Linear Power Supply - Heat Dissipation Problem

What type of PSU is better?

  • Linear Power Supply

    Votes: 0 0.0%
  • Switching Power Supply

    Votes: 0 0.0%

  • Total voters
    0

Thread Starter

StephanG

Joined May 21, 2017
20
Hi, I have build an Linear Power Supply similar like that on the attached picture.
I use and 230:6x2 transformer that is insulated in epoxy.
Line (socket) that is connected is about 235V-245, 50Hz (EU).
Instead the C1 470u of the pic use 2200u, and instead C3 100n block, I use 100u electrolytic.
The load that is connected is an led strip, that makes less than 200mA.
The transformer is rated for 1A output like the LM7812 and the Diode Bridge.
My problem is that the Transformer and the Volt Regulator (LM7812) get very hot (60-80Celsius).
Transformers in normal condition get warm. But mines get very hot.
My question is: Why the transformer and 7812 get very hot even the load is less then 200mA, and transformer and 7812 are rated for 1A.
Thanks in Forward.

 
Last edited:

bertus

Joined Apr 5, 2008
20,049
Hello,

When you are using a 12 Volts transformer with the 200 mA load, the 7812 would dissipate about 1 Watt.
This should not let it run hot.
The problem is likely C3 that you made 100 uF in stead of the recommended 100 nF.
Now it is possible that the 7812 is running hot due to oscillations by the wrong value and type of C3.

Bertus
 

Thread Starter

StephanG

Joined May 21, 2017
20
Hello,

When you are using a 12 Volts transformer with the 200 mA load, the 7812 would dissipate about 1 Watt.
This should not let it run hot.
The problem is likely C3 that you made 100 uF in stead of the recommended 100 nF.
Now it is possible that the 7812 is running hot due to oscillations by the wrong value and type of C3.

Bertus
OK, will try to change the C3 from 100uF elec to 100nF block. Thanks.
 

bertus

Joined Apr 5, 2008
20,049
Hello,

The value of the buffer capacitor will have hardly influence on the heat dissipation.
With the larger capacitor the ripple will be lower.

Bertus
 

crutschow

Joined Mar 14, 2008
23,324
A 1A RMS transformer must be derated to about 55% of that for a rectified capacitor DC output due to the high peak RMS currents that the rectification generates.
The transformer is thus limited to about 0.55Adc output current, which is still well below your supposed load.

It sounds like the load is more than you think.
Have you measured that actual current load from the LEDs?

Does the transformer get hot when the LEDs are disconnected?
 
Last edited:

Thread Starter

StephanG

Joined May 21, 2017
20
A 1A RMS transformer must be derated to about 55% of that for a rectified capacitor DC output due to the high peak RMS currents that the rectification generates.
The transformer is thus limited to about 0.55DC output current, which is still well below your supposed load.

It sounds like the load is more than you think.
Have you measured that actual current load from the LEDs?

Does the transformer get hot when the LEDs are disconnected?
I measured the LED strip with an multimeter, the result was 182mA.
And the transformer and 7812 are warm after the LEDS are disconnected, but less than connected.
 

Thread Starter

StephanG

Joined May 21, 2017
20
I changed the 100uF capacitor to 100nF block, and the transformer and 7812 dissipate less heat than previously.
 

JackBerg7

Joined May 22, 2017
3
I have about the same situation, my project consist of
1 x 12v dc 2A power adaptor
1 x MT3608 DC-DC up converter
1 x L7812cv
16 x 5mm Piranha super flux mixed colors (Red,Yellow,Green) current load 71mA

The circuit worked fine but the 7812 was overheating with a load of 71mA
The original setup was the MT3608 output to 24v DC feeding it to the 7812.

Then I re-adjust the MT3608 to an output of 14v DC,instead of 24v DC
I did run the circuits for over one hour, the 7812 is now at room temperature.
 

R!f@@

Joined Apr 2, 2009
9,647
I have about the same situation, my project consist of
1 x 12v dc 2A power adaptor
1 x MT3608 DC-DC up converter
1 x L7812cv
16 x 5mm Piranha super flux mixed colors (Red,Yellow,Green) current load 71mA

The circuit worked fine but the 7812 was overheating with a load of 71mA
The original setup was the MT3608 output to 24v DC feeding it to the 7812.

Then I re-adjust the MT3608 to an output of 14v DC,instead of 24v DC
I did run the circuits for over one hour, the 7812 is now at room temperature.
Why would you want to boost a 12VDC supply and then reduce it back to 12VDC to light up some LEDs o_O
Is it because you can ? :rolleyes:
Or you have energy to spare ? :confused:

You can get rid of the boost converter, plus the 7812 and use a CC driver running off the 12V Adapter and light the LEDs. You save space, cost, energy and run the LEDs like it is suppose to.
 

MrAl

Joined Jun 17, 2014
6,478
Hello,

In your poll you only have two choices, the linear OR the switcher. In real life we have another choice though that is better than either one alone, and that is a power supply with a switcher front end and linear back end. That way we get the best of both worlds: a switcher to keep power dissipation low and a linear to keep the voltage smooth and well regulated.

The rule of thumb for heat dissipation and temperature rise for small heatsinks or no heatsink is a 60 degrees C rise for every 1 watt of power dissipation across an exposed surface area of one square inch. That means a package like the 78xx series regulators with no heat sink could get up to 80 degrees C in a 20 degree C ambient with just about 1 watt of power dissipation after some number of seconds of run time.

If there is some part that is defective that could cause a higher than normal temperature rise in different parts of the circuit.
 

R!f@@

Joined Apr 2, 2009
9,647
I did not vote, cause there is nothing to vote for.
I have both, and they are both good for what ever purpose I use them for.
 

JackBerg7

Joined May 22, 2017
3
Hi R!f@@

thank's for the comment, first let me explain,,

my project consist of a small enclosure box, the only power supplies I have are
12v DC, I did connect the 7812 alone to the LED's board, and it's sink about 36mA
and the LED's were not as bright as with a direct 12v DC.

Since I need a regulator on board to add one more circuits I did tried the MT3608 as
a booster, since the 7812 need an higher input voltage than the output.

your question make sense.

I did rework the project board, and it's stable at 71mA and the first half of the LED's are
very bright.





 

R!f@@

Joined Apr 2, 2009
9,647
Hi R!f@@

thank's for the comment, first let me explain,,

my project consist of a small enclosure box, the only power supplies I have are
12v DC, I did connect the 7812 alone to the LED's board, and it's sink about 36mA
and the LED's were not as bright as with a direct 12v DC.

Since I need a regulator on board to add one more circuits I did tried the MT3608 as
a booster, since the 7812 need an higher input voltage than the output.

your question make sense.

I did rework the project board, and it's stable at 71mA and the first half of the LED's are
very bright.
Still it makes no sense.
No matter what, you are getting 12VDC from a 12VDC
The 7812 can supply 1Amps max, while your adapter can go 2Amps. Not relevant to your current load demands but you see what I am trying to say.

How are you running the LEDs. Series parallel combination ?
Piranha super flux takes different Vf for different colors.
 
Last edited:

JackBerg7

Joined May 22, 2017
3
I know it's seem not making any sense, in theory
I'am the first to agree with your reply.

but when I build this project, this is what I found, and for what I have on hand
it took a different turn.

by using a 12v dc to an 7812 direct to the LED's board, I'm loosing half of the brightness

and in the other hand if I'm using mt3608 to the 7812 input from 12v to 14v dc, I gain an extra
36mA, and the Led's board reflect the same brightness.

strange but true.
 

crutschow

Joined Mar 14, 2008
23,324
Well, I would just use a low drop-out regulator to regulate the 12V, but if you are happy with such a kludge circuit, than that's all that matters. :)
 

Tonyr1084

Joined Sep 24, 2015
3,578
From the onset of this post I've noticed something stated:
Hi, I have build an Linear Power Supply similar like that on the attached picture.
SIMILAR? What is yours like? Would you post a schematic of what you're working with. All we know is that you've somewhat copied something you researched - which many of us have done. I'm assuming you've modified yours in some way that has yet to be stated; unless you care to revise your statement and say you've built one exactly like the one pictured. Maybe you ARE saying that. Maybe the only modifications you made was in using different value components. Even then, showing us what you built will help us understand some place you've made a mistake. We all make them. No shame in asking for direction and correction. So please post a diagram of what you've built. If you don't know how to make one, draw a picture of it then take a picture of that and post the picture. We can make sense of a lot of things.

Help us help you. Show us what you've made.
 
Top