ADPD2210 in low light application - how to deal with 240 nA static bias

Helllo,

I am kindly asking advice on the following matter:

We are considering to use the ADPD2210 as front-end for a photodiode in a very low light application. The

expected photo-currents are in the range of 1-1000 pA.

Since the ADPD2210 always exhibits a static bias current of 240 nA on the output pin the question is how

to further process this signal which will only slightly increase when the current due to the PD is added.

Following the suggested circuit designs in Fig. 25/26 from the datasheet the dynamic range of the measurement

would be rather small since the static bias would limit the size of the  Rs respectively Rf in order to gain the signal.

Is there a smart circuit design that would allow to measure very small currents with sufficient gain and not be

limited by the static bias current. Something like a differential measurement ...

Any advice is much appreciated!

Best regards
Peter

  • 0
    •  Analog Employees 
    on Sep 25, 2020 1:56 PM 1 month ago

    Hi Peter,

    Your question is actually very interesting and it requires some brainstorming. Also, there is a critical apps related question. Is the photocurrent continuous or are you using pulsed LEDs/light sources? Obviously, if the light source/PD current is continuous, it's harder to distinguish that from the output offset, but with pulsed systems, you could do smarter designs.

    Our field application engineers could help you with your design as well, you can fill out the below form so that someone could be assigned to you to help:

    https://form.analog.com/form_pages/support/integrated/techsupport.aspx

    This could be critical if your application is confidential. 

    Best regards,

    Yigit

  • Hi Yigit,

    thank you for the reply!

    > Is the photocurrent continuous or are you using pulsed LEDs/light sources?

    Initially the photocurrent will be DC in order to keep the design simple and learn about the actual signal levels

    available in the instrumentation. Moving on to a modulated light source and synchronous detection is under

    consideration. In that case using one of the mixed signal AFE from ADI might be the most efficient approach.

    Considering the low photocurrent in our application can you suggest a particular model of the ADPDxxxx series?

    My impression is that some external analog signal processing will be necessary to increase the current to a level

    which allows the internal TIAs of the ADPDxxxx to function properly. What's your assessment on this?

    Thanks again for you support!

    Best

    Peter

  • 0
    •  Analog Employees 
    on Sep 28, 2020 1:57 PM 1 month ago in reply to PeterTrollens

    Hi Peter,

    I see. I think the most well-suited option will be the newest one (ADPD4100/ADPD4101 - only difference is SPI or i2C) as you signals are low, and you will need a low noise AFE to distinguish your low level signal. 

    ADPD4100 has a float mode which allows low input current measurements. It basically allows some time of accumulation of charge on photodiode and dump the charge on photodiode very rapidly, which allows low current measurements. If you have ambient light, float mode is not recommended. Therefore, you could use the float mode if you don't have ambient light in your low current measurements. If your signal will be in low end of pA range, you may still need some more amplification on top of ADPD2210. ADPD4100 can process very small currents but pA may be low to see a meaningful code increase. It has to be characterized in your system, it is hard to tell right away.

    ADPD4100 allows differential input measurements, so you could either use two current sources and measure differentially, or if you two voltage sources, you could use a sensing capacitor between two inputs of ADPD4100 to measure differential voltage. There is an input resistor between the input and the TIA, therefore, it allows to measure voltage sources. 

    I would still recommend you to fill out the form that I sent, so that some field apps or central apps engineer may be assigned to you. We do not have a direct recommendation, but for a pulsed system, I could imagine some resistive load with few switches and capacitors could solve your offset canceling issue. You could basically charge two capacitors with a switched network so that you could get differential voltage across a third capacitor. So, one resistor at the the output of ADPD2210 to turn current into voltage, two switches to control charging of two caps at different times and two switches to create differential voltage on a third cap, which you could use as a sensing cap for voltage measurements with ADPD4100. This is NOT an official recommendation, I am just imagining in my mind now. So, please fill out the form that I sent to get a closer help from a central or field apps engineer.

    Best regards,

    Yigit

  • Hi Yigit,

    thank you very much for your elaborate support - much appreciated !

    The ADPD410x devices seems a very attractive device - especially the float mode looks like an interesting mode of operation - not only for PDs. As you say, well will need to determine the actual photo-currents in our instrument before deciding which amount of analog signal conditioning will be required. Good to know that several options will be available once we know exactly.

    Thanks again!

    Peter

  • 0
    •  Analog Employees 
    on Sep 30, 2020 1:27 PM 1 month ago in reply to PeterTrollens

    You're very welcome Peter! Yes, ADPD4100 is a very flexible AFE, so it could enable interesting applications. And as you said, you'll need some initial characterization on a prototype.

    Good luck with the design!

    Best regards,

    Yigit