• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 47 subscribers
  • Views 985 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

LMV771 Input near/below GND (photodiode transimp. amp)

Martin55843
Intellectual 400 points
Other Parts Discussed in Thread: LMV771, LMV774

Hi,

I have a photodiode transconductance amplifier that I wanted to implement with a Microchip MCP601. Application is single supply +3 V. I wanted to use the quite practical configuration from figure 4.12, page 15 in the datasheet:

http://ww1.microchip.com/downloads/en/DeviceDoc/21314g.pdf

This is quite interesting as no additional (potentially noise-inducing) bias voltage is necessary. The trick is that IN- is driven slightly below GND by the photodiode, which in turn is being compensated by a positive voltage at the output. This works as V(CMR) is specified as Vss-0.3 ... Vdd-1.2 V. Of course the necessary range is only a few mV (or uV) below GND due to the high open loop gain (not sure how input offset voltage comes in play here).

As I have an LMV771 (774) elsewhere in the design (because I need the lower noise), I wonder if I can throw out the MCP601 and use the LMV771 instead (in order to reduce part count). Now V(CM) is specified as 0 - 1.8 V for the LMV771. The question is whether the LMV771 would also work with IN- slightly below GND as in figure 4.12 above. 

Any ideas?

  • 0
    TI__Guru 52536 points
    Hello Martin,

    The circuit in 14.2 is a classic grounded-input photo-diode trans impedance amplifier.

    If operating correctly, the inputs should both be at ground (± the amps offset voltage). Pulling -IN below +IN would bypass the TIA feedback and hard-rail the output.

    Current must be pulled out of the non-inverting node to make the output move positive. The diode is operating in current mode since the TIA is forcing 0V across the diode.

    Negative bias is normally applied to photo diodes to reduce the capacitance of the photodiode, which can be significant.

    For that circuit, the LMV774 would work fine. It can sense 100-200mV below V_ withput any problem. The main limitation is the ESD protection didoes starting to conduct and adding current.

    Regards,
  • 0 in reply to Paul Grohe
    Intellectual 400 points

    Hello Paul,

    thanks for the answer! I am well aware of TIA configurations and the 14.2 is classic for dual supply op-amps, I'd say. But going into a splitting-hair-mode, I'd argue that in a single supply configuration, this circuit is outside the specified range for a LMV771. Assuming the output is at 1 V by some photocurrent (ie. 1 uA with 1 MOhm feedback), the inverting input is (DC gain 100 dB, 100000x) -1V/100000 = -10 uV. As VCM is specified as 0 - 1.8 V, this is not covered by the spec (also assuming zero offset voltages/currents). 

    Is it a valid assumption that for any op-amp with VCM starting from 0V, it will operate also down to a few (hundred?) negative millivolts (bearing the input protection diodes in mind)?

    It would be nice if this was specified as in the Microchip datasheet!

    Regards, Martin

  • 0 in reply to Martin55843
    TI__Guru 52536 points

    Hello Martin,

    Yes. For a single supply the input pin will be slightly negative, due to the source pulling current out of the input. Technically, it is violating the CMR specification.

    For most "ground sensing" and R-R input  amps, the common mode can extend 200mV to 300mV below V- (at room temp - but common mode range can shrink over temperature). Many can go further, but become limited by the ESD clamps.

    The Vos vs Vcm graphs on page 8 show the 771 common mode extends well down to about -300mV, so this application should not be an issue at the rail and should be similar to the Microchip.

    Regards,