I am a relatively newbie here and I am trying to build this project... so do not kick me hard and forget any stupid errors you see on the project. I appreciate any suggestions, criticism, etc., preferably with instructions on what to do. Talk to me like I am 3 years old. Show pictures... 😃

enter image description here

Open this image on another window, or you will not see it with enough zoom.

This is a 4 stereo input switch. You press the key and you select the input key you want to send to the output.

First look at my power supply. I am using LM7810 and I have a resistor divider. So, My power supply is delivering +5V, 0 and -5V.

By having "negative voltage", I can setup the audio on the middle of the range and supply the MAX359 with positive and negative voltages.

I turn the whole thing on and it works almost perfectly. Every press of the switch and the next input is selected.

Now look at pins 8 and 9 of MAX359E (the switch). These lines enter on the potentiometers. The midle of the potentiometers go to the amplifier (TEA2025) and the other side of the potentiometer goes to -5V.

I have two problems with this circuit:

  1. Noise. It is humming like hell on the output.
  2. Another thing is that I am not sure if some voltages are correct. Suppose I put the multimeter's positive probe after C16 and the negative probe at -5V. If input U4 is not the selected input, that voltage is 5V as expected. As soon as I select U4, the voltage drops to 2.42V. I was not expecting that to drop.

Any ideas? Please remember to talk to me like I am 3 years old. Thanks.

  • 5
    \$\begingroup\$ Your power supply ground will not work. You can not 'divide' a 10V supply into two 5V one with 100K resistors, even worse is omitting the capacitors there. And no, I can't explain in words for a 3 year old why that does not work. The average three your old does not understand 'impedance, resistance etc. \$\endgroup\$
    – Oldfart
    Apr 7, 2018 at 20:43
  • \$\begingroup\$ Use two 9 volt batteries, and get better performance. Then sit back and think about why these 2 batteries do work. \$\endgroup\$ Apr 8, 2018 at 4:26
  • \$\begingroup\$ Try looking at the answers to this question to see what a proper power supply splitter looks like. \$\endgroup\$
    – Neil_UK
    Apr 8, 2018 at 4:40

1 Answer 1


Now look at pins 8 and 9 of MAX359E (the switch). These lines enter on the potentiometers. The midle of the potentiometers go to the amplifier (TEA2025) and the other side of the potentiometer goes to -5V.

  • Why? R7 should be a volume control, right? Then the lower end should go to GND, not V-. Because your signal has positive and negative voltages.

  • U2 has the same problem, outer GND is tied to -5V instead of GND. This will give you serious clipping.

Please solve these two problems first. Then, you may need to have the GND tighter than 50µA, the maximum current your voltage divider R3/R4 allows. Think of an active voltage divider using e.g. an op-amp unity-gain follower behind this resistor network. Also, create the virtual GND directly from the input, then use one 7805 and one 7905 to regulate +5V and -5V separately.

  • \$\begingroup\$ thanks I will modify the circuit. BTW, can you point me to a kind of circuit that may be suitable for this power supply. \$\endgroup\$
    – Duck
    Apr 7, 2018 at 23:58
  • \$\begingroup\$ This is a simple virtual GND: i.stack.imgur.com/pNmSy.jpg The purpose of the op-amp is providing more current into GND without having an excessive current from V+ to V-. \$\endgroup\$
    – Janka
    Apr 8, 2018 at 2:45
  • \$\begingroup\$ I have did everything you said: 1) connected the other side of the pots to GND. Connected all grounds of the input and output jacks to GND. Added an op amp to stabilize the virtual ground and now the op amp output is the new GND. Now, when I turn it on, I hear a sound like an helicopter on the output. I do not have an oscilloscope but I tried to measure +V and -V and they are oscillating like hell \$\endgroup\$
    – Duck
    Apr 8, 2018 at 15:46
  • \$\begingroup\$ If it's "helicopter sound", chances are you are still having the wrong reference potential somewhere so the 100/120Hz from the supply are added to your signal. Without having an oscilloscope, debugging this is complicated. You should cut down your device into several parts you could debug separately. \$\endgroup\$
    – Janka
    Apr 8, 2018 at 15:51
  • \$\begingroup\$ so, just to confirm: 1) I should connect this virtual GND to the ground of all audio jacks and to the other side of the potentiometers, so I have the audio in the middle of the DC rail. The ground of all ICs must be connected to V- and the power of all ICs to V+, right? \$\endgroup\$
    – Duck
    Apr 8, 2018 at 17:04

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.