This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

LMH32401: Photodiode Current Pulse Amplifier Figures of Merit - Comparing LMH32401 to MAX40213

Part Number: LMH32401
Other Parts Discussed in Thread: OPA855, OPA859

I'm trying to make a circuit to detect single-photon pulses on an avalanche photodiode (APD).  I'm starting with an S14422 APD.  

I don't have a model for the current pulses associated with a photon.  I have been testing an APD module, C13367.  The module's amplifier circuit outputs pulses in the 100mV, 10ns range.  I don't know what the signal from the APD itself looks like.

Not having a circuit model for my APD, I'm trying to understand the figures of merit for transconductance amplifiers.  

LMH32401 has built-in feedback resistors of 1k and10k.  That is small compared to the MAX40213, whose gain options are 150k or 750k.  My first thought is a part with this much gain would have poor high-frequency response, but look at this nice square 10ns pulse: 

Does TI have a high speed part comparable to MAX40213?  Are there tradeoffs that come with such high transimpedance?  What figures of merit should I be looking at to see the tradeoffs?  

  • Hello Charles,

       We do have higher speed parts where higher gain resistors can be used. For output pulses in the 100mV, would you need these higher gain options? Also, for a 10ns pulse, the system bandwidth needs to be at least around 35MHz. 

       However if you are using these modules rather than an APD, you would not need TIAs, but rather a simple voltage feedback amplifier to amplify the mV signals if necessary. Will you be switching over to just an APD later on in the design and these modules are for testing purposes? 

    Thank you,
    Sima

  • Will you be switching over to just an APD later on in the design and these modules are for testing purposes? 

    That's exactly it. I should have been clearer on that point.  I have a circuit for counting the 100mV 10ns pulses working.  

    Now I need a transimpedance amplifier to interface directly to the APD.  

    Do you know of a TI part similar to MAX40213, and what tradeoffs are associated with MAX40213's high gain?

  • Hello Charles,

    We do not have a single amplifier with TIA gain as high as the MAX40213. What we would recommend is to use an amplifier like the LMH32401 followed by the  OPA855 in a difference amplifier config to create the requisite gain. One advantage of this method is that the resulting gain is well controlled. Notice the MAX40213 device to device gain varies from 400kOhm to 1100kOhm - that is a variance of (1100-400)/750 = 93.3% over its typical value. The LMH32401 gain variation is around (22.5-17)/20 = 27.5% and the gain of the OPA855 as a difference amplifier can be within 1%. This large gain variation is only a concern if you plan to build these systems in larger volumes over a period of time.

    Alternatively you could also use the OPA859 as the TIA and configure it in a high gain.

    Finally we have a new amplifier, the OPA3S2859 which has integrated switches and can be used as a programmable gain TIA. This will allow some dynamic adjustment to maximize your SNR without saturating the output from the APD. The max gain of the OPA3S2859 and OPA859 can be set to > 100kOhm.

    1. Have you narrowed down the APD you plan to use - in the 1st thread you mentioned S14422 but that is an MPPC/SiPM?

    2. You mention a 10ns pulse - is this a square pulse or a gaussian shaped pulse?

    • If square what is the rise/fall time of the pulse?
    • if Gaussian is that 10ns pulse width representative of full-width at half-maximum?

    You did point to the pulse shape of the MAX40213 - notice those pulses are with saturated inputs in the mA range. The linear range of the MAX40213 is only 1V/750kOhm = 1.33uA. There are no pulses for a linear signal as far as I can see in the datasheet. However if you look at their frequency response graph below that is around 18-dB of peaking. I also assume they meant the y-axis to be dB-kΩ, rather than dB-Ω.

    Thanks,

    Samir

      

  • Hi Charles,

    the interesting question is, why do you want to use a TIA in combination with a APD? Is it to detect single photon hits only? Or do you want to achieve a high performance mesurement on the pulse timing? Or do you want to measure the energy of photon or the number of photons hitting the APD at the same time and you are interested in a highly linear circuit?

    As the datasheet says that MAX40213 is optimized for LIDAR applications - or by other words- for optical distance measurements. Here the plain echo detection with high time accuracy and fast overload recovery is what counts. But a highly linear TIA with precise gain and wide linear frequency response is not essential for this sort of application. This should be clear to you when using the MAX40213 Relaxed

    So maybe it's a good idea telling us what exactly you desire to do?

    Kai

  • why do you want to use a TIA in combination with a APD? Is it to detect single photon hits only? Or do you want to achieve a high performance mesurement on the pulse timing? Or do you want to measure the energy of photon or the number of photons hitting the APD at the same time and you are interested in a highly linear circuit?

    Thanks for your reply.  The APD I'm using is S14422 rightly points out this part is an MPPC, which consists of many APDs in parallel.  I am only trying to quantify light; time-of-flight does not matter.  My plan is to have a mechanical relay switching between two circuits: one for counting photon pulses when measuring very low light levels and another for integrating current.  The photon counting will be more accurate in low-light levels because it's digital counting of pulses.  For higher light levels, the frequency of overlapping photon pulses will increase, so I'll get better accuracy integrating current.  

    The reason I think I should care about bandwidth is a I think (could be wrong) that I'll need a lot of gain to detect the pulses.  My experience with measuring small amounts of lights is from PMTs and from the C13367 module.  I imagine (could certainly be wrong) the MPPC array of APDs in that module produces tiny amounts of current, which the module's amplifier (a black box to me) turns into the 100mV 10ns pulses I am counting, is like a high gain high-BW transimpedance amp.  

    So far I cannot find app notes about what I am doing or about what is going on in the C13367's black box.  The C13367 is not important, though; my goal is to use S14422.