Digital Potentiometer for specific gain values in op-amp circuit

[mdszy]

I'm working on a design for something that takes a signal in which is approximately 0.6Vpp. My goal is to be able to vary the amplitude of this signal anywhere from 0.1Vpp to 5Vpp in 0.1V steps. My plan was to use a resistor divider to bring the signal from its original 0.6Vpp to 0.1Vpp and then use an opamp and a digital potentiometer to vary the gain from 1V/V - 50V/V.

My question is, is this feasible with an inexpensive (Hopefully ~$1/chip in small quantities) digital pot? I have a feeling it is, as long as I choose the right R1 and R2 values. However, it seems like a lot of digital potentiometers have strange resistance values for the taps, since the number of steps is a binary multiple number (64, 128, 256, 512...)

I've attached a schematic. I'd appreciate any pointers as to how to go about figuring out what resistance values I need for this to work.

Also, I based the schematic on this application note provided by Microchip.

[Evan.Cornell]

You could take a look at Variable or Programmable Gain Amplifiers:
http://www.ti.com/amplifier-circuit/pga-vga/overview.html

[schmitt trigger]

To figure out the actual gain, write the opamp gain formula in an Excell spreadsheet, then with the autoincrement drag function create a column with all the possible ohmic values, plot the results.
You could also use a SPICE simulator and utilize a .STEP function and do the same.

Beware! Be sure to also check out the digipot's end-to-end tolerance.

[mdszy]

To figure out the actual gain, write the opamp gain formula in an Excell spreadsheet, then with the autoincrement drag function create a column with all the possible ohmic values, plot the results.
You could also use a SPICE simulator and utilize a .STEP function and do the same.

Beware! Be sure to also check out the digipot's end-to-end tolerance.

Ah, right! I totally forgot excel existed, haha. Thank you, I'll give it a whack!

[metrologist]

I modeled the attached. Changing the 1k to 166k gives +/-5V output.

[mdszy]

I modeled the attached. Changing the 1k to 166k gives +/-5V output.

Right. I understand how to get the top/bottom end gains. I just need to make sure that I can get a digital pot and choose my resistors such that I can get 1V/V, 2V/V, 3V/V ... all the way up to 50V/V. That's the issue I'm having.

[ArdWar]

Pick a digital potentiometer with specific value. Use it as inverting amplifier feedback resistor in rheostat (variable resistor) mode. Pick input resistor that is 1/64 of the digital potentiometer resistance (might need trimmer), and you'll get 1v/v per step if 6bit or fraction of that if higher.

Of course that'll still subject to the digital potentiometer's linearity and stability (might not hold the same value at all signal voltages, frequencies and temperature, for example)

[mdszy]

Pick a digital potentiometer with specific value. Use it as inverting amplifier feedback resistor in rheostat (variable resistor) mode. Pick input resistor that is 1/64 of the digital potentiometer resistance (might need trimmer), and you'll get 1v/v per step if 6bit or fraction of that if higher.

Of course that'll still subject to the digital potentiometer's linearity and stability (might not hold the same value at all signal voltages, frequencies and temperature, for example)

Thank you so much! This is exactly what I was looking for. I was having a massive brain block, I knew there was a simple way about this, and this will work! Thanks again!