Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

DIY 30V 5A Power Supply

Status
Not open for further replies.

Voltz

New Member
Im making this my first project, a power supply to power all my breadboard projects, It has to be variable, 0-30VDC and provide up to 5A, - so far I know I need


  • 1 x 220000uF (nearest value to 138000uF) for 3V Ripple at 5A
    1 x 230VAC to 30VAC Transformer
    1 x Bridge Rectifier
    1 x LM723C Variable Regulator (It can do above 10A apparently with an external Transistor [But I don't know what that means :confused: and 0-37VDC although I only need it to do 0-30VDC)
    1 x Red LED (Just to tell when the power supply is on)
    1 x 5A Fuse (I'm using an auto circuit break)
    1 x 275V Varistor (For Surge Protection)
    Some Resistors (I'm struggling with the wiring diagram for the LM723C)
    Some Capacitors (Same as above - for the Voltage Regulator)
    A Potentiometer to control the output Voltage
    A Chassis
    A Voltmeter
    An Ammeter

My main issues are
1.) Where can I get a capacitor that big
2.) What's an external Transistor for the Voltage Regulator
3.) The Wiring diagram for the Regulator

I'm also planning to use a Rotary Switch to change the output waveform from AC (direct from the transformer), DC (Rectified, Smoothed and Regulated), Square Wave (Same as DC but the output run through a 555 astable Timer) and Sawtooth (555 Timer again) - both 555 Timer circuits can have variable capacitors to change the frequency
Im working on a schematic in Eagle but its proving tough - any advice?
 
The capacitor looks like overkill, 47,000µF should reduce the ripple to 1V when 5A is drawn but we don't know the ripple requirements because you haven't posted a schematic.

Please post a schematic for the project you're building.
 
The capacitor looks like overkill, 47,000µF should reduce the ripple to 1V when 5A is drawn but we don't know the ripple requirements because you haven't posted a schematic.

Please post a schematic for the project you're building.

Im wirking on it - but how did you arrive at 47000uF?
 
For 1V of ripple the general rule of thumb is 10,000µF per Amp of current drawn.

For 2V of ripple it's 5000µF/A

The basic formula is:

C = 10000×I/V

C is in µF, A in amps and V is the maximum ripple voltage.

Your ripple requirements might be different so I don't know.
 
A 30V, 5A linear regulator will require a huge heat-sink for the pass transistor, since the worst-case power dissipation will be 150W.
 
For 1V of ripple the general rule of thumb is 10,000µF per Amp of current drawn.

For 2V of ripple it's 5000µF/A

The basic formula is:

C = 10000×I/V

C is in µF, A in amps and V is the maximum ripple voltage.

Your ripple requirements might be different so I don't know.

I re did my calculation (after discovering I used the wrong units in my figures) My new value needed is 13833uF - closest value is 15000uF which is MUCH easier to find so that's sorted, now what's an external transistor?
 
Moyo, there are many circuits on the net for the 723, even if you go to the manufacturer's web site there are application ccts...

but to answer the question, the external transistor is the device connected externally to the 723 to provide the power..the 723 alone isnt man enough, basically is connects to the 723's output pin (base) the collector is conneted to the supply +ve and the output is the emitter connection via a small value current limitting resistor that provides current limitng for the supply...A typical transistor for this would be a 2N3055 or two even for that amount of power, you may need a driver transistor stage to precede the 3055's ( a BD139 maybe)..but the webcircuits will help...if not just ask..:)
 
Last edited:
A 30V, 5A linear regulator will require a huge heat-sink for the pass transistor, since the worst-case power dissipation will be 150W.

A good way to get round that problem, is to use a centre tapped transformer and take the power from the tap at lower voltage settings.
 
A good way to get round that problem, is to use a centre tapped transformer and take the power from the tap at lower voltage settings.

Heat won't be an issue, I'm fine with having massive heat sink and the Power supply with be served by two 80mm 12V Fans

As for the L723 - I've found the wiring diagrams but the chip doesn't seem to be in Eagle :/
 
Last edited:
It's not just heat, it's also a good design practice.

Good lab power supplies have relays that turn various taps every 5-10V. Wasting power in the regulator produces heat, reduces efficiency, reduces stability and reduces the lifespan of the product.

Not a big deal for a couple of amps. 30V @ 5A and it starts to become a problem.
 
Agreed, it'll also save you money, when it comes to the heatsink and fans.

How do you plan to go down to 0V?

The datasheet says the minimum voltage is 2V.
https://www.electro-tech-online.com/custompdfs/2010/05/LM723PDF.pdf

I'm not very familiar with the LM723, there's probably a way to get it down to 0V but I don't know about it.
 
Let me tell you one thing.
You have to be an expert to use a power supply capable of delivering more than a few amps.
The reason is simple.
You will be powering and testing equipment capable of drawing 3 or more amps and this means the cirucit will be quite low impedance.
Furthermore, any fault that develops will involve a lot of damaged components and may even damage trackwork.

Even a 3-amp power supply is not for beginners and if you are having trouble designing and building this type of supply, you should be looking at smaller and simpler projects to work on.
If you are adamant about requiring 5 amps, the simplest solution is to buy some 12v 1.5amp-hr batteries for $15.00 and use them as a supply. You can then add a single strand of wire from a length of flex to act as a fuse.
 
Last edited:
Any power supply can damage components if the circuit is incorrectly assembled. Try connecting an standard 5mm LED to a power supply with a current limit of 100mA and see how long it takes before it burns out.

I'd advise adding an adjustable current limit to any bench top power supply over an Amp or so.
 
**broken link removed** a simple 30V, 5A supply. It uses an LM338 which adds short-circuit and over-temperature protection. Note the large heat-sink.

It's minimum output voltage is 1.3V but that's generally low enough for most experimental requirements. If you really need to go to zero there is a technique shown in the National LM338 data sheet that biases the device from a negative voltage reference.

Edit: There is also a circuit at the end of the data sheet that shows how to provide a current limit by adding a transistor and a shunt resistor. To make the limit adjustable you could switch in different values of shunt resistors or use a power rheostat for the shunt resistor. The shunt resistor must dissipate about 3.5W at 5A.
 
Last edited:
That won't be able to provide 5A at lower voltages because the LM338 will limit the current when the safe operating area protection kicks in.
 
Last edited:
**broken link removed** a simple 30V, 5A supply. It uses an LM338 which adds short-circuit and over-temperature protection. Note the large heat-sink.

It's minimum output voltage is 1.3V but that's generally low enough for most experimental requirements. If you really need to go to zero there is a technique shown in the LM338 data sheet that biases the device from a negative voltage reference.

Thank you very very much, looks much easier to wire than a 723

At the moment, I'm thinking styles - Either Old School or Techy
If I do Old School the case is metal and everything is analogue
If I do techy the case will be plastic and everything is digital, I just need to know how to make a potentiometer digital, being the lazy kind I was thinking a motor to twist the nob connected to a forward and backwards button (+ and -), it sounds stupid though and I'd really like to learn how to make a digital potentiometer
 
A 30V, 5A linear regulator will require a huge heat-sink for the pass transistor, since the worst-case power dissipation will be 150W.

Hi Carl,

the size of a heatsink can be reduced significantly using the proper heatsink.

The LAM4100 has a thermal resistance of 0.7K/W using a 12V fan. It decreases further using a 24V fan.

Here's a datasheet.

Regards

Hans
 

Attachments

  • LAM4100.pdf
    980.1 KB · Views: 517
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top