I just built a basic power board for a project that needs a +5V, GND, -5V outputs. Current requirements are tiny (100ma max, and <<10ma normal)

Here is my schematic: (Pretty much straight from the 7805 datasheet)

The problem is that both +5 and -5 lines ripple by roughly .25V. Oscilliscope reading:

I tried adding some electrolytic caps (47uf) to smooth things, but it had almost no effect.

What do I need to do to smooth these lines? I'm using them for an ADC that needs to take accurate readings between 0 and 3.3V, and this is destroying accuracy. I'd like to get this ripple down by an order of magnitude or 2.

That doesn't look like ripple; it looks like a high frequency oscillation. Try some really low inductance capacitors (such as disk ceramics) right at the output leads of the regulator...and the input one, if necessary.

 

After some further thought I realize that the circuit in my post #7 is likely the simplest way to do what the op wants. Since the regulators will either sink or source current there is no limitation on the positive load current relative to the negative load current (within the regulator's current/power limits) as I originally thought.

 

Re the OP's circuit, you could replace R1 with a piece of wire, change C2 and C4 to 47uF, and put a 56 ohm resistor across C4.

After that it should do more or less what the OP wanted.

It also might need a large input cap, say 1000uF 25v across V1. Some "DC" plugpacks have a rectifier but no smoothing cap.

 

Re the OP's circuit, you could replace R1 with a piece of wire, change C2 and C4 to 47uF, and put a 56 ohm resistor across C4.
.................

That would appear to work as long as the positive current is not greater than 89mA more than the negative load, and you don't mind wasting 89mA of negative current.

 

Listen to Crutschow.

Your virtual ground is maintained 5 volts above the negative input. If there is ripple on the negative input with respect to your measurement reference, then you will see ripple on your virtual ground and as well as +5. Get a 7905 and your problems will be solved.

 

Listen to Crutschow.

Your virtual ground is maintained 5 volts above the negative input. If there is ripple on the negative input with respect to your measurement reference, then you will see ripple on your virtual ground and as well as +5. Get a 7905 and your problems will be solved.

That's not what I suggested. How would you connect the 7905? If you follow the current path from the plus side of the 12V supply through for various load values to the minus side of the 12V supply you will see the problem.

 

You're right -the part that I was thinking of but failed to reference was the fact that the regulator being used for the negative supply needs to have load current than that of the positive regulator. Sorry about leaving that part out.

If the unregulated voltage is split and a virtual ground created from that, a 7905 could be used.

That brings me to one other thought. Svdsinner said that there is ripple on his virtual ground, which leads me to ask "Why is your virtual ground is not the ground you use as a reference?

And while typing, I would point out to Mr. Svdsinner that the use of a positive voltage regulator as a negative regulator is not a standard application for the part.

 

...................
If the unregulated voltage is split and a virtual ground created from that, a 7905 could be used.
...........................

If you do that then the positive voltage will be unregulated. What you would need is a 7810 or 7910 (10V) regulator and split that voltage with a virtual ground.