Audio Summing Amplifier

Thread Starter

Malsch

Joined Mar 19, 2011
23
Hi, I am designing a summing amplifier with 3 inputs. The inputs can be any type of wave (from a signal generator for example). The output wave can then be input into a power amplifier and then output trough speakers/headphones.

The design is ready and working on Proteus. The lecturer told me that all inputs and outputs have to be buffered. This is what i didn't understand. How do you buffer inputs/outputs? with an op-amp on each I/O? and what is the purpose of buffering them?

Also is there a maximum voltage input of the power amplifier?

The design is attached. The gain of each wave can be controlled by a potentiometer and the total 'volume' can also be controlled. The capacitor is used to roll off any unwanted frequencies (above 20khz).

Thank you!
 

Attachments

jegues

Joined Sep 13, 2010
733
The lecturer told me that all inputs and outputs have to be buffered. This is what i didn't understand. How do you buffer inputs/outputs? with an op-amp on each I/O? and what is the purpose of buffering them?

Also is there a maximum voltage input of the power amplifier?

The design is attached. The gain of each wave can be controlled by a potentiometer and the total 'volume' can also be controlled. The capacitor is used to roll off any unwanted frequencies (above 20khz).

Thank you!
You can create a buffer by simply implementing a voltage follower configuration using another op amp. You can also use a common collector or common drain configuration if you are looking to implement using either BJT or MOS transistors.

The purpose of buffer is to essentially eliminate any loading issues you may experience between the source and the circuit that you're connecting.

For example, imagine I have a circuit that is simply a resistor and I want to apply a voltage across it.

Well, I hook up my voltage source across my resistor and I'm only seeing 90% of the source voltage.

What's going on? There are loading issues present. Because the source itself has some internal resistance, the source voltage is being divided across the 2 resistances.

You can prevent this by implementing a buffer between the two. A buffer will provide with "infinite resistance" such that the drop across the internal resistance of the source is negligible.
 

Adjuster

Joined Dec 26, 2010
2,148
Buffering in this context probably does refer to the provision of separate input and output amplifiers. This may be done for a number of reasons, principally to isolate signals from each other, so that there is minimal "leakage" of signal e.g. from one input to another.

In addition, buffer amplifiers may help to obtain suitable input impedances, typically high at the input(s) and low at the output(s).

Buffer amplifiers are often, but not necessarily, unity-gain voltage followers. These could for instance be added right at the inputs of a circuit like yours.

Many power amplifiers, but not all of them, are designed for a standard line input level of 775mV RMS, or about 1.1V peak. http://en.wikipedia.org/wiki/Line_level
 

Thread Starter

Malsch

Joined Mar 19, 2011
23
Hey, thanks for your fast answers. Now i understand much better :)

Is this the correct way to implement the buffers? (3 for inputs and 1 for output in attached file). Could I have used the 74LS365 (has 6 buffers) also instead of op-amps?

Thanks Again!
 

Attachments

jimkeith

Joined Oct 26, 2011
540
Hey, thanks for your fast answers. Now i understand much better :)

Is this the correct way to implement the buffers? (3 for inputs and 1 for output in attached file). Could I have used the 74LS365 (has 6 buffers) also instead of op-amps?

Thanks Again!
Nice try, but no cigar--the 74LS365 is a digital device with only 2 (or 3) output states so it will not work for audio (linear) applications--will sound very bad...

Try a quad op amp (perhaps LM324) connected as voltage follower as in your last sketch--yes that is correct.
 

Adjuster

Joined Dec 26, 2010
2,148
That's right, for this job you need linear buffer amplifiers, not logic devices. If you look at the data-sheet for a 74LS365 you will see that it is not suitable. http://www.futurlec.com/74LS/74LS365.shtml

That's something you might think to do in future, by the way. It is a good idea to check the manufacturers' data for any device that you are planning to use.
 

crutschow

Joined Mar 14, 2008
34,280
...............

That's something you might think to do in future, by the way. It is a good idea to check the manufacturers' data for any device that you are planning to use.
It's not just a good idea, it's an absolute requirement.:)

One note: The adjustment for each input level will be rather non-linear with pot position due to the loading of the 100k pot outputs from the 47k input resistor. A 10k (or lower) pot value would be better.
 

Thread Starter

Malsch

Joined Mar 19, 2011
23
Thank you for your advices. They improved the design :)

I'm using the UA741 because i have a bunch of them at home, and this is a simple experiment I'm building on a breadboard.

Thanks everyone and have a happy new year!
 

Audioguru

Joined Dec 20, 2007
11,248
The 741 opamp is 44 years old and was never intended to amplify audio.
It is noisy so its hiss level is high.
It has trouble above only 9kHz. Audio goes to 20kHz.

Your input opamp buffers do not have a DC voltage (0V) at their inputs so they probably will not work.

Your output 741 buffer ruins the very good output of the NE5532 audio opamp.
 

Adjuster

Joined Dec 26, 2010
2,148
For the benefit of the OP, Audioguru has made an excellent point that seems to have escaped the rest of us. Operational amplifiers require defined input voltages: bipolar ones like the 741 require quite significant input bias currents.

There are therefore some things missing from your schematic, but I fear incurring displeasure for spoon-feeding too much information. Suffice to say therefore that if you look up the data-sheet for the 741, you can find out how much bias current is likely to be required, which may give an idea of a reasonable upper limit for the value of the missing parts. If you have been properly taught about bias current problems, you may want to add a few more bits for better balance.

Having said that, on top of advantages already mentioned, a modern op-amp may have a lot less bias current [but still it won't tolerate a floating input].
 

Thread Starter

Malsch

Joined Mar 19, 2011
23
hey, thanks for replying. You do make a very valid point. The noise produced by the input buffers is also amplified by the audio op-amp and noise is also increased at the output buffer. So i replaced the 741 with the same audio op-amp being used. I included capacitors at the inputs to remove any dc voltages and to create a high pass filter at about 20Hz. To supply the op-amps with dc voltage, i added a pull down resistor. Is the pull down resistor at the output buffer necessary though?

Thanks.
 

Attachments

Audioguru

Joined Dec 20, 2007
11,248
The OP27 is an excellent low noise audio opamp. Its input current is so low that R4, R8 and R11 can be 100k then the input coupling capacitors can be a much smaller and less expensive 0.1uF (use film capacitors) and R5 can be replaced with a piece of wire.

R6 does nothing and can be removed since the output of U3 is 0VDC.
Output buffer opamp U1 also does nothing and can be removed since the output of U3 has an extremely low output impedance.
 
Top