Need Help with a Basic Design Question

[Cigarsnob]

Hey everyone,

I feel like an idiot for having to ask this, but I need some help with a calculation question regarding an opamp gain based off an NTC thermistor (10kohm). In short, an APD is being temperature monitored to adjust the bias voltage. As the temperature increases, I want the bias voltage to decrease. I'm using an NTC thermistor to monitor the temperature and adjust the gain of an amplifier, which adjusts the high voltage regulator. See link below for the opamp configuration and a chart for the APD being used.

http://imgur.com/a/h5G6J

I want the APD to have a gain of 50. The chart given plots gain vs bias voltage for three different temperatures (0C, 25C, 50C). The opamp with the thermistor is configured in a non-inverting amplifier. The second page of the link has a little table of TC, VHV (High voltage), Divide by (this is how much the high voltage is divided by for the feedback), and the gain of the amplifier under question.

My question is, how do I calculate the gain of the amplifier? When I first tested the circuit, I left out the thermistor and just did a simple non-inverting amplifier to give me a gain of 1.33, which gave me a high voltage of 115V. Based off the APD chart, I know from 0C to 50C that there's a change of 80V over a 50C change. Which gives me a 1.6V change per degree C. Does this play into the equation of determining the gain?

I was thinking about taking the three temperatures for a gain of 50 for the APD. With that gain of the APD and the bias voltage required, I know the gain of the opamp (given in the table shown on page two as previously mentioned). From this, I have three unknown resistors and three gains. If I calculated the gain equation correctly (see page 1 for the equation, 1+k, where k is some resistor combination shown), then I know it needs three different gains. Solve the equations for each of the unknown resistors (3 total). But I feel this is incorrect as I'm using the same equation, but each equation just equals a different gain.

My apologies in advanced if this is confusing. I'm just trying to type out everything that I'm thinking. Any help is welcomed.

Thank you

[tszaboo]

I think that the problem is that you want to solve a non-linear problem, with linear tools. If I were you, I would enter the NTC characteristics, and the APD (is that the photodiode BTW) characteristics in Excel (or equivalent) and go from there. Calculate the gain for every 5 Kelvin, and see if you can solve it with a setup.
But as far as I see, it might not be trivial. It might be more time economical to solve the problem with a microcontroller look-up-table.

[Cigarsnob]

I think that the problem is that you want to solve a non-linear problem, with linear tools. If I were you, I would enter the NTC characteristics, and the APD (is that the photodiode BTW) characteristics in Excel (or equivalent) and go from there. Calculate the gain for every 5 Kelvin, and see if you can solve it with a setup.
But as far as I see, it might not be trivial. It might be more time economical to solve the problem with a microcontroller look-up-table.

Thanks for the feedback. The APD is an avalanche photodiode. I'll go the route of Excel (probably matlab though) and take it from there.