I'm building an absorbance device with a UV LED and a photodiode. See the attached pdf for specifics. Basically, the LED shines light through a solution and the photodiode detects how much light is blocked. As concentration goes up, the amout of light reaching the photodiode goes down. This device is for the Pharmacy department at my university.
I'm using Adafruit's ADS1115 16 bit ADC with the Adafruit Python library. It provides 16x gain so I haven't tried using an external op-amp circuit. I'm also using a 5v switching power supply as Vdd for all of the components.
I have a working prototype, but it has a few problems. First, there is a fair amount of noise, even with the primitive .16 Hz RC filter and software averaging. More importantly, the signal drifts upward very slowly, and I'm not sure why.
Below are a few improvements I've been considering. Let me know if they're worthwhile or if I should do something else.
I'd like to replace the 5v power supply with a 9v power supply and a regulator. I also intend to attach a capacitor from the ADC's Vdd to ground. I'm not sure what capacitor value to use, though.
The diodes are currently connected with jumper wires. I'd like to connect them with shielded wires instead. Both are in a TO-5 package. The photodiode has 3 pins: cathode, anode, and one connected to the housing. The LED has a cathode and an anode, with the anode connected to the housing. Both are mounted in an aluminum enclosure, so their housings are electrically connected. I'm not sure what kind of shielded wire I should use or how I could ground it since the housings are essentially connected to 5v.
I'm also open to using a different ADC, to using a microcontroller with a built in ADC, or to any other option that would work better.
I'm using Adafruit's ADS1115 16 bit ADC with the Adafruit Python library. It provides 16x gain so I haven't tried using an external op-amp circuit. I'm also using a 5v switching power supply as Vdd for all of the components.
I have a working prototype, but it has a few problems. First, there is a fair amount of noise, even with the primitive .16 Hz RC filter and software averaging. More importantly, the signal drifts upward very slowly, and I'm not sure why.
Below are a few improvements I've been considering. Let me know if they're worthwhile or if I should do something else.
I'd like to replace the 5v power supply with a 9v power supply and a regulator. I also intend to attach a capacitor from the ADC's Vdd to ground. I'm not sure what capacitor value to use, though.
The diodes are currently connected with jumper wires. I'd like to connect them with shielded wires instead. Both are in a TO-5 package. The photodiode has 3 pins: cathode, anode, and one connected to the housing. The LED has a cathode and an anode, with the anode connected to the housing. Both are mounted in an aluminum enclosure, so their housings are electrically connected. I'm not sure what kind of shielded wire I should use or how I could ground it since the housings are essentially connected to 5v.
I'm also open to using a different ADC, to using a microcontroller with a built in ADC, or to any other option that would work better.