Open KonradMcClure opened 4 years ago
Frequencies can be seen in the images above. They average around 14.8kHz, though the 90deg baseband appears to be at 15.3kHz...
Did some testing to see what might be the cause. I dropped the RF frequency down to 5MHz, and the LO frequency to 5MHz + 10.7kHz. Reading the Baseband 180deg and the LO_I test pins I saw this:
Since the difference in LO - RF (5.0133 - 5.0 MHz) can be seen in these images to indeed be the the frequency of IF (13.3 kHz), we can rule out it's the MUX that's causing this issue.
Here I've taken a reading of the LO_I pin and CLK0 pin on the same settings. I'm stretching the sample rate of the Digilent Analog Discovery 2, which sits at a max of 100MHz, but it's relatively clear. We would expect that the CLK0 pin would read as 4(5,000,000 + 10,700) = 20.042MHz which is what was put in the arduino code for the si5351 test. However over time we see it average out to 20.022 MHz (running longer than the image above). We only seem to be getting half of the 410,700 that we would expect. Luckily the Johnson counter seems to be pretty faithfully dividing by 4 given that 20.022/5.0134 = 3.994
I set up the other outputs of the si5351 to see if they all read the same. It appears the Analog Discovery 2 reads a different value depending on whether my other probe is reading the other clock signals....
Or the local oscillator...
Or nothing on the other probe...
I think my equipment just might not be good enough for this
The bandbase signals (0, 90, 180, 270 degrees) should come out of the MUX at 10.7kHz, yet they appear to be around 14-15kHz. This could be an issue with clock generator, speed of the johnson counter circuit, or the speed of the multiplexer itself...