TIA Frequency Response
I am designing a TAI using this tool from Analog devices. The circuit diagram and it's frequency response obtained from the design tool is given below.
Circuit Diagram:
I simulated the same circuit in LTspice the only change I made is instead of dual supply I used single supply.I did not considered photodiode capacitance here.I will be biasing the photodiode as as shown in the first image.
In my circuit I am getting a bandwidth of only 10Hz!!!. Please find my simulation results below.The simulation file is also attached here.
May I know where I went wrong.
When I kept16nF.The simulation is failing
2 answers
You are accessing this answer with a direct link, so it's being shown above all other answers regardless of its score. You can return to the normal view.
The labels on your expected frequency response graph and the schematic of what you simulated are too small to see, so I can only answer with general observations.
- The Analog Devices schematic shows only 371 fF across the feedback resistor. Specifying such a small capacitance to 3 digits is silly when the inevitable parasitic capacitance could easily be several times that unless you very carefully use unusual construction techniques.
- Taking the 371 fF capacitance across 1 MΩ at face value (again, silly), the low pass rolloff frequency is 430 kHz. In reality, it will be less than that because of the real capacitance being higher, but not anywhere as low as 10 Hz.
- To get 10 Hz rolloff with 1 MΩ, you'd need 16 nF. That's much larger than stray capacitance, so would need to be deliberate. Again, because the schematic of what you're simulating is too small to read, I can't say what you may have done wrong.
- The capacitance of the photodiode doesn't matter because the voltage across it is constant. The cathode is held at 3.3 V, and the anode at 0 V. Since the voltage across the capacitor doesn't change, no current flows thru it, and it becomes irrelevant.
I replaced the schematic with new one with improved readability.
My volunteer time here feels abused due to the initial unreadable schematic, so I'll be brief.
The fact that your simulation is failing means something with the simulation configuration is bad. Try putting in a simple step of 1 µA and see what happens.
1 comment thread
The mistake you made is seems quite trivial. LTSpice recognizes your 20M as 20mHz (milihertz). If you want 20MHz, then you need to write "20MEG". That's all.
However, when I ran your schm, I got a bandwidth of about 1MHz, which didn't make sense, as I expected a pole at about 430kHz with a 1st-order roll-off. There seems to be some extraneous zero that cancels this pole and it rolls-off with the op-amps parasitic poles (~-40dB roll-off, so 2nd order).
That's as much as I found out. You probably have to characterize the op-amps in LTspice and in Analog's website to see who's wrong. But by the looks of it, I'd tend to trust the Analog's website one.
0 comment threads
0 comment threads