Summing opamp doesn't seem to work

[boylesg]

If I open all 3 switches the opamp output is below 1V.

If I close 1 of the switches the output goes up to 3V, which seems about right according to the calculations.

But if I close another switch then the output remains at 3V.

So it wont sum the 3 voltage like it is supposed to.

Can anyone tell me what I have done wrong.

I am referring to the circuit here: http://www.visionics.a.se/html/curr... and Averaging Amplifiers/Images\Image288.gif

 

[chuckey]

You HAVE to use the negative input of the opamp for summing, also there must be negative feed back to this point so it is a virtual earth.
Frank

 

[boylesg]

You HAVE to use the negative input of the opamp for summing, also there must be negative feed back to this point so it is a virtual earth.
Frank

What's with the non-inverting summing amplifier I have been reading about all over the place then?

I don't understand.

 

[ZASto]

Check the attached file.

 

[boylesg]

Check the attached file.

Cool! Thanks for that - it is rather useful.

I have found this whole thing rather confusing because it seems that on most electronics websites there is not clear differentiation between a voltage adder and a voltage averager.

I have discovered that applying multiple voltages to the positive input of the opamp gives you a voltage averager rather than a adder.

Which explains why when I implemented it I could see no difference in the voltage output as I added successive inputs.

What happens if you pass a steady negative voltage into an inverting opamp? Do you get the steady positive voltage?

 

[betwixt]

I think the misunderstanding is about where the virtual ground comes from. It is possible to sum voltages with a single supply but not in the configuration in the schematic.

The idea is that the summing point is at virtual zero volts, obviously not connected to ground but held at ground potential by the output of the op-amp. The feedback current has to cancel the current through the input resistors so at the junction they are equal and opposites, hence the zero result. If you connect positive voltages at the inputs, in order to cancel them the op-amp output has to swing negative and for that to happen it needs a negative supply.

Brian.

 

[boylesg]

I think the misunderstanding is about where the virtual ground comes from. It is possible to sum voltages with a single supply but not in the configuration in the schematic.

The idea is that the summing point is at virtual zero volts, obviously not connected to ground but held at ground potential by the output of the op-amp. The feedback current has to cancel the current through the input resistors so at the junction they are equal and opposites, hence the zero result. If you connect positive voltages at the inputs, in order to cancel them the op-amp output has to swing negative and for that to happen it needs a negative supply.

Brian.

Well I can't find a schematic for a voltage adder based on a single supply. The only circuit I have found is that voltage averager.

Any way I found that if I pass the negative voltage output of my regular voltage adder through and inverting opamp then I get the postive voltage I am after to pass into my arduino.

It will work out rather neatly. I have will need 6 x dual opamps, one side of which will be a regular voltage adder and the other side will be a unity gain inverting opamp and the output of this I can then pass into my arduino analog pins.

I have found I can reduce the output voltage of the voltage adder by making its Rf smaller than the input resistors and that way I can make sure the voltage that my arduino analog pins sees is always less than the 5V maximum. I will add some additional external protection on them to make sure.

So each key on my piano keyboard will have to have a comparator (12 x quad comparators) to drive its capacitive sensor and the output of each of these comparators will go into the inputs of the 6 x voltage adders in assorted combinations according to a table of 6 bit binary values.

The voltage adders will mean I can 'stack' the 1's for multiple simultaneous key presses and have a polyphonic organ and up to about 8 note chords (I suspect it would be rather difficult to do an 8 note chord on a piano keyboard).

I made a power supply circuit board with a couple of linear voltage regulators (9V for the adders and inverters and 5V for the comparators) and UC3843B charge pump to provide -9V or close enough. I salvaged a number of these PWM chips from ATX power supplies and thought I may as well start learning how to use them rather than the just using the familiar 555.

Had a bit of a hiccup with this latter because I soldered it up and it wouldn't oscillate. I eventually discovered it wouldn't oscillate on the 5V that I had powered it with but it would oscillate quite happily on 8-15V. Is there a minimum supply voltage on these PWM chips?