Help with TL082 opamp with very low input signal change

Fausto,

What is the current you are expecting from the photodiode?  If the current is on the order of the bias current of the TL082 (30pA) it might be difficult to detect.  The TL082 has an offset voltage of ~3mV, and a 10M resistor would only give a 300 uV output from the diode current and another 300 uV from the bias current.  You could pick a better opamp.  But I suspect something else is wrong.  Even small photodiodes have outputs of 10's or 100's of nA in low light.  Turn on the room lights and see that you have a response (120Hz).  It is possible that something could be wired incorrectly.  Also note that you probably need a feedback cap to control your noise gain.  TI has a lot of good app notes on TIA design. 

But to answer your question, there should be a "dead zone".  It's just a question of having sufficient signal (and low enough noise).

Good Luck!

Tim

Tim:

Thank you for your replay

I am hoping to be able to see a change of 50pA or less

It is the current change (Less than 0.01HZ)  that I am interested, this is why is so low. The DC current is about 350nA.

Is there a better opamp that you could recommend?

Thank you again

Fausto Bartra

Is the DC current due to the background?  Is 0.01Hz the lower limit or the upper limit?  What is your measurement bandwidth?  Do you need DC response?  0.01Hz is going to be a problem.  You need to look for amplifiers with low 1/F noise. 

Tim

The DC current is background signal. (we do not used)

0.01HZ is the upper limit, basically DC levels

This circuit is part of the control loop to maintain a solid DC level (constant light output < 0.01% deviation)

Thank you

Fausto

That is certainly going to be a challenging problem.  I like the OPA627 (expensive!) and OPA827 for photodiode applications when I don't have power limits.  I think the OP-27 (or OP-37 if your diode capacitance is high enough to keep your noise gain > 5) has lower current noise.  While they have very good voltage and current noise specs, they have far more bandwidth than you need.  Like I said, 1/f noise will be important.  I'm hoping one of the engineers from TI jumps in...possibly they have a suggestion for a better amplifier.  I've always wondered about using a chopper in this case...OPA2188 for example...but I have never tested it as my electro-optics applications don't have to worry about DC response. 

Good Luck!

Tim

Tim and Fausto,

The OP27, OP37-type op amps (and their successors) have quite high input bias current and current noise. They are not well-suited to a transimpedance circuit with a 10M feedback resistor. A FET-input op amp is the correct choice in this case. The OPA627 and OPA827 are good recommendations. The OPA140 is another very good option.

Regards, Bruce.

And this is why I wanted Bruce to chime in ;)  Now I don't know where my brain was this morning...he's right...those amps are terrible.  I think we were using them back in the dark ages because they were in our flight stock.  Of course, that was back in the days of the OPA111...

Thanks Bruce!

Tim