Post Go back to editing

Log Detector Output

Thread Summary

The user is experiencing a high DC offset (1V) in the output of an AD8309 log amplifier when measuring a 60-66MHz input signal centered at 63MHz. The support engineer suggests this may be due to noise or interference, recommending the addition of power supply decoupling capacitors and ensuring the part is shielded. The output voltage in the datasheet is measured relative to ground, and the limiter is disabled, which is good for maximum sensitivity.
AI Generated Content
Category: Hardware
Product Number: AD8309

Hello,

We have implemented the following circuit with AD8309. We intend to obtain the Vlog of the input signal in the range of 60MHz to 66MHz at the center frequency of 63MHz. The input signal is as follows: 

The corresponding output signal is as follows: 

The circuit that we have used is as follows & the inductor & capacitor values used in the front end are corresponding to 60MHz: 

We had the following questions: 

a. The output of the log amplifier has a DC offset. Theoretically this should exist, but this offset in our circuit is high (1V). Is this normal ? If not, are there any techniques to reduce it ? 

b. The graph provided in the figure 8 of the AD8309 datasheet of RSSI output versus input level in dBm shows the output ranging from 2.4V for 20dBm to about 0.4V for -90dBm. Does this consider the DC offset or is it a peak-to-peak measurement? We were able to get similar graph if we consider the DC offset as well.

Request you to provide some suggestions on the above as this is required for a customer project.

With Regards,

ARK

  • Case CS-521851-K1N4P9 has been canceled in MSD
  • With no signal at the input, the AD8309 will should have a dc offset of around 400 mV. You can see this in Figures 7, 8 and 9 in the datasheet. 

    I would not call the 1V you are seeing, a dc offset. What I suspect is going on is that the part is detecting and measuring noise or an interference signal. A 1V output corresponds to around -45 dBm of input signal which is around 1.3mV of rms noise. To mitigate this you need to eliminate the noise or shield the part. I would experiment with putting  power supply decoupling capacitors on the individual power supply pins (in case RF is getting in through the power supply pins). Another thing you can try is to turn off you main RF signal source. If the 1V is still present, that means that there is interference coming from somewhere else.  

    I note that you have the limiter disabled. This good practice if you are targeting maximum sensitivity from the circuit. When the limiter is enabled, the large limiter output signal can leak through the board, back to the input and degrade sensitivity. But you have the limiter disabled so this mechanism is not at play here. 

    Regarding your second question, all output voltages dc voltages, measured relative to ground. 

  • Thanks Enash,

    We will try experimenting with the decoupling capacitors and also check output with no signal input. We have isolated the power supply for the log detector from that of the other digital section parts to mitigate the noise through the ground. There is no other RF input source except for the log detector input. 

    Regards,

    ARK

  • Hi @enash

    We added a 10uF capacitor in parallel to the existing one for power supply & measured the output power. There was a DC offset of around 383mV. 

    However, when the RF input of 10dBm was provided, this offset would increase to 1V as stated earlier: 

    Could this be due to the input from the instrument itself?

    Regards,

    ARK

  • Hi   : Could you please provide any suggestions?

    Regards,

    ARK

Before You Switch


Switching languages will make ADI Explorer unavailable. Resume your session by switching back to English and reopening ADI Explorer.