Problem with 5v power supply

[bigone5500]

I have built a simple power supply using a 7805CT regulator, a 47uF cap, and a 104 ceramic cap. I also have a red led with a 100 ohm resistor for power indication.

I have it connected as follows:



My problem is when I apply 12v to the regulator it gets hot fast. If I disconnect the LED then it's ok. I have screw terminals for the input and for the output.

*edit* schematic updated.

 

[KrisBlueNZ]

Try redrawing your circuit:

• The 0V rail (negative side of input and output voltages) should run horizontally only, all the way along the bottom of the diagram;
• The regulator should have its input on the left and its output on the right;
• Positive voltages should be higher up in the diagram than the 0V rail;
• Input voltage should be shown;
• Part number for the regulator should be shown.
  
• All text should be horizontal and read from left to right.

Once you've done that, the problem should become clear to you. If it doesn't, repost it here.

 

[bigone5500]

Even though the schematic is not a standard one, it should be clear as to how I have it connected. I can redo it but I do not know what I am doing wrong. I will post an updated one.

*edit* ... LOL! HAHA! I see what did wrong on the schematic. However, I do have it connected right on my PCB.

 

[davenn]

hi Jared

still 2 main things wrong with your schematic
1) you didn't follow Kris's point 2 .... as a result your PSU is back to front ( rotate it 180 deg with input on the left side of the diagram )

2) your LED is around the wrong way

3) not a major problem but ... C2 should be labelled 0.1uF not mF

4) I would up C1 from 47uF to at least 220uF or more


Dave

 

[bigone5500]

hi Jared

still 2 main things wrong with your schematic
1) you didn't follow Kris's point 2 .... as a result your PSU is back to front ( rotate it 180 deg with input on the left side of the diagram )

2) your LED is around the wrong way

3) not a major problem but ... C2 should be labelled 0.1uF not mF

4) I would up C1 from 47uF to at least 220uF or more


Dave

Ok. I fixed it. But as I said before, I have the regulator connected correctly on my PCB. If I take the led out of the circuit, then I get 5.03 volts at the output. The regulator does not get hot. With the led it gets warm fast.

 

[bigone5500]

Here is the board.

 

[KrisBlueNZ]

Are you sure the input voltage is 12V?

Are you sure the LED's series resistor is 100Ω?

Assuming the LED's forward voltage is about 2V the LED current will be about 30 mA. This will cause the regultor to dissipate about 0.2W (if the input voltage really is 12V) which will make it detectably warm but not noticeably warm with that heatsink on it.

 

[bigone5500]

Are you sure the input voltage is 12V?

Are you sure the LED's series resistor is 100Ω?

Assuming the LED's forward voltage is about 2V the LED current will be about 30 mA. This will cause the regultor to dissipate about 0.2W (if the input voltage really is 12V) which will make it detectably warm but not noticeably warm with that heatsink on it.

Input is a sealed lead acid 12v 20AH battery. Resistor is brown black black.

 

[davenn]

is the LED actually lighting up ?

what happens to the output voltage of the reg when the LED and resistor are connected
does it drop significantly or stay ~ 5V ± 0.1V ?

 

[bigone5500]

is the LED actually lighting up ?

what happens to the output voltage of the reg when the LED and resistor are connected
does it drop significantly or stay ~ 5V ± 0.1V ?

LED lights up when connected to the battery. Output voltage is 4.97v.

 

[KrisBlueNZ]

Brown black black is 10Ω

 

[bigone5500]

Brown black black is 10Ω

And with that I will go and hide...

 

[davenn]

Doh ... I didn't even pick up on that

 

[bigone5500]

Doh ... I didn't even pick up on that

HAHA! Neither did I!!!

Heck, I was just testin' ya!

 

[KrisBlueNZ]

You may have damaged the LED too. 300 mA is probably too much.

 

[bigone5500]

You may have damaged the LED too. 300 mA is probably too much.

And I thought it was a blinking LED...

 

[hevans1944]

And I thought it was a blinking LED...

Ho ho ho! Like the little Christmas tree lights? Must be a pretty good LED if it somehow goes into thermal shutdown at 300 mA and then lights up again after it cools off. Oh, wait, you were kidding us about the "blinking," riiiight?

This must be the time of year to replace LEDs. Working late into the evening, I burned out two of them on the same day this week, one because I read the color-code wrong on a 1/4 watt current-limiting resistor, another when I accidentally shorted two wires together on the breadboard, probably connecting the LED directly across the power supply rails. The 12 V 1.5 A voltage-regulated power supply didn't even hiccup, but the resistor was toast, as were the two LEDs.

Good thing they were just two out of a hundred or so I purchased for about five bucks (late in the last century) from Mendelson Electronics, a surplus store here in Dayton. Lessons re-learned: (1) wear the magic magnifying 2X four-diopter OptiVISOR and (2) use spaghetti insulation over those pesky long leads of the resistors and capacitors until they are transferred from the breadboard to a circuit board. <sigh> This was so much easier twenty years ago. Plus, I was allowed to ride my motorcycle then.

 

[bigone5500]

Ok. I got around to 'repairing' the resistor that someone colored wrong. So I put a 100 ohm in and the regulator doesn't get hot anymore. The temp is very negligible. So...I am wondering if it really needs a heatsink.

 

[hevans1944]

... I am wondering if it really needs a heatsink.

That depends on what kind of a load you plan on using. Or did you just want to light up the LED (which I presume it does now)?

 

[bigone5500]

At most it will power an Arduino mega 2560.

And yes the LED is shining nicely.