Non-inverting voltage amplifier effecting my Transimpedance Amplifier's output

[SidhdantThePenguin]

I am currently testing a 2 stage Picoammeter (which I have soldered onto a PCB) whose overall design is:

    1. Input current goes into a Transimpedance amplifier with a gain of 150M.
    2. Output of TIA goes into a RC Filter with a corner frequency of 10Hz (10k and 1.5uF).
    3. Output of RC Filter goes into a Non-Inverting Voltage Amplifier with a gain of 7 (50ohm and 300ohm).
    4. Output of Non-Inverting Voltage Amplifier goes into a RC Filter again with a corner frequency of 10 Hz same as before.

The entire circuit is powered through a 9V battery feeding into a railsplitter to get us a bipolar supply of ±4.5V to power all the OpAmps.

Here is the Spice Schematic:



Here is the PCB Schematic:



Here is the view of the actual PCB:





I have thoroughly verified the working of the Railsplitter, Transimpedance Amplifier and first RC Filter Stage, as in the output is a fairly linear function of the input current.

Problem: I have attached a 2 pin jumper to either connect/disconnect my non-inverting voltage amp (marked as A and B in the schematic and actual PCB) to the output of the RC Filter. The Non-inverting voltage amp works fine on its own (when you disconnect it from the previous RC Filter and supply a input voltage). The very peculiar problem is that when the two stages are disconnected even a simple act of touching the node B with a wire drastically alters the voltage at node A (for example with a input current of 1nA I get the output voltage to be around -145mV but touching the node B changes it to 145mV). I am stuck with this issue for quite some time as I very much want these two stages to be independent of each other and am not able to figure out why is the voltage amp affecting my RC Filter output eventhough they don't share anything?

[ledtester]

In your first post instead of the [img] tag, use:

Code: [Select]

Spice Schematic:

[attach=1]

PCB Schematic:

[attach=2]

Actual PCB Board:

[attach=3]


[SidhdantThePenguin]

Thanks a lot

[mtwieg]

Schematics are too low-resolution to read.

You could just attach the LTspice simulation files too.

[Alex Nikitin]

For starters, try to increase the values of the feedback resistors (50 Ohm and 300 Ohm) by about 200 times (say, 10K and 62K).

Cheers

Alex

P.S. the AD8597 is not a good choice of an opamp for the second stage.

[MariuszD]

Check if the circuit is not oscillating.
The schematics are unreadable, but the virtual ground circuit looks like something that might not work.

[David Hess]

When you disconnect jumper AB, the non-inverting input is left floating, which the operational amplifier is not going to like.

Maybe at a resistance to ground on the non-inverting side of the jumper to pull it to zero when the jumper is removed?

[Alex Nikitin]

When you disconnect jumper AB, the non-inverting input is left floating, which the operational amplifier is not going to like.

Not just the opamp, but the circuit altogether won’t like it, as the saturated opamp output will produce a lot of current into the rail splitter due to very low values of feedback resistors.

Cheers

Alex

[SidhdantThePenguin]

When you disconnect jumper AB, the non-inverting input is left floating, which the operational amplifier is not going to like.

Not just the opamp, but the circuit altogether won’t like it, as the saturated opamp output will produce a lot of current into the rail splitter due to very low values of feedback resistors.

Cheers

Alex

Weirdly enough I see the previous stages working in a correct manner when the jumper is disconnected (ie the input of the voltage amp is left floating). However the moment I short the jumper or try to attach a resistor to the input of the voltage amp  to ground I see the previous stages stop working correctly, which is why I found it so weird.

A floating input of the opamp saturating the output/loading the battery and virtual ground circuit would have made sense but I am observing the complete opposite where as long as I keep the input floating the previous stages work fine, the moment I try to "unfloat" the input the previous stages go haywire

[SidhdantThePenguin]

I did try attaching a 10k/1Meg resistor from the input of voltage amp to the ground without connecting the jumper but that too affects my previous stages which is the strange part as until I connect the jumper the only thing these two stages share is a power supply so I am not sure as to why is voltage amp affecting my previous stages. My previous stages work fine and as expected if I disconnect the jumper and leave the input pin floating