Op amp buffer and summing amp in series not working.

[poodina]

Hi

I have been trying to get the following circuit to work, it works in sections but does not perform as desired when connected as in the diagram.
Here goes the explanation.
I have to read a pulse in a ADC of PIC18F4620, the pulse varies between -5 to +5v, from an analog circuit. Since the ADC can only read between 0 to 5v have to level shift the signal and use a a zener diode to control the voltage at 5V , So I got the buffer working for impedance matching, and I am using a summing amp to level shift the zero to 2.5V [this circuit works on it's own perfectly].
The problem arises when I connect both together - when the input is 0V the buffer output is zero the output of the level shifter has to be 2.5V, but the output of summing amplifier increases by 1v [to 3.5V] ,I need help in this interfacing.

Could you please let me know what I am doing wrong?

 

[keith1200rs]

You have a 10V input range and 5V output range so you feedback resistor should be half the input resistor.

Keith

 

[albbg]

In principle, when you put zero at the input of the buffer, you should see at the output of the adder the same voltage you have at the inverting input (set by R24). However I didn't understand where and why you use a zener diode. If you have a pulse between -5V and +5V you can directly, by means of opamps shift it to the wanted level of 0 to +5V summing +5V and setting the gain to 0.5 (is the same suggestion of keith1200rs)

Furthermore I think it will be better to move the offset from the inverting input to the non inverting one changing the sign from -2,5V to +2.5V (simply moving the -12V input of the divider to the +12V) in order to avoid any current flowing among buffer and/or feedback resistor to offset circuit and viceversa (in some case this could be cause the problem you are experiencing).
I also will increase the value of all the resistor (factor of 20 or more) to eliminates the need of high wattage resistor and trimmer.

 

[keith1200rs]

Yes, bigger resistors would be good - I don't know what a 741 can drive as I haven't used one for 25 years. There are thousands of far better opamps to choose from.

Keith

 

[poodina]

In principle, when you put zero at the input of the buffer, you should see at the output of the adder the same voltage you have at the inverting input (set by R24). However I didn't understand where and why you use a zener diode. If you have a pulse between -5V and +5V you can directly, by means of opamps shift it to the wanted level of 0 to +5V summing +5V and setting the gain to 0.5 (is the same suggestion of keith1200rs)

Furthermore I think it will be better to move the offset from the inverting input to the non inverting one changing the sign from -2,5V to +2.5V (simply moving the -12V input of the divider to the +12V) in order to avoid any current flowing among buffer and/or feedback resistor to offset circuit and viceversa (in some case this could be cause the problem you are experiencing).
I also will increase the value of all the resistor (factor of 20 or more) to eliminates the need of high wattage resistor and trimmer.

Thank you for the reply, The zener was used at the point where the first diode is connected between the output of adder and ground [replaced the diode with a zener]. I was using this as the pulse is not a constant -5V to+5v it crosses 9V some times, it's from an analog circuit, I have to read it and display a bar graph on a LCD, so to protect the controller ADC pin from receiving higher voltage than 5 , I was using zener to controller the +v to 5volts.

I have used higher resistance values also, to stop the voltage flow. Let me see if I got what you explained. I keep the circuit same, change the OFF set, from +v input terminal of the op-amp [instead of grounding it] and and set the gain of the adder to -0.5, so I get the full range of the voltage. Now The off set has to be +5V at pin 3 of the op-amp.
Will try the circuit and post the results. Thanks a million for the reply.

Added after 6 minutes:

Yes, bigger resistors would be good - I don't know what a 741 can drive as I haven't used one for 25 years. There are thousands of far better opamps to choose from.

Keith

Thank you for the suggestion .
I got the idea of 0.5 gain that I can read the equivalent of the full range, Could you suggest some good op-amps, for this particular circuit. We have tried using OP07DNZ, But the output are as stable as the 741.

Added after 2 hours 4 minutes:

albg , hi
I tried to reconstruct then circuit as you have mentioned with,by shift the reference to R21 to pin 3 [+v input ] of the op amp, and using +5V ref at that point, also changed the resistor values to 22K [R21 and R20] and the feedback to 11K R27.
The output follows the input with gain 1, when there is no singel at R20.

Please if you could help me with a circuit diagram it would be great help.