PIN photodiode light detector circuit looks wrong.

The company where I am working have implemented a light detector circuit involving PIN photodiodes and Opamps. (as attached)
The problem I see is that the LMV321 opamp has a 3V3 rail and yet a 2V7 input, which is above the common mode input range of the LMV321 opamp.

Do you agree that this circuit is suboptimal?

LMV321 Opamp datasheet:
https://www.ti.com/lit/ds/symlink/lmv321.pdf

PD15-22C photodiode datasheet
https://media.digikey.com/pdf/Data Sheets/Everlight PDFs/PD15-22C_TR8.pdf

I would say it is borderline. It would be out of spec at 2.7V supply but well inside spec at 5V supply, the data sheet doesn't show the exact relationship between the two.

What puzzles me more is how it works, I can't tell if the diode is used in photo-voltaic or photo-conductive mode. I'm guessing photo-conductive in view of the relatively small feedback resistor value. PIN technology is normally used where high speed is needed but the LMV321 only has a narrow bandwidth and the feedback capacitor further reduces it.

Brian.

What puzzles me more is how it works, I can't tell if the diode is used in photo-voltaic or photo-conductive mode. I'm guessing photo-conductive in view of the relatively small feedback resistor value. PIN technology is normally used where high speed is needed but the LMV321 only has a narrow bandwidth and the feedback capacitor further reduces it.

Click to expand...

Thanks, i too am puzzled.....it looks to me that the Left hand opamp is "supposed" to see its output go high when light is incident on the photodiode there.
..and it looks like the Right hand circuit is "supposed" to see its output go low when light is incident on its photodiode.

Another problem to me is that each circuit probably acts just like a follower (opamp buffer) when no light is incident on its respective photodiode...and this would make the respective output be 0.6V and 2.7V respectively...and these are a little too near the microcontroller's high/low threshold of 2.2v/0.8v. (the opamps feed into a 3v3 microcontroller)

- - - Updated - - -

I would say it is borderline. It would be out of spec at 2.7V supply but well inside spec at 5V supply, the data sheet doesn't show the exact relationship between the two.

Click to expand...

Thanks, it looks like the common mode input range extends to up to a volt below the opamp's rail. (so up to 2.3v with the 3v3 rail, at least thats what i "assume" from the datasheet)
Would you agree that any operation whatsoever that risks any input ever going outside the input common mode input range of the opamp is very bad news and must be stopped.?
After all, if ever an opamp input goes above the input common mode range, the output of the opamp is likely to latch high and stay latched high, even if that input falls back into the common mode input range of the opamp.

They look like photoconductive (photocurrent) mode
and a "trying way too hard" transimpedance amp.
You could just convert current to voltage across
the feedback resistor (or a same-impedance, right-
bias network) alone. Other than drive strength.

Optical detection in SWIR and shorter (visible) does
not maybe need a PIN diode, because penetration
depth is probably less than depletion width even
for 3.3V Si PN diodes.

Seems like for pennies more a true rail-rail I/O amp
could be substituted, or else the right hand amp's
bias could be brought lower and the offset made up
somewhere down the line. Not clear at all, why the
two different setups.

Thanks, do you think the circuit might "get away" with the input being above the common mode input range of the opamp?

The difference would be too small to cause latch-up so I don't think any damage would occur.
Even accepting the op-amp is OK, if it works in photoconductive mode it should have a current leak to the other supply rail and in photovoltaic mode the feedback resistor should me much higher value, maybe 1M or more.

Brian.

I think there may be two "latch-up" modes being discussed
here. One is 4-layer (SCR) latchup between the power rails.
An input exceeding the VCM limit doesn't cause this but one
exceeding the rails by a Vbe (and with enough trigger
current behind it) might. This should be spec'd in the Abs Max
tables and defeated by proper I/O pad and attached-device
design (guardrings and such).

The other looks to be a functional / behavioral deal, a
latched circuit state but without excess / destructive
supply-supply current, stemming from the front end
pair being choked off. Both PNPs need to go above
(Vdd-Vbe-Vce(sat)) to choke the front end currents
off. One is pinned to the 2.6V reference on the right
hand amp, so this would not occur (but is pretty
close, maybe too close for low temp comfort). As
long as one input remains below that threshold the
amplifier will have the right sense because currents
will steer appropriately.

Any op amp or comparator with a phase inversion
anywhere in the Vcm range should be shunned, of
course. And I prefer that (no phase inversion) to
encompass the supply rails range regardless of
whether common mode range for performance,
does or not. But it can be hard to pick out from a
datasheet, and if the part acts wrong in this respect
it might just not be mentioned.

Thanks, do you think the circuit might "get away" with the input being above the common mode input range of the opamp?

Click to expand...

A possible result for circuit 2 with 2.7 V common mode voltage is that the output is initially high and always stays high, not matter what the photodiode signal is, without needing any latch-up or phase reversal effects.

There's a 50% change that the offset of the folded cascode steers the output high if both inputs are above the common mode limit and the differential pair is cut off. In the other case (output low with input stage cut off), you get a poor TIA operation with the output biased roughly at the common mode limit.

The behavior depends on type variations, so some exemplars may work while others fail.

Thanks, i believe i agree with you. I never ever use opamps in any situation where they would suffer an input outside of the input common mode range. -But this company where i work have done this, as in the top post.
I believe the circuit must be changed, do you agree?
There are planty of opamps with rail to rail input operation, and theyre barely a cent more in price.