Smoothing Capacitor in home made PS

When I recently built a custom HTPC in an old tape deck it made me realize that testing LED's really requires a solid and stable power supply. I am designing and building one now (yes I am a total novice) and would like a sanity check on a few things. Attached is the diagram and the application will be for testing LED's, LCD's, and eventually setting up a laser projector of some type for my stereo system (I think I will use two to three 5 - 8mw lasers at this point so that is what I am trying to design for).

Biggest question is on the smoothing capacitors. It looks to me like, here in the US, if I want to keep the voltage drop under 10% for .5A then I will need 42000uf of capacitance. That seems pretty high considering the power supply I found in a not so old stereo receiver only had two 6800uf capacitors. There is a lot of information on this on the internet but it seems to be more about theory than practice. Any Advice?

Second question, my plan is to do basic AC-DC conversion, smoothing, and regulation in the first half of the PS for outputs that are stable. Then those outputs will feed to manual switches. The switches will alow me to plug into modules that will further refine and smooth the DC power down to the precise power outputs I want. I want to protect the power supply, the modules, and the devices I plug into the final outputs. Should I build the protection in at the output modules since they will provide the final power outputs or in the first half before the power is switched. My thought was the power modules. Also I have assumed that this is overload protection, or is there other terms I should be researching?

If the diagram is too small to read let me know please.

That diagram is incorrect unless you mean 24.4peak VAC (AC voltage is generally stated in RMS values). Actual voltage after rectification for 24.4VAC RMS would be ~34VDC. I'll assume 34VDC after the rectifier, and you can always change the values later. The lm317 will need to have at least 2.5V of headroom to operate properly.

For a fullwave rectifier, the peak-to-peak ripple voltage is: V = I/2fC where I is current, f is AC frequency, and C is capacitance.

For the 24V line, a 10% ripple would require:

V = I/2fC
3.4 = 0.5/(120C)
C=0.5/(2.4*120)
C≈1740µF

Running the numbers for the other two voltages yield ≈2440µF for the 13V, and ≈7300µF for the 5V.

Thanks a million for the formulas and the figures. The VAC I listed is what I read off the multimeter since I could not find a datasheet for the transformer. I figured with no load it would be slightly off but I thought it would only be ~.2v. Then I read it would be a an average of 1.4v drop from the rectifier. Thanks and I will see if I can figure out where I went wrong.

Is the frequency 120 because the full wave rectifier uses both halves of the AC waveform?

The LM317 does a good job of smoothing out a lot of the noise in signals. You can take a look at this copy of the TREAD power supply schematic (attached to message) and adapt it to your needs. It was designed for clean power for audio systems.

Here is an archived link of how to build that power supply.

Great info, looking over it and the faq has answered a number of questions I am having. I have already started a slightly modified design of a schematic I found and will definitely be adding in the filter capacitors at the bridge diodes. LM317 Linear power supply Regulator selector 1.5V,3V,4.5V,5V,6V,9V 1.5A circuit : electronic circuit diagrams Latest, schematics, electronic projects

Thank you for taking time to provide these links.

Note that the LM317 is only good for about 32V input, 34V after regulation is probably pushing it a bit, and it's going to get HOT. And I mean extremely hot.

I do have a number of big aluminum heatsinks to choose from and planned to use one. As well the transformer has a couple of secondarys that I didn't have slated for usage so I will probably work one of them to a fan. Thank you for the heads up.

I want to thank all of you for your help. I have just completed the first circuit I have ever built, beyond putting a resistor in front of an LED on a custom case, and it is somewhat of a success. Instead of having the outputs at 1.5V,3V,4.5V,5V,6V,9V, they are instead at 1.38v,5.04v,7.25v,7.78v,8.72v,10.52v.

Close enough that I should be able to work the bugs out. Thanks again for the information.