Skip to main content

Log Detector

[caption id="attachment_894" align="aligncenter" width="540"] (almost) complete Log Detector[/caption]

I seem to have taken all my module photos before installing the voltage regulators. Anyway, other than that, this is the completed log detector. The log detector has a few responsibilities. In the standard spectrum analyzer, it's the last step in the "RF" signal chain. What that means is that on it's input, the final I.F. (intermediate frequency) is rectified into a DC voltage proportional to the log of the input. Because we use dB so extensively in characterizing RF circuits this is not only convenient, but it also assigns ADC bits evenly along the logarithmic range. It's not used in the basic spectrum analyzer, but in the VNA (vector network analyzer) the "limited-output" is used. This output is essentially the same as the input, but amplified to be at a constant output level. This is directly used by the Phase Detector of the VNA. The input to the log-detector IC is very high impedance, which is the purpose of the transformer seen on the right of the PCB.

[caption id="attachment_896" align="aligncenter" width="540"] Log detector test setup[/caption]

The testing configuration for the log-detector is shown above. I've soldered a coax cable from the output of the detector to the ADC. The center pin of the video filter switch is directly connected to the input, so I've attached the oscilloscope probe to it. Also, I've attached a small antenna to the input connector. I tested the detector with the input left open, capped with a 50 ohm dummy load, and with this antenna. I was trying to get an idea of what the noise floor is in each of these configurations.

[caption id="attachment_898" align="aligncenter" width="473"]![]( detector-dummy-load.jpg) Log detector output with 50 ohm dummy load[/caption]

There was no difference in noise floor between the dummy load and open input. Each configuration was close to, but a little less, than -100 dBm. Also, I had the video filter switch set to filter the input as much as possible. It should be noted that the average value with the log detector off is about -110dBm.

[caption id="attachment_897" align="aligncenter" width="473"]![]( detector-antenna.jpg) Log detector with antenna[/caption]

With the antenna attached, the noise floor rose to about -85 dBm. This is probably because of the stray EMI around my house. It was cool to see that the amplitude increased when my hand was near or touching the antenna.

[caption id="attachment_899" align="aligncenter" width="473"]![]( log-detector.jpg) Log detector with nearby transmitter[/caption]

I performed the above test on another day, so the noise with the antenna attached was a little different. Anyway, the point of this test was to observe whether the system was capable of detecting nearby radio signals. In this case, I'm keying a radio set to 50mW at 6 meters probably a foot away.

At this point, it appears that the log detector is in good working shape. I'm pretty happy about that, and I hope things keep going well.