Post Go back to editing

LTC6560

Hi, I am currently working on a LIDAR prototype and I decided to use PIN photodiode to receive the laser. I am looking forward to use LTC6560 TIA. It says it is ideal for APDs. Can I use it for PIN and how the circuit differ if I use PIN?

Parents
  • Hi ROrta290,

    The expected LTC65650 bandwidth / speed will depend on the sensor (PIN) capacitance. For example, SFH203 PIN photodiode from Osram shows 11pF of capacitance:

    https://dammedia.osram.info/media/resource/hires/osram-dam-5488331/SFH%20203%20FA_EN.pdf

    You can use the LTC6560 LTspice model to estimate what the response / bandwidth will be with this much capacitance on the input.

    The best response / speed will be expected with a low capacitance APD as the sensor.

    Most TIA's have a similar dependence of speed on input capacitance.

    Regards,

    Hooman

  • Hello again ROrta290,

    Here are some more thoughts from our expert(s):

    The large TIA gain of the LTC6560 (74kV/A) would be beneficial in the case of the PIN Photodiode which has no gain (unlike the APD), we believe. The device would work with the PIN PD, but make sure the connection is such that the light pulses pull current out of the TIA input, if DC coupled. That's because the LTC6560 is optimized for fastest response this way. You could also AC couple the PD to the input, if that suits your application. Ditto the bandwidth limitation because of the larger PD capacitance noted earlier.

    Hope these responses answer your questions.

    Regards,

    Hooman

  • Hi again after a long time,

    I implemented the design recently. But unfortunately, the output I obtained from LTC6560 is not as desired. I shared the schematic of the design and the PIN photodiode I use is ODD-5W. The problem is that the output of LTC6560 saturates near 500mV. With this design I expect to have approximately 1.7Vp-p. In other words, I could not benefit from the high gain of the LTC6560. Do you have any recommendations related with the design or any clue about what can be the pronlem? Thank you.

    Sincerely,

    Orta

Reply
  • Hi again after a long time,

    I implemented the design recently. But unfortunately, the output I obtained from LTC6560 is not as desired. I shared the schematic of the design and the PIN photodiode I use is ODD-5W. The problem is that the output of LTC6560 saturates near 500mV. With this design I expect to have approximately 1.7Vp-p. In other words, I could not benefit from the high gain of the LTC6560. Do you have any recommendations related with the design or any clue about what can be the pronlem? Thank you.

    Sincerely,

    Orta

Children
  • Hi Orta,

    I understand you used a PIN diode on the input of the LTC6560 (AC coupled per your LTspice schematic) on a bench board / setup and you had some issues, but you're finding different results in the LTspice simulation plot you attached. Is that correct?

    What exactly was the measured results that is different from the LTspice plot you attached (which seems to show the 1.7Vpp signal you've noted you need)? Perhaps you can attach scope photos of the measured results? Keep in mind that with a 11pF input capacitance, there will be bandwidth reduction and it may be that your pulse width is too narrow to allow the LTC6560 to respond (I'm just guessing here because you've not provided these timing details).

    You may want to tie the OMUX pin to ground just to make sure your device output is active (even though there is a pull-down on that pin) - just to make sure.

    Let me know if you have any more details or preferably scope photos of what you measure please.

    Regards,

    Hooman

  • Hi Hooman, 

    Yes, you understood correctly. I shared the scope photos I have with me right now. The green waveform is the LTC6560 output.  Also, the pulse width is 20ns.

    Thank you,

    Orta

  • Hi Orta,

    Is it possible that your TIA input current is much higher than the 30uA you're simulating with LTspice? With the input AC coupled as you've shown in your schematic, a saturating event (i.e. high input current) can cause a long recovery time similar to the scope photo below and I'm wondering if that's what you're seeing here?

    Is there any way you can DC couple the input to see if this is related to input AC coupling?

    Regards,

    Hooman

  • Yes, I will try and inform you as soon as possible. So, can the low peak-to-peak amplitude at the output be also related with the saturating event? It seems that the figure you posted has also 500mV output in recovering state. 

    Thank you,

    Orta

  • Hi Orta,

    I'm thinking that you may be getting the low peak-to-peak amplitude possibly because you get into saturation and do not really recover in time for the 20ns duration of your pulse width.

    Please remind me again why you had to AC couple the input? With DC coupling response to a large input current is much better and this is the preferred method of input coupling. If you like to eliminate dark current or somehow need to cancel any input offset current, the following circuit is recommended where LT6015 acts as a servo loop to make sure the "average" output is at Vset (assuming low duty cycle TIA input current pulses).

    Here is a simulation file for the circuit above:

    LTC6560_DC_cancel.asc

    Let me know if you find out if DC coupling is an option and whether or not you find out the TIA input current was >> 30uA during your testing?

    Regards,

    Hooman