Tutorial on the Theory, Design and Characterization of a CMOS Transimpedance Amplifier

In this episode, Shahriar and Shayan discuss the design and characterization of a deceptively simple CMOS inverter-based transimpedance amplifier. The the large and small signal behavior of the CMOS inverter is discussed and measured using the Keithley 2450 and 2460 source meters. The transient response is also measured using a Keysight MSO-S series oscilloscope.

The small signal gain of the circuit is calculated from small signal parameters which are extracted directly by measuring the devices I/V characteristics. The NMOS/PMOS devices used are from an ADL1105 quad-discrete transistor IC. Through the use of a shunt-shunt feedback, the CMOS amplifiers is converted to a transimpedance amplifier which is capable of amplifying the current from a photo-detector diode by a gain of 30kV/A. The feedback theory is used to calculate the gain of the amplifier. The slides for this tutorial can be downloaded here.

9 comments

  1. Larry says:

    I will add this to my blog as well. Thank you for the informative.

  2. tchicago says:

    I think it is easier to explain the TIA as an amplifier with a zero input impedance. This amplifier is sensitive to the current flowing from the photodiode, converting current on input to voltage on output. Then, since the input point remains at zero voltage, the parasitic capacitance does not need to be recharged, so there is less input shunting effect from the capacitance. Therefore, the frequency response improves.

    (it is possible that this all was mentioned on video but I missed that)

    • Shahriar says:

      Yes, but that is a highly simplified explanation and an idealized model. In order to correctly characterize the system, feedback theory must be applied which shows the input impedance of the amplifier is reduced by the amount of feedback (1 + BA). The input impedance become zero only if the open-loop gain of the amplifier is extremely large (essentially infinite). In this case, the open-loop gain of the amplifier is only ~15V/V.

      • tchicago says:

        Yes, that would be the introductory motivation part of why doing it that way. Then the characterization shows the real world parameters when applying this theory.

        Speaking of putting digital parts into the linear region, I remember articles in old soviet amateur magazines about building a radio receiver on a Russian analog of classic 7400. The chip was called K155LA3 according to the soviet naming nomenclature. Adding some smart feedback loops to force gates into the linear region, magnetic antenna and a tuning cap, and there you go, a working LW and MW radio. HF amp, detector and the audio amp – I think they were all made of those four NANDs.

        • Shahriar says:

          Those Russians built some amazing stuff with minimal components. I would love to see that article if you happen to find it.

  3. Jorge Garcia says:

    Excellent video!! Please continue to make more like these.

    I especially enjoyed the complementation of theory and measurements.

    Thank you for all of the work you do.

    Best Regards,
    Jorge Garcia

  4. Nick says:

    Great lecture.

    Why are many photodiode tia created with two opamps? It has something to do with noise I think.

    Where can I download your slides? It is not under the word ‘here’

    Thanks,
    Nick

  5. qeeee says:

    Enjoyable video, good harmony between you two.

    Now I know how to solve a similar problem I had characterising an OTA without information about W/L. What formula and parameters can I use to extract that information from the datasheet?

Leave a Reply

Your email address will not be published. Required fields are marked *

  • Social Media

    twittergooglepluswordpressyoutuberedditrssemail
  • Administrator

  • Archives