Post Go back to editing

using LTC6268 as a TIA, but has a sinusoidal waveform which is 300Mhz in the output port

Category: Hardware
Product Number: LTC6268

Dear Sir,

   Using LTC6268 as a transimpedance amplifier (photodiode mode), with V+ as the 5V power supply, V- as GND, the non-inverting input connected to a 1.25V reference source, and the inverting input connected to the positive terminal of the photodiode. Rf is 9.09MΩ. Currently, a sinusoidal waveform of around 300MHz with an amplitude of 100mVpp is measured at the output. Below is the schematic and oscilloscope waveform measurements.so could you help for analysing why this happened?

Parents
  • This might be a result of the measurement setup.. What's your scope's input impedance? Typical scopes have relatively high input capacitance... I'd also verify that C20 is in fact 0.1pF. Large capacitive paths to ground at the output can lead to a larger feedforward path in the amplifier that, at high enough frequencies, could result in this feedforward path dominating (rather than the feedback path). In other words, you're generating peaking in your transfer function which leads to oscillations.

    That's just a guess, but I just posted about this phenomenon - I observed peaking in simulations around 1GHz so I wanted to ask the community if they've experienced something like this and it seems you may have. My suggestion - try higher impedance scope probes first (i.e. connect a 10M resistor in series with your probe). If that resolves the issue (or shifts the oscillation frequency) then that's definitely the problem. If not, then there may be some other issue.

    Best,

    Andy

Reply
  • This might be a result of the measurement setup.. What's your scope's input impedance? Typical scopes have relatively high input capacitance... I'd also verify that C20 is in fact 0.1pF. Large capacitive paths to ground at the output can lead to a larger feedforward path in the amplifier that, at high enough frequencies, could result in this feedforward path dominating (rather than the feedback path). In other words, you're generating peaking in your transfer function which leads to oscillations.

    That's just a guess, but I just posted about this phenomenon - I observed peaking in simulations around 1GHz so I wanted to ask the community if they've experienced something like this and it seems you may have. My suggestion - try higher impedance scope probes first (i.e. connect a 10M resistor in series with your probe). If that resolves the issue (or shifts the oscillation frequency) then that's definitely the problem. If not, then there may be some other issue.

    Best,

    Andy

Children
No Data