• State TI Thinks Resolved
  • Locked Locked
  • Replies 9 replies
  • Answers 3 answers
  • Subscribers 49 subscribers
  • Views 1886 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

Log amplifier

PRAHLAD Supeda
Expert 6200 points

Other Parts Discussed in Thread: TL441, LMH6881

My customer is looking for log amplifier with following specification:
Input current range:  10 nA to 150 mA
Input current pulse width: 6ns to 10ns (pulse width = ON time)   
Amplifier response time: <5ns

I couldn't find log amplifier >10mA. Please suggest way to achieve this, if we don't have a part.

Regards,

Prahlad

  • 0
    TI__Mastermind 25865 points
    Hello Prahlad,

    We have a limited portfolio of log amplifiers. As you noted, none of them have the desired range, much less the bandwidth required for this application.

    Please understand that this is not an easy problem to solve. Essentially the customer is trying to quantify more than 7 decades of current that requires a bandwidth of hundreds of megahertz. Furthermore, the lowest level of current (10nA) requires amplifiers with pA or fA of input bias current, which is typically not found in wide bandwidth devices.

    So, unfortunately we don't have a solution for this application that I am aware of.
  • 0
    TI__Guru**** 169936 points
    Prahlad,

    The fastest log amplifier should be the TL441. Even it might have problem with a 6ns pulse.
    Can you range detect and pre-amplify based on input signal?
    A TL441 can do 80dB range easy (with support circuitry) and 120db in theory only.
  • 0 in reply to Ron Michallick
    Expert 1057 points

    a436380.pdf

    I'm attaching the datasheets of two sensors which i need to work on.HARLID sensor gives current output and APD sensor comes with a preamplifier and gives voltage output.I need to detect minimum 100nW in HARLID and 0.5nW to 5nW in APD( LLAM 1060nm).Pulse width (ON time) for incident beam will be between 10 to 30 ns as well as above 100ns. Frequency will be between 1 Hz to 1 KHz (incident beam frequency).

    Can you suggest a multi-stage log amplification circuit or AGC.
    Please suggest solution.

    Regards

    Bivin

    LLAM_updated.PDF

  • 0 in reply to Ron Michallick
    Expert 1057 points

    Hi Ron

    I'm attaching datasheet for  two sensors.One is HARLID and another is APD (LLAM 1060)

    We need to measure the current pulse width in case of HARLID and voltage pulse width in case of APD (APD we buy as module which includes a preamplifier, internal circuit shown in datasheet).

    The minimum power incident on APD will be 0.5nW to 5 nW .We tune APD at 350V,1mA HV supply to get a responsivity of 200kV/W at 27C(temperature). 

    The minimum power incident on HARLID will be 100nW.

    Pulse width (ON time) in both cases will be 10 to 30 ns and sometimes more than 100ns.

    Request you to suggest a solution to get this pulse width.

    Please try the scope of AGC along with log amplifier.

    Also consider the scope of multi-stage amplification and pulse detection.

    You can find a sample circuit for HARLID which was designed for continuous laser beam.(Our requirement is pulsed laser with 10ns ON time)

    7610.a436380.pdf

    3312.LLAM_updated.PDFRegards

    Bivin George

  • 0 in reply to Bivin George
    TI__Guru**** 169936 points

    Bivin,

    Let's see if I can guess the output voltage into recommended load.

    Minimum 0.5nW * 200kV/W = 0.1mV

    Minimum 5nW * 200kV/W = 1mV

    I see your range is now only 20dB

    This signal is too low for TL441; it needs at least 10mV.

    Start a new post in a high speed amplifier forum to find such an amplifier.

    https://e2e.ti.com/support/amplifiers/high_speed_amplifiers/

  • 0 in reply to Ron Michallick
    Expert 1057 points

    Hi Ron

    In case of APD


    Output will be voltage(module comes with a preamplifier).

    The maximum incident power it can detect will give a saturation voltage of 3V/0.9V with respect to 1K and 50 ohm load resistors respectively.Maximum detectable power comes to be around 16.5uW for a responsivity of 200kV/W.So the range is more than 20 dB. Minimum power incident will be 0.5 nW.

    In case of Harlid

    Output is current.

    Output current varies between 10nA to 150 mA.

    Minimum detectable  power is 100nW.

    Regards

    Bivin George

  • 0 in reply to Bivin George
    TI__Guru**** 169936 points
    Bivin,

    Sorry for the delay, I was travelling.
    AGC is a good solution if the amplitude doesn't change quickly and there is time for AGC to settle on power up or signal appearance.

    The best solution depends on the signal being continuous series of pulses or not. (Needed for AGC)
    Do the pulses carry intelligence (data)?
    How is the data stored? (in amplitude,pulse width, pulse spacing, or some other means)

    The TL441 has hardware log sections that only have a 30dB each. Multiple sections can be use but the section inputs must be externally scaled. The pulses may be too short (amplifier delay to large) to use multiple sections for more active dB range. The input impedance for TL441 is 500 ohms
  • 0 in reply to Ron Michallick
    Expert 1057 points
    Hi Ron

    Pulse don't carry any information.Our requirement is to measure the ON time width as well as pulse repetition frequency.

    Thanks & Regards

    Bivin George
  • 0 in reply to Bivin George
    TI__Guru**** 169936 points
    Bivin,

    Sorry for the delay. I have given this much thought; the auto gain control is the most effective solution.
    However, this solution is well outside my product line. I suggest starting a new post in high speed amplifier forum.
    Try putting this part number in the post's subject. LMH6881