General propose Bipolar Power Supply

[Fantasma25]

Hello. I've been working recently on some guitar effects pedal design, but most of them use Opamps, and therefore I require +/- 12 or 15v supply voltage. I also don't want to make a transformer PSU for every design since they are heavy and bulky. I want to be able to use any generic unregulated wall AC/DC converter (maybe even a laptop charger) and through some regulators and rail splitting circuits inside the pedal design, feed all the Opamps. Does anybody know any good (and preferably cheap) design that can output about 100 to 200mA?  Thanks!

[Zero999]

A cheap way to do this is to only power it from mains adaptors with an AC output and use two half wave rectifiers to get a bipolar output.

[Trey]

Hi. I know this wasn't your question, but do you really need +/- 12-15V (that is, 24-30V) supply? Perhaps you are working up something unusual, but many commercial guitar pedals work from a single 9V battery or simple ~9V DC wall-wart (note: typically center pin is _negative_). Many modern Op-Amps (eg, TL072) will work well for instrument-level (<2V) signal with a "single" 9V supply -- one just needs to bias an AC-coupled input signal up to the mid-point (~4.5V) which can be obtained many ways. I've designed/built several effect pedals with a simple resistor divider with a capacitor to produce the ~4.5V "ground" if not much current is needed from it (just bias/reference); or used an op-amp follower of that mid-point for more current (rarely needed). Hope this helps -- again, I know it wasn't your original question. Cheers.

[Fantasma25]

Hi. I know this wasn't your question, but do you really need +/- 12-15V (that is, 24-30V) supply? Perhaps you are working up something unusual, but many commercial guitar pedals work from a single 9V battery or simple ~9V DC wall-wart (note: typically center pin is _negative_). Many modern Op-Amps (eg, TL072) will work well for instrument-level (<2V) signal with a "single" 9V supply -- one just needs to bias an AC-coupled input signal up to the mid-point (~4.5V) which can be obtained many ways. I've designed/built several effect pedals with a simple resistor divider with a capacitor to produce the ~4.5V "ground" if not much current is needed from it (just bias/reference); or used an op-amp follower of that mid-point for more current (rarely needed). Hope this helps -- again, I know it wasn't your original question. Cheers.

That is a good tip. I guess I'm used to work with that voltage because of the university labs, but there is really no need for such a high voltage in the signal, as most pedals are not power amplifiers. Nevertheless, I'm currently working on a Mu-Tron III clone pedal and the schematics show that it was used with two 9v batteries (for the +/- rails). I'm planning on using this circuit to generate the +/- rails from a laptop charger, but I'm not sure if it will be able to feed enough current to the opamps.

What do you think?

[Zero999]

Many modern Op-Amps (eg, TL072) will work well for instrument-level (<2V) signal with a "single" 9V supply -- one just needs to bias an AC-coupled input signal up to the mid-point (~4.5V) which can be obtained many ways.

The TL072 isn't specified at voltages below 10V, the graphs on the datasheet and other specifications such as common mode input range and output voltage swing, suggest it will stop working at voltages below 7.5V but it could be higher, depending on how it's used.

For maximum battery life and stability, use an op-amp specified down to 6V or less. The TLC072 is a good replacement for the TL072

That is a good tip. I guess I'm used to work with that voltage because of the university labs, but there is really no need for such a high voltage in the signal, as most pedals are not power amplifiers. Nevertheless, I'm currently working on a Mu-Tron III clone pedal and the schematics show that it was used with two 9v batteries (for the +/- rails). I'm planning on using this circuit to generate the +/- rails from a laptop charger, but I'm not sure if it will be able to feed enough current to the opamps.
What do you think?

That will work but beware the TLE2426 is unstable under certain combinations of load current and capacitance, see the datasheet. Connecting two 4.7uF capacitors, each with 100nF in parallel, from the output to either supply rail, giving a total effective output capacitance of 9.6uF should avoid any problems, as well as reduce the output impedance at high frequencies.

EDIT:
Is the laptop PSU output connected to earth? If so, this could be a problem if you connect your circuit's 0V terminal to another device's 0V with is connected to earth. If this is the case the laptop's negative rail with be 0V relative to earth, leaving your 0V rail at half the PSU voltage relative to earth, which will be short circuited if you connect it to another device who's 0V rail is also connected to earth.

[dannyf]

What do you think?

The rail splitter can be replaced with just another opamp (like ne5532) + resistor pair. If the opamp is wimpy, put a pair of transistors on its output.

Or even a simple 3-terminal voltage regulator of about 1/2 of the rail voltage. For example, in your case, a 5v 7805/78L05 would work.

[Zero999]

If you go for the LM78L05, you need to beware the regulator cannot sink current, only source it, so you need to connect a load resistor from the output to the negative rail, allow it to allow for this.

[Fantasma25]

Thanks for your answers. I'm kinda digging the regulator approach, but I think I'll do it with the opamps at first, as it will make the project more "universal", being able to power it with a more variety of voltages. I'll also try to use low power opamps so that I don't have to use 20+ v AC/DC converters.

EDIT:
Is the laptop PSU output connected to earth? If so, this could be a problem if you connect your circuit's 0V terminal to another device's 0V with is connected to earth. If this is the case the laptop's negative rail with be 0V relative to earth, leaving your 0V rail at half the PSU voltage relative to earth, which will be short circuited if you connect it to another device who's 0V rail is also connected to earth.

Thanks for the tip!! I was not considering that 

[Trey]

As an example, the voltage splitter, input buffer, and drive section of a Tube Screamer TS-808 looks something like this:
.
The design is not ideal as the input impedance is too low, but that would be easily solved by bootstrapping.

If you want to accommodate various input voltages, splitting the rail down the middle is advisable.  Most guitarists I know (I play bass) use 9V wall-wart to their pedal boards, though a few use 18V. IMHO a well-designed pedal should work well at 9V (like most) up to 18V (in case of higher supply for other pedals on the board), with the only difference being more headroom (likely not used) at the higher voltages.

There are numerous pedal/audio-oriented sites  (eg, Geofex, BeavisAudio) where you can see example schematics and read domain-specific technical information that may help.

You might find LTSpice or similar helpful for simulating various options -- though of course that is no substitute for bread-boarding a working(!) circuit. 

Many modern Op-Amps (eg, TL072) will work well for instrument-level (<2V) signal with a "single" 9V supply -- one just needs to bias an AC-coupled input signal up to the mid-point (~4.5V) which can be obtained many ways.

The TL072 isn't specified at voltages below 10V, the graphs on the datasheet and other specifications such as common mode input range and output voltage swing, suggest it will stop working at voltages below 7.5V but it could be higher, depending on how it's used.

I think TI specifies their TL072 with a total supply voltage down to 7V (and up to 36V). http://www.ti.com/product/tl072