1
\$\begingroup\$

The signal generator outputs a sine wave into an LED which emits light onto a photodiode. The photodiode is connected in a TIA circuit and then the output is displayed on an oscilloscope.

The amplitude of the output signal stays the same up until 700 kHz. After this point, as I increase the frequency of the input signal, the amplitude of the output signal increases and it keeps increasing until 1.9 MHz and then it starts to fall. I am not sure what the issue is.

enter image description here

\$\endgroup\$
5
  • 2
    \$\begingroup\$ In general, you have a frequency response problem, but you knew that. What is actually causing that is impossible to guess from what you've told us so far. If you have a schematic of the the amplifier, post that, and the schematic of the diode's connection to it. If it's a commercial amplifier, link to a data sheet. Also the oscilloscope model number, and any components that you use to connect the amplifier to the scope. \$\endgroup\$
    – Neil_UK
    Commented Apr 5, 2023 at 19:43
  • \$\begingroup\$ This photodiode has a frequency spectrum that approximates the human eye and it is optimized for linearity. It is not a good choice for high-speed applications. \$\endgroup\$
    – Mattman944
    Commented Apr 5, 2023 at 20:55
  • 1
    \$\begingroup\$ See this article: e2e.ti.com/blogs_/archives/b/precisionhub/posts/… \$\endgroup\$
    – user1850479
    Commented Apr 6, 2023 at 2:21
  • \$\begingroup\$ What is the output signal if you interrupt the optical path? Above 700 kHz you may also receive direct RF radiation. \$\endgroup\$
    – Jens
    Commented Apr 6, 2023 at 15:23
  • \$\begingroup\$ How do you know that all the increase/decrease in signal is due to your detector? \$\endgroup\$
    – D Duck
    Commented May 3, 2023 at 21:21

2 Answers 2

Reset to default
3
\$\begingroup\$

The problem is the BPW21's self-capacitance and the op-amp's lack of gain. If the BPW21 was an ideal photodiode with zero self capacitance then the LF351's limited bandwidth of 5 MHz would still be a problem because there is very little open-loop gain at 1.5 MHz (and that's just the fundamental frequency of the squarewave you are trying to amplify as per your earlier question on this subject): -

enter image description here

So, you really do need an op-amp that has a significantly higher bandwidth. But, the self capacitance of the BPW21 (quoted as 48 pF) will mean that the op-amp has to drive that capacitance via the 2.2 kΩ feedback resistor and this makes it worse. At 1.5 MHz, this is an impedance of 2210 Ω. Some data sheets quote the BPW 21 having a capacitance that is much higher so you need to know what item you have bought and where it came from.

So, the op-amp is working a lot harder and the gain rises. This is your problem; a mixture of poor op-amp choice and a photodiode capacitance that is also going to ruin your square wave shape. Here's a simulation to show you the gain peaking versus photodiode capacitance (C1 below): -

enter image description here

And, if your BPW21 is like some of the devices you can find on the internet then the peak in the AC spectrum could be easily more than 10 dB.

\$\endgroup\$
6
  • \$\begingroup\$ I am inputing a sine wave not a square wave. The amplitude of the output sine signal increases as I increase the frequency higher than 700kHz. This is also due to what you have explained? \$\endgroup\$
    – rockky007
    Commented Apr 5, 2023 at 20:51
  • \$\begingroup\$ @rockky007 I have added a spectral response to show the peaking you can get. I've varied the photodiode capacitance from 0 pF to 40 pF to demonstrate the effect. If your BPW21 has an unknown brand, the self-capacitance may be very much more than what I've indicated. \$\endgroup\$
    – Andy aka
    Commented Apr 6, 2023 at 14:11
  • \$\begingroup\$ Yes I had a look thank you very much \$\endgroup\$
    – rockky007
    Commented Apr 6, 2023 at 19:49
  • \$\begingroup\$ @rockky007 please view what is said in this link regarding site etiquette and good practice --> electronics.stackexchange.com/help/someone-answers \$\endgroup\$
    – Andy aka
    Commented Apr 7, 2023 at 9:27
  • \$\begingroup\$ @rockky007 if you are happy with one of the answers then you should formally accept it like this: Image of what to do. \$\endgroup\$
    – Andy aka
    Commented May 3, 2023 at 20:47
1
\$\begingroup\$

Most likely the TIA circuit has a peak in its response. This may also point that the op-amp in this circuit has less phase margin left around 2MHz, although that doesn't have to be the case.

Less likely is that the drive circuit has a resonance.

I am not sure what the issue is.

I guess you have some unsaid requirement as to the flatness of the response and the bandwidth needed. What are they? Edit the question to add this requirement. And show your circuit schematic!

\$\endgroup\$
3
  • \$\begingroup\$ I have added the schematic in the question \$\endgroup\$
    – rockky007
    Commented Apr 5, 2023 at 20:11
  • \$\begingroup\$ I have added the schematic but I am not sure what you mean by requirement \$\endgroup\$
    – rockky007
    Commented Apr 5, 2023 at 20:24
  • \$\begingroup\$ The rise in amplitude concerns you for some reason. Such a reason should by driven by the performance required of the circuit. So you should have some response bandwidth and flatness specification that you then design to. Without a spec, the circuit works fine since there’s nothing that I see that I requires that the gain not rise. \$\endgroup\$
    – Kuba hasn't forgotten Monica
    Commented Apr 6, 2023 at 13:53

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.