3-Channel Mixer

[Rory]

I'm not much of an electronics expert os bare with me.

I have just made the 3 channel mixer at:
http://www.electronics-lab.com/projects/audio/009/index.html

When I use it there are huge amounts of destortion unless I turn that channel down to almost nothing which renders it useless as a mixer. Is it a component that is causing this or should I just turn the inputs devices down greatly?
Which component in the circuit is limiting the volume it can reach before distorting? and is it possible to fix it?

Help would be much appreciated. Thanks

Rory

[audioguru]

Hi Rory,
Welcome to our forum. ;D
That mixer circuit is one of the worst circuits that I have seen! It is too simple to work properly.

I used the extremely old BC109 transistor about 40 years ago in my first job at Philips. Its current gain has a very wide range of values from 200 to 800. The simple mixer circuit has its base bias resistor selected for a transistor with a current gain of around 220, so an old BC109 with low gain would work nearly the same as a typical 2N3904 transistor. If the transistor has high gain then the circuit will produce extreme distortion like you have.

I simulated the circuit with a 160mV peak input signal (113mV RMS) with a typical 2N3904 transistor and the bottom of the output waveform was just beginning to clip. I replaced the 2N3904 with a 2N5089 transistor with a typical current gain of 600 and the output was extremely distorted and was still very distorted with a very small input level. If the transistor has its max gain of 800 then there won't be any output.

This is the year 2006, not 1966. So make a mixer with a good opamp then there isn't a problem with distortion or amount of current gain. ;D 

[Rory]

Thanks for the help

[MP]

Rory,
The project does not go into depth about what type of input signals it can handle. From your description about how you turn it down to almost nothing to get rid of the distortion, is a symptom that you are overloading the mixer with your input signal. You need a different design. What are the signal specs? Do you know the voltage of your signal? Current? Frequency range?

MP

[audioguru]

The transistor in the mixer project doesn't have any negative feedback to cope with a transistor that could have a current gain of 200 or 800 or anywhere in between.

Look at any transistor design course and it will show transistor amplifier circuits with an emitter resistor or with the base bias resistor connected to the collector for negative feedback. Many transistor circuits use a voltage divider for base bias so that the current gain of the transistor doesn't upset its bias.
This circuit has its only base bias resistor connected directly to the supply!

Simple arithmatic on the original circuit:
1) A current gain of 200. The 1M base resistor will have a current of 8.4uA. Therefore the collector current is 1.68mA and the collector voltage idles at 5.3V and the top of the waveform will clip first.
2) A current gain of 500. The collector current will try to be 4.2mA and the collector voltage will idle at 0.1V. The output will be half-waves with severe distortion.
3) A current gain of 800. The collector current will try to be 14.8mA and the collector voltage will idle at 0.1V. It will take a very high signal level to bring the transistor out of saturation for a small part of the waveform, with severe distortion.

I changed the position of the base resistor to the collector to add some negative feedback. The circuit works well with a 2N3904 transistor with a current gain of 230, and also works well with a 2N5089 transistor with a current gain of 600. I reduced the input level from a peak of 160mV before to a peak of 150mV so that the 2N5089 transistor doesn't clip its bottom too much.

It is interesting to see that the two different transistors create slightly different amounts of voltage gain.