Paired Op Amp Input Offset Correction

P

pnielsen

Member

• Feb 6, 2018

• #1

Each of the op amps shown in the attached circuit diagram is powered from differing rails (+, gnd and gnd, -) to produce two identical, but DC shifted, outputs.

In addition, the input signal is referenced to the negative rail instead of ground. There is a reason for this relating to balanced battery drain.

I tried adjusting the offset to each amp with R2, R3 and R7, R8 but canot bring it within range. Can anyone please offer a solution?

The final circuit will probably use LM358's.

 

Attachments

• offset_op_amps.png

  62.5 KB · Views: 248

K

kubeek

Well-Known Member

Most Helpful Member

• Feb 6, 2018

• #2

It would be better if you posted your asc file. Try powering the opamps from the edge rails, I think then it should work correctly.

 

P

pnielsen

Member

• Feb 6, 2018

• #3

As requested, here is the .asc file.

Not sure what you mean by "edge rails".

 

Attachments

• Dual Polarity Driver LM358 -2V Input.asc

  2.4 KB · Views: 221

K

kubeek

Well-Known Member

Most Helpful Member

• Feb 6, 2018

• #4

Like this

 

Attachments

• Dual Polarity Driver LM358 -2V Input2.asc

  2.4 KB · Views: 217

K

kubeek

Well-Known Member

Most Helpful Member

• Feb 6, 2018

• #5

If the input signal really needs to be centered about -4.5V, then you will need to add a level shifter to move it within the common mode range of the opamps.

 

P

pnielsen

Member

• Feb 6, 2018

• #6

Yes, that works with a few resistance changes. But in your .asc you changed the return on the signal input to ground. How can I make the circuit work with a return to the negative rail as originally drawn? There is a reason I would like it this way. Thanks.

 

Attachments

• Dual Polarity Driver LM358 -2V Input2.asc

  2.4 KB · Views: 202

P

pnielsen

Member

• Feb 6, 2018

• #7

kubeek said:

If the input signal really needs to be centered about -4.5V, then you will need to add a level shifter to move it within the common mode range of the opamps.

Click to expand...

I tried this with a divider across the rails to pin 3 of each amp but could not get it to work. Not enough adjustment possible by simply changing the resistor values. Would you have a sample circuit?

 

K

kubeek

Well-Known Member

Most Helpful Member

• Feb 6, 2018

• #8

What exactly is the input signal? Maybe you could couple it with just a capacitor.

 

P

pnielsen

Member

• Feb 7, 2018

• #9

Bump. Can anyone offer a solution to the circuit in my OP?

The signal generator (not shown) and two op amps are powered from a simple voltage divider off a 9V battery. Current draw must be balanced to keep the rail voltages equal. Hence the genrated signal fed into the op amps is referenced to -4.5V instead of ground, but the amps are powered separately from each leg of the divider.

Basically, I need two op amps with adjustable input level shift and amplification. Nothing I have tried has worked so far.

https://www.electro-tech-online.com/attachments/offset_op_amps-png.110756/
https://www.electro-tech-online.com/attachments/dual-polarity-driver-lm358-2v-input-asc.110758/

 

K

kubeek

Well-Known Member

Most Helpful Member

• Feb 7, 2018

• #10

Really? Maybe, just maybe, if you answered my questions I would provide you with what you want to know.
So I guess I will have to ask again: What is the signal generator doing? What voltage is that signal and what freqency or is it a DC signal? Does the signal go below the -4.5V rail?
Why not just show the generator? Why don´t you use an opamp for the virtual ground and stop worrying about balanced loading? What is the goal of the complete circuit?

 

P

pnielsen

Member

• Feb 7, 2018

• #11

The signal is a 3Vpp 10Hz sine wave referenced to the -4.5VDC rail. It does not go below the -4.5VDC rail.

The "generator" is a SOMO 14D module that plays back from a micro SD card. It is powered by 3.3VDC from a 5VDC regulator with two diodes in series. The regulator's common is -4.5VDC which explains why the SOMO 14D's signal is aslo referenced to that rail and not ground.

This is a battery powered circuti. I prefer to keep the component count low, e.g. without added active parts.

The circuit is intended to drive a magnetic coil through the 220R curent limiting resistor. Apart from that, I would like to learn how to implement a DC level shift in a situation such as described above.

 

K

kubeek

Well-Known Member

Most Helpful Member

• Feb 7, 2018

• #12

Ok, so the signal is 3Vpp centered about -2.85V?

 

K

kubeek

Well-Known Member

Most Helpful Member

• Feb 7, 2018

• #13

This I think should be what you want, but might need some tweaking of the output offsets to get the output offsets you want.

 

Attachments

• Dual Polarity Driver LM358 -2V Input3.asc

  2.6 KB · Views: 206

P

pnielsen

Member

• Feb 7, 2018

• #14

OK. This looks promising. Thanks. I will play with the values and report back tomorrow.

 

P

pnielsen

Member

• Feb 7, 2018

• #15

Here is the end result. Pretty good I would say, thanks to your help.

 

Attachments

• dual_polarity_amps_breadboard.jpg

  158.2 KB · Views: 211
• dual_polarity_amps_output.jpg

  147.6 KB · Views: 222
• dual_polarity_amps_schematic.png

  51.3 KB · Views: 216

K

kubeek

Well-Known Member

Most Helpful Member

• Feb 7, 2018

• #16

I am glad glad that we worked it out in the end. Also i would suggest getting some high efficiency step-down converter to power that SOMO 14D, otherwise you are wasting a lot of power on the linear regulator and drop diodes.
And i am sure there will be a lot of better choices for the amplifier than LM358, 1 to 2mA is quite a high static current for being powered by a 9V battery.

 

P

pnielsen

Member

• Feb 7, 2018

• #17

All good points. Once I verify the desired effect, I will look into upgrading the circuit. Thanks again.