• State
  • Locked Locked
  • Replies 1 reply
  • Subscribers 47 subscribers
  • Views 420 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.

High frequencies on the LMP7721 evaluation board

Jay Chang16
Prodigy 60 points
Other Parts Discussed in Thread: LMP7721, LMP7715

I have the LMP7721 multi-function evaluation board and would like to use it in TIA mode with frequencies in the 1-10 MHz range.  Will I run into any issues with either the LMP7721 or the LMP7715 guard driver?  It states in page 10 of the users' guide:

"The bandwidth of the guard driver circuit is determined by R1 and C5. The default R1 and C5 values are 10K and 1000pF, which creates a 16 KHz driver bandwidth. For higher bandwidths, these values can be decreased down to 100Ω and 82pF. But some bandwidth limiting should be used to prevent peaking of the buffer at high frequencies, which could cause instability of the entire system."

Should I replace R1 and C5 with 100Ω and 82pF ones since I'm significantly higher than 16 kHz?  Just out of curiosity, how are the values of R1 and C5 determined/calculated?  All of the voltage follower sites I've read place a direct short between the inverting input and output so I'm not entirely sure how to calculate those values.

Thank you!

  • 0
    TI__Guru 52396 points
    Hi Jay,

    The board was designed with low-frequency, sub pA measurements in mind - which usually means low (KHz range or lower) frequencies.

    One thing you have to be careful of is phase lag between the input signal and the guard. Any phase lag will decrease the bootstrapping effectiveness.

    You need some bandwidth control on the buffer to prevent peaking. Unfortunately, the follower is the most 'peakiest' configuration. Directly shorting the feedback on the buffer can cause peaking since you loose BW shaping ability. Peaking will cause issues because the effective guard gain is higher than 1 in the "peaking" region.

    These higher frequencies tend to easily capacitively couple back into the input through the guard trace. And any time you have a positive gain greater than one going through a non-inverting gain stage - you get a nice oscillation. So the buffer gain needs to be kept under 1 at all frequencies to prevent feedback through the input.

    Try 100 ohm and 120pF on the buffer feedback (~13MHz) as a start to minimize peaking. If you cannot eliminate the peaking, then you may need to add a large value resistor (~10k or so) between the guard and ground to slightly attenuate the guard signal (creating a voltage divider against the 100 ohm output resistor).

    Regards,