Limiting Output Voltage of the Opamp to avoid saturation

[Kakeh]

Problem:

my photo diode receives pulses of width 10ns-150ns and repeating at rate of 1Hz-50KHz

the current from photo diode depending on incident light can go from 10nA-100mA,

in a nutshell, my application tries to digitalize the analog pulses for pulse width measurement

so i have 'only' two photo diodes to cover the dynamic range,after breaking the dynamic range into multiple channles like 1uA-5mA(less than 1uA its hard for me to measure 10ns pulses for several reasons,find here, here) and 500uA-100mA, there are some special reasons like optical attenuation before one sensor which made to break the photo diode ranges into overlapping one.

so for my first channel 1uA-1mA which is where i want to do I-V conversion using a TIA rather than using a resistor,so for a gain of 1K i was able to achieve proper gain 60dB at 100MHz, so i will not change the gain now,

through this 1uA will be 1mV and 5mA will be 5V, now that is not enough i have to put one more gain stage of 20V/V to read my 1mV, so the second stage would saturate with a input of 500uA itself, which is a set back to my approach (to stick to the problem i did not post second stage)

Approach 1:

so i used this techinique of using diodes in my loop so that the ouput would be limited to the cutoff of the diode, but this is leading to oscillations in other words unstability of the opamp
Results:

for a current above 450uA the opamp starts slightly oscillating, for a given current of 1mA on the right you can see completely unstable, i thought the problem is due to switching time of diode being 5ns only, so i changed it to schottky which resulted full fledged unstability, so there must some problem with opamp or the setup, please guide

Approach 2:

limiting the output current of the photo diode itself to 500uA ? using a current limiter, which did not workout properly, because the approach itself affects the frequency response and also adds distortion to the signal, the question quoted tries to limit the current to 5mA it can also be applied for 500uA, find it here, as its failed i don't want to bring it here.

Kindly suggest me an alternate approach to tackle this, or any modifications to existing design to get rid of the problem, to put it short i want to cover the range of 1uA-5mA

 

[Baluncore]

Kindly suggest me an alternate approach to tackle this, or any modifications to existing design to get rid of the problem, to put it short i want to cover the range of 1uA-5mA

Firstly; please add .txt to your LTspice.asc file, make it LTspice.asc.txt and attach it to your next post.
That way we will model the same circuit.

You have used a +/– 5V supply, but the op-amp is not rated to 10V. It is only specified from 3.1V to 5.25V maximum. I have not studied the op-amp specs so I cannot recommend a better choice of supply voltages at this time.

I think if you must use diodes then selection will be critical.
You have selected a diode that has a high capacitance when slightly forward biased.
Since you use back to back diodes you will always have worst case capacitance.
You do not need the 1pF cap in the feedback if you use diodes because diode capacitance is greater.
Your pulses are unidirectional so you only need one clamp diode.
The selected diodes are slow at 4ns. There are faster diodes available, or you might pre-bias them.

 

[Tom.G]

That opamp is only stable at gains of at least 10. See page 1, 1st paragraph, last sentence of:

http://cds.linear.com/docs/en/datasheet/626810f.pdf
"It is a decompensated op amp that is gain-of-10 stable."
Put a resistor (or a voltage divider) in series with the feedback diode(s).

 

[Kakeh]

please find the LTspice file, i tired of of putting a series resistor with diode which made it work as like there is no diode,

 

[Tom.G]

I think there is something wrong with the simulation. Put a probe on the opamp output and on the inverting input.

 

[Kakeh]

I think there is something wrong with the simulation. Put a probe on the opamp output and on the inverting input. View attachment 102101

yes if you see the output its not limited to diode voltage but to 2.5V which is saturation of opamp