I am building a power supply with an LM7812 regulating one of the outputs.

In order to reduce heat, I added a PC fan (2 wire) to the circuit with the red wire sharing the 12V output and the black wire connected to ground.

Everything works fine until I connect a load (Like a stop lamp drawing over 1A) to the 12V output. When I connect the load, the PC fan stops and doesn't start even after removing the load. The only way I can get the fan moving again is to turn off the power supply and back on again.

I read an article suggesting that the PC fan has a hall sensor in it. Is this what is causing the fan to stop? Is there something I can add to the circuit to keep the fan happily moving even when sharing a load? I have 1 10uf cap on the output.

 

Have you measured the voltage out of the 7812 under heavy loads? If it is falling significantly under 12V, the fan may be shutting itself off and won't reset until power is disconnected and reconnected.

 

No, I haven't measured. I'll check that out tonight. I'll also test using a lighter load. Any suggestions to keep the voltage from dropping?

 

No, I haven't measured. I'll check that out tonight. I'll also test using a lighter load. Any suggestions to keep the voltage from dropping?

The LM7812 has a built-in protection circuit that shuts it down if it overheats (due to heavy loads,) but the output can sag for a while before it shuts down. Although some manufacturers claim as much as 1.5A from the LM7812, I find that it is much more suitable for .5A or less.

The higher the input voltage to the 7812, the more power it has to dissipate. What is the input voltage (under load) you are using? And is the input supply robust enough to supply more than 1A? If not, that may be your problem.

And as usual, a schematic of your circuit would certainly help spotting problems.

 

Thanks Tracecom,

The fan shuts off instantaneously when the load is connected so I'm thinking the LM7812 doesn't have time to heat up. It could be that the lamp draws more than the 1.5A @ 12V? I may need to test with something a little less power hungry.

The transformer that I'm using is rated at 24V @ 2A.

I have another output that is regulated by an LM317 that I am able to get over 1A output with the lamp at around 9-10 V...The fan continues to operate when I'm using the LM317 output.

Sorry, I don't have the schematic available on this computer.

 

I think that connecting the fan to the regulator output is a silly idea. Why? because it's probably going to cause more heat generated than it helps dissipate .
Maybe some more experienced members can suggest a way to connect the fan directly to the 24 volts using pwm as regulation.

Once again ... my opinion

edit*
Maybe using two much smaller 12 volts fans in series and controlling them using pwm

 

The fan shuts off instantaneously when the load is connected so I'm thinking the LM7812 doesn't have time to heat up. It could be that the lamp draws more than the 1.5A @ 12V? I may need to test with something a little less power hungry.

The transformer that I'm using is rated at 24V @ 2A.

I have another output that is regulated by an LM317 that I am able to get over 1A output with the lamp at around 9-10 V...The fan continues to operate when I'm using the LM317 output.

Sorry, I don't have the schematic available on this computer.

If your power input is rated at 24VDC, (it is DC, right?) it could be actually producing 28V or more under load. Whether it's 24 or 28, the LM7812 is having to dissipate a lot of power with a 1A to 1.5A load, which is turned into heat. You need to measure the actual voltage and current between the 24V source and the input to the LM7812, as well as the voltage at the output under load. You may have to decrease the load or beef up the circuit.

I do think the fan is a worthwhile idea if you are going to load the 7812 heavily. With a good heat sink and the fan moving air across it, I believe it will help. As Shagas pointed out, running the fan on the input voltage instead of the output would reduce the load on the LM7812.

And do post your schematic (and some photos) when you can.

 

Since the maximum current through the LM7812 is about 1A, why not add a power transistor to the circuit so that most of the current is through the transistor and not the LM7812 (the LM7812 will still regulate the voltage)?

 

OK, I was able to get the schematic and a photo of the project (in progress)...In the photo, the Regulators from left to right are: LM7805, LM7812, and LM317.

I'll be able to take measurements tonight but here is a link to the transformer I'm using: http://www.radioshack.com/product/i...2703&filterName=Type&filterValue=Transformers

Thanks for all the help with this. Will I need to use a voltage divider to reduce the voltage if I put the fan before the 7812?

Thanks!

 

Thanks for all the help with this. Will I need to use a voltage divider to reduce the voltage if I put the fan before the 7812?

Thanks!

You'll need something to knock down the voltage to 10 - 14V. Conversely, you can get a 24V fan and connect it directly. However, the fan must be tolerant to the peak output of your 24V transformer, which can be as much as 34V.

 

Random comments:

I have never used schottky diodes in a full wave bridge rectifier; they may be fine, but it's new to me.

Measure the input to the regulators; I am afraid it's approaching the limit.

Why are the circuits for the LM7805 and LM7812 different? Both should have a .33μF across the input and a .1μF with a 10μF across the output. I don't think C3 is needed as close as the 2200μF cap is.

How wide are the tracks on the PCB that are expected to carry a couple of amps? It's hard to tell in the photo.

 

Measure the input to the regulators; I am afraid it's approaching the limit.

I'll measure the voltage going into the regulators. The LM7805 and LM7812 are rated to 35V. The LM317 is rated to 40V diff between input and output, so yes it may be close.

How wide are the tracks on the PCB that are expected to carry a couple of amps? It's hard to tell in the photo.

The traces are 0.15" going into the split to the regulators. The remaining traces are 0.07".

Why are the circuits for the LM7805 and LM7812 different? Both should have a .33μF across the input and a .1μF with a 10μF across the output. I don't think C3 is needed as close as the 2200μF cap is.

The reason I have one less cap on the 7812 is it shares the same input as the 7805. I didn't think I needed to duplicate that cap, but if the location of the cap to the regulator is significant, I can certainly add one in there.

Thanks again.