Simple headphone amp with mixing inputs?

[tuck1s]

I'm looking for a headphone amp to allow me to practise guitar while playing along with backing tracks on ipod.

My effects pedal produces a healthy low impedance output and is volume controllable. (In fact it can drive headphones OK on its own).

I've looked at the CMoy basic circuits but I'm thinking an inverting design with mixing resistor inputs might be more suitable.

So there's nothing 'guitar-specific' about this application; it's just line-level mixing then an amp/buffer.

Would be nice if the circuit can be expanded to 'n' inputs instead of just 2 (e.g. practise with a friend) and 'n' outputs but that's strictly optional at this stage.

Ideally I'll run it off the same 12vdc power supply as the effects pedal (by chaining the DC through the amp and out). So would be good to have a Cmoy-like negative-rail generator thingy. Dunno how good that PSU is, so maybe include some regulation (or at least heavy duty smoothing).

I would appreciate some pointers in the right direction.

[jman 31]

HERE is my design. It works great for me! I use it with my bass guitar and my 6 string.

[tuck1s]

Great, thanks for sharing. Interesting to see the guitar stage, although I have a pedal that gets me up to line level already.
Do you get reasonable battery life out of the LM386's? They can kick out nearly a watt, if I recall correctly. I was wondering about using something a bit lower powered..

[ohnoezitasploded]

I'm currently building a practice amp with the TDA7052A, and am pretty happy with the quality of audio I'm getting out of it. The two things I've learned about it:
1. It wants 6V. Not 9 or 12, as it will get hot enough to melt your protoboard.
2. Even with 6V, it gets very warm when driving it at the full 1W output. Be sure to make some provision for heatsinking or airflow.

[madtecchy]

Originally Posted by ohnoezitasploded

I'm currently building a practice amp with the TDA7052A, and am pretty happy with the quality of audio I'm getting out of it. The two things I've learned about it:
1. It wants 6V. Not 9 or 12, as it will get hot enough to melt your protoboard.
2. Even with 6V, it gets very warm when driving it at the full 1W output. Be sure to make some provision for heatsinking or airflow.

Hi i think you may find it is a combination of the supply voltage and speaker impedance that makes you chip run hot on higher supply rails.
What is the impedance rating of your speaker????

[ohnoezitasploded]

Originally Posted by madtecchy

Hi i think you may find it is a combination of the supply voltage and speaker impedance that makes you chip run hot on higher supply rails.
What is the impedance rating of your speaker????

I'm using an 8 ohm speaker, the rated impedance of the chip. It still gets hot when I'm running it full out at 40x gain, but at a 6V input it hasn't melted my protoboard again.

[gootee]

Originally Posted by tuck1s

I'm looking for a headphone amp to allow me to practise guitar while playing along with backing tracks on ipod.

My effects pedal produces a healthy low impedance output and is volume controllable. (In fact it can drive headphones OK on its own).

I've looked at the CMoy basic circuits but I'm thinking an inverting design with mixing resistor inputs might be more suitable.

So there's nothing 'guitar-specific' about this application; it's just line-level mixing then an amp/buffer.

Would be nice if the circuit can be expanded to 'n' inputs instead of just 2 (e.g. practise with a friend) and 'n' outputs but that's strictly optional at this stage.

Ideally I'll run it off the same 12vdc power supply as the effects pedal (by chaining the DC through the amp and out). So would be good to have a Cmoy-like negative-rail generator thingy. Dunno how good that PSU is, so maybe include some regulation (or at least heavy duty smoothing).

I would appreciate some pointers in the right direction.

You can easily expand to n inputs. Check out http://www.national.com/an/AN/AN-20.pdf and http://www.national.com/an/AN/AN-31.pdf . Look at the summing amplifier circuits, e.g. Figure 4 in AN-20. There is an equation, there, for the value of the resistor from + to ground, which says that in order to minimize the offset error due to the input bias currents, that resistor's value should be equal to the parallel combination of all of the summing junction resistors and the feedback resistor. So if you wanted to add another resistor to the summing junction (to add another input), you would just lower the value of the resistor from + to ground, per the equation.

For n outputs, just use a unity-gain buffer amplifier to make each output (See Figure 3, AN-20). You could add as many as you like.

You can insert a summing amplifier before basically any other amplifier, to make n inputs, and can convert many existing amplifier circuits to make them into summing amplifiers (to get n inputs without adding another amplifier stage). And you can insert n parallel unity-gain buffers after almost any other amplifier (with appropriate levels; otherwise voltage divider then buffers), to make n outputs.

For some better (than just a single opamp) but still very simple low-level audio amplifier circuits and topologies, check out this downloadable Walt Jung book from Analog Devices: ADI - Analog Dialogue | Op Amp Applications Handbook . (Remember to also look at section 7 and add the proper RF filters, etc.) Some of the composite amp and line-driver amp circuits in that book look really good.

Cheers,

Tom