Two points to note. First, the common SI5351A boards are DC coupled, so the chip tries to supply DC current into the 50 ohm load. Additionally if one uses a 50 ohm load on the output of the Si5351A it takes it out of the limiting range, so the even harmonics degrade. So if you AC couple the output then it will stop the DC issue, (say 100nF 0.1uF) and then if you add an additional series resistor, such as a 470 or 1 kohm, then the even harmonics will improve markedly, say to -50 dB or better. The datasheet states that it has lowish output Z, but it is a CMOS device, and should drive higher Z loads. Keep up the great videos
Thanks for your kind words! And also thanks for the additional notes. In practise amateur use, the output of the SI5351A will be added to an amplifier or mixer, indeed with low impedance (the datasheet mentions 85 Ohm). I will check your suggestions and share results in an other video.
One could use the series resistor driving into a 50 ohm gain block (either a MMIC or a discrete 50 ohm amplifier using the value of the series resistor to provide output power setting, or a biased emitter follower, with sufficient collector current to drive a 50 ohm load at the required RF power (A series output resistor of 39-47 ohms would then provide around 50 ohm source impedance).
Thanks again for your additional notes. About your last suggestion: indeed I did something similar in my 6 meter transceiver. I combined the series resistor with an attenuator (th-cam.com/video/jP5sGiyjWzc/w-d-xo.html ).
I was told that the capacitance of the scope probes accounts for rounding of the square wave as frequency increases. Your test supports this explanation as the spectrum analyser does not show any decrease in harmonics at higher frequencies, and it is the harmonics that form the square edges. Try the x10 setting on the probe and you will observe the waveform changes shape due to a different (lower?) capacitance. Another thing to try is to try different si5351 output settings, 2, 4 or 8MA. Keep up the interesting videos! You can learn a lot by simply measuring RF!
Thanks for your kind words Paul! Indeed capacitance of scope probes will influence what you see on the screen of the oscilloscope. In this specific case, the 50 Ohm cable runs from the SI5351 to the oscilloscope and is terminated there. The output impedance of the SI5351 is a bit higher but the mismatch is not that much. Settings (in the sketch) is a good idea to try, I did not do that. On the other hand, if the harmonics are not relevant, you likely want to have the most output. For example in my 6 mtr transceiver (50 MHz). Oscillator about 72 MHz, second harmonic 144. In combination with the bandpass filter (50 MHz) it has no influence (I think 😀).
Two points to note. First, the common SI5351A boards are DC coupled, so the chip tries to supply DC current into the 50 ohm load. Additionally if one uses a 50 ohm load on the output of the Si5351A it takes it out of the limiting range, so the even harmonics degrade. So if you AC couple the output then it will stop the DC issue, (say 100nF 0.1uF) and then if you add an additional series resistor, such as a 470 or 1 kohm, then the even harmonics will improve markedly, say to -50 dB or better. The datasheet states that it has lowish output Z, but it is a CMOS device, and should drive higher Z loads.
Keep up the great videos
Thanks for your kind words! And also thanks for the additional notes.
In practise amateur use, the output of the SI5351A will be added to an amplifier or mixer, indeed with low impedance (the datasheet mentions 85 Ohm).
I will check your suggestions and share results in an other video.
One could use the series resistor driving into a 50 ohm gain block (either a MMIC or a discrete 50 ohm amplifier using the value of the series resistor to provide output power setting, or a biased emitter follower, with sufficient collector current to drive a 50 ohm load at the required RF power (A series output resistor of 39-47 ohms would then provide around 50 ohm source impedance).
Thanks again for your additional notes.
About your last suggestion: indeed I did something similar in my 6 meter transceiver. I combined the series resistor with an attenuator (th-cam.com/video/jP5sGiyjWzc/w-d-xo.html ).
I was told that the capacitance of the scope probes accounts for rounding of the square wave as frequency increases. Your test supports this explanation as the spectrum analyser does not show any decrease in harmonics at higher frequencies, and it is the harmonics that form the square edges. Try the x10 setting on the probe and you will observe the waveform changes shape due to a different (lower?) capacitance.
Another thing to try is to try different si5351 output settings, 2, 4 or 8MA.
Keep up the interesting videos! You can learn a lot by simply measuring RF!
Thanks for your kind words Paul!
Indeed capacitance of scope probes will influence what you see on the screen of the oscilloscope. In this specific case, the 50 Ohm cable runs from the SI5351 to the oscilloscope and is terminated there. The output impedance of the SI5351 is a bit higher but the mismatch is not that much.
Settings (in the sketch) is a good idea to try, I did not do that.
On the other hand, if the harmonics are not relevant, you likely want to have the most output. For example in my 6 mtr transceiver (50 MHz). Oscillator about 72 MHz, second harmonic 144. In combination with the bandpass filter (50 MHz) it has no influence (I think 😀).