• State
  • Locked Locked
  • Replies 2 replies
  • Subscribers 46 subscribers
  • Views 797 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

LMP7721 closed loop bandwidth

Marco sartore
Prodigy 30 points
Other Parts Discussed in Thread: LMP7721, LMC6482, LMC6041, LMC6081

Hello,

I want to use the LMP7721 to amplify a 10nA to 50nA input current. In order to get a corresponding 1 to 5V output range, I need a 100MOhm feedback resistor. Supposing to use an almost perfect PCB (as its EVAL board from TI, for example), how high can I hope to go in bandwidth ?

Things go better if I accept to have a smaller feedback resistor, but in that case I will need a second (voltage) amplification stage to go to the desired output range. Will this be a suitable solution or I will add too noise to the output signal ?

Many thanks in advance for your considerations,

best regards

Marco Sartore

  • 0
    TI__Guru 54166 points

    Hello Marco,

    It will all depend on the amount of capacitance on the inverting node (the transimpedence amp "input").

    Any capacitence (including strays, cable and/or sensor capacitance) on that node will cause large peaking, which must be compensated for by placing a cap across the feedback resistor.

    The compensation cap usually is roughly the same as the input capacitence, or a little more.

    Search for "transimpdence amp" or "Photodiode amp" and you will find many appnotes and articles describing how to select the compensation cap.

    AN-1803 is a good start:

     http://www.ti.com/lit/an/snoa515/snoa515.pdf

    Of course, this cap limits the bandwidth. And fractions of a picofarad make a big difference at these impedance levels. 1pF and 100M is already 1.5KHz - and thats just the first pole. Remember that the surface mount resistor itself will have a minimum of 0.25pF across it just sitting there...not including pad strays.

    The 7721 eval board has 11pF at the input (with guarding on). Most if that is the amplifiers input capacitance since the external capacitance is cancelled by the guarding.

    So just the board alone, with full compensation, will give you just a few hundred Hz of BW.

    The best S/N ratio is achieved using the largest feedback resistor value possible. It is better to use a large value resistor than to use a smaller resistor and add gain - but you may have to do this if you need more BW....as you discovered...

    To maximize BW, the amplifier and strays should have as low a capacitance as possible. Due to the low-noise nature of the 7721 - it has rather high input capacitance. Try a device like the LMC6041, LMC6081 or LMC6482, which have about 2-3pF input capacitence. The amplifer noise is basically irrelevent since is is swamped by the noise of the resistor.

    And use only as much BW as you need...excess BW is just extra noise.

    Regards,

  • 0 in reply to Paul Grohe
    Prodigy 30 points

    Dear Paul,

    many thanks for your valuable and professional answer.

    In any case among those described, I argue that one can rise the bandwidth to some kHz at most, and that very high bandwidths such as 1MHz or more is simply impossible with these low level currents at the input to amplify. Or maybe there is a completely different circuit topology (and devices to utilize) that can match these limits ? (I have read some post about switched capacitors or so...).

    Many thanks in advance,

    Marco