0

I am trying to create a photodiode circuit for this sensor so that I can connect its output to a microcontroller. I am following the document: https://www.ti.com/lit/an/sboa220a/sboa220a.pdf for my design. However, I am unable to get similar plots on the AC analysis after running a simulation of the design on Tina-TI.

For my situation, the following parameters are my design goals:

  • Input current: 0A to 100uA (based on reverse light current from Fig. 3 of https://www.vishay.com/docs/81169/vbp104fa.pdf)
  • Output voltage: 100mV to 3.3V (as I intend to connect this to a microcontroller ADC pin)
  • BW: kept at 20kHz (not sure if this is relevant to my scenario)
  • Supply: Vcc is 3.3V as it will run on the same power rail as the microcontroller, Vee=0V and Vref = 0.1V.

Using the parameters above, I have worked out the following values for the different components:

  • R1 = 32kOhm
  • R2 = 240kOhm, R3 = 7.5kOhm (based on a resistor ratio formula)
  • C1 <= 4.97uF (based on R1=32kOhm and bandwidth of 20kHz)
  • GBW - I was unable to work this out. For this situation I had been looking at using a LMP7221 opamp, which has an input capacitance of 11pF. I assume this is equivalent to Cd + Cm (differential and common mode capacitance). Then I used 17pF for junction capacitance of diode, I assume this is equivalent to diode capacitance at 3V of the photodiode I am using.
  • I kept C2 = 1uF

Using the following parameters above, I have generated the following schematic and plots. Schematic

AC Sweep

DC Sweep

I think the AC sweep does not look correct so I am not sure which parameter I might have entered correctly.

user1397215
  • 147
  • 4

1 Answers1

1

  1. R1C1=T is about 150 ms is far too low for 20 kHz with that needs a T~ 50us/2pi

  2. There is no good reason for Vref to be more than needed if you have a 1M load and 50 Ohm source. The output swing is rated for 2k and 10k loads. If you insist, at least use proper values of 8.45k instead of 7.5k to target 100 mV for Vref.

  3. Then change R1 from 32k to 30.1 k for 0.5% gain error with an output swing from 0.1 to 3.1V and a more accurate log range near 0.1V for dark current and dim light.

  4. Bode is correct for the C1 chosen in error but 20 kHz is for what light source? 100 MHz sweep is based on what assumptions?

Tony Stewart EE75
  • 1
  • 3
  • 54
  • 182