To find the root cause of the spurs, I worked in 2 steps:
?
1. I took the front panel board out of the case and connected it with jumper wires to the main board. My objective was to eliminate possible EMI from the Pi Pico into the RF board.
?
This did not make a difference for the spurs, as they just went down by 0.5dB which is the resolution of the TinySA.
?
2. In the si5351v2.c code I changed 8MA to 2MA. This results in approximately 5dBm less output power (33.5 vs 38.1dBm) but also a relative reduction of the spurs by approximately 7.5dB from -28.1dBc to -35.6dBc:
?
?
Reduced output power is most probably due to the fact that the zBitx uses a diode ring mixer instead of the KISS-Mixer used in the sBitx and the oscillator power to the ring mixer is provided directly by the SI5351 CLK2 output (in reality an MS5351). Also, the IF diode ring mixer gets its oscillator signal directly from the SI5351 CLK1, whereas in the sBitx there is a buffer amplifier with two 2N3904.
?
It could be that the much higher load on the SI5351 outputs, results in some intermodulation in the SI5351, which finally shows up as spurs.
?
Given that these spurs are a showstopper, to transmit with the zBitx as the spurs are not at least 40dB suppressed, I am looking forward to further investigation and direction by hfsignals.
?
On Mon, Apr 14, 2025 at 11:39 PM, Ron, N6YWU wrote:
If zbitx/si5351v2.c is the correct place to look, then in lines 243, I see the variable drive_current being set to 2ma. But then that variable is being ignored, and the Si5351 is being configured in line 249 :
?
set_freq_fixeddiv(clk, pll, frequency, pll_div,?
? ? ? ? ? ? ? ? ? ? SI5351_CLK_DRIVE_STRENGTH_8MA);
?
Changing this to 2 mA might help reduce the spurs according to KX4Z (Thanks!).
?
- Ron
?
?
--
73, Joerg - DB2OO
