.5A will be available at high voltage settings, but at lower voltages the IC cuts its max current down to only about 200mA to protect itself.
Therefore consider using a 24V center-tapped transformer (12V-0V-12V). The full secondary winding can feed the rectifier for high output voltage and current, and a switch can select only half the secondary winding (the tap) for lower output voltages and a high current.
Use a transformer rated for at least 2A for the project to provide 1.5ADC output. The extra power from the transformer is used to heat the rectifier and LM317.
Use a big heatsink with thermal grease for the LM317. ;D

 

[ruden]

i already made that thing

instead of buying a rectifier, i bought 4 General Purpose diode, 1N4001

Disadvantage:
      the max out current is very low

 

[Alun]

Which is hardly suprising considering the above is only rated to 1A, the diodes need to be rated to more than 2.12A for an output current of 1.5A, I'd recommend >2.5A though.