开云体育

Hombrew sBitx Frequency Cal - a tale of 3 Frequencies

#homebrew #sBitx

Dean Souleles All Messages By This Member #108948 Hi all - Seeking wisdom (to borrow a term from FFT lore) from the crowd about a frequency problem wiht my Homebrew SBITX.??if you happened to listen Soldersmoke #250 I talked about it there and I'll illulstrate it here.? First I should explain the diffferences between my build and the HF Signals build.? My digital board consists of a Raspberry Pi 3b+ and a WM8731 codec prototype board from .? My SI-5351 an Adafruit style module which does not have a precision tuned crystal referency oscilaltor and therfore it requires a crystal calibration offset. First issue - I zero beat 10MHz WWV and used the displayed value to calculate the crystal calibration offest I needed for the frequeny display to be exactly 10MHz.? After adjustging the offset the radio is perfectly alligned at 10MHz.? To linearity I set the frequency to 15MHZ and found the WWV center frequency was approximately 200Hz too high (15.000.200). I then tuned to 5MHz and the center frequeny was approximately 200 Hz too low (4.999.800).? ?Pictures below (signal in the picture is from a signal generator, not WWV, to make it easier to see. \ Does anyone have any thoughts on why the display frequency should vary so much from the tuned frequency and why the delta would vary with the HF frequency? Second issue - in testing the codec prototype board stand-alone (that is disconnected from my analog receiver),? I injected a 24KHz sine wave into the line-in and expected to see a peak at exacty 0 on the waterfall display,? I did not. The waterfall shows the peak about 600Hz below the waterfall center point.? In discussing with Farhan the best guess is the audio prototype board crystal reference is off frequency.? ?I'm going to try a different card to see what happens, Would appreciatte any thoughts on either of the above. 73, Dean KK4DAS
Gordon Gibby KX4Z All Messages By This Member #108949 开云体育 Can you command the audio board to generate a 4 kHz tone and measure it on the waterfall? That might allow you to check the crystal.?? Or do the reverse ? Gordon Kx4z? toggle quoted message Show quoted text On Feb 12, 2024, at 14:13, Dean Souleles <dsouleles@...> wrote: ?Hi all - Seeking wisdom (to borrow a term from FFT lore) from the crowd about a frequency problem wiht my Homebrew SBITX.??if you happened to listen Soldersmoke #250 I talked about it there and I'll illulstrate it here.? First I should explain the diffferences between my build and the HF Signals build.? My digital board consists of a Raspberry Pi 3b+ and a WM8731 codec prototype board from .? My SI-5351 an Adafruit style module which does not have a precision tuned crystal referency oscilaltor and therfore it requires a crystal calibration offset. First issue - I zero beat 10MHz WWV and used the displayed value to calculate the crystal calibration offest I needed for the frequeny display to be exactly 10MHz.? After adjustging the offset the radio is perfectly alligned at 10MHz.? To linearity I set the frequency to 15MHZ and found the WWV center frequency was approximately 200Hz too high (15.000.200). I then tuned to 5MHz and the center frequeny was approximately 200 Hz too low (4.999.800).? ?Pictures below (signal in the picture is from a signal generator, not WWV, to make it easier to see. \ Does anyone have any thoughts on why the display frequency should vary so much from the tuned frequency and why the delta would vary with the HF frequency? Second issue - in testing the codec prototype board stand-alone (that is disconnected from my analog receiver),? I injected a 24KHz sine wave into the line-in and expected to see a peak at exacty 0 on the waterfall display,? I did not. The waterfall shows the peak about 600Hz below the waterfall center point.? In discussing with Farhan the best guess is the audio prototype board crystal reference is off frequency.? ?I'm going to try a different card to see what happens, Would appreciatte any thoughts on either of the above. 73, Dean KK4DAS
Evan Hand All Messages By This Member #108954 Hi Dean, I burned the Si5351 in my DE.? I repaired it with an Adafruit clone board.? I found two lines (45 and 46) in the si5351v2.c program. ? ? ?//int xtal_freq_calibrated = 25012725; // crystal oscillator? ? ? ?int xtal_freq_calibrated = 25000000; // tcxo I uncommented the first and commented the second.? I still had to use different values of the xtal_freq_calibrated = line to get the sbitx frequency to agree with my CW frequency counter.? You can do the same with another receiver or spectrum analyzer. 73 Evan AC9TU
Dean Souleles All Messages By This Member #108958 On Mon, Feb 12, 2024 at 05:12 PM, Evan Hand wrote: I burned the Si5351 in my DE.? I repaired it with an Adafruit clone board.? I found two lines (45 and 46) in the si5351v2.c program. ? ? ?//int xtal_freq_calibrated = 25012725; // crystal oscillator? ? ? ?int xtal_freq_calibrated = 25000000; // tcxo Thank you Evan - I have already done exactly as you suggested.? That is how I got the receiver display frequency to align at? exactly 10MHz.? But it is still off at 5MHz and 15MHz and not linearly, its more like a bell curve. I've done much more analysis than I described at first.? I have?measured the BFO to make sure it is exactly on 40.035 - and I still see a delta between the displayed frequency and the tuned frequency that vaires with frequeny as you saw in the original post.? I have looked at both clocks with multiple frequency counters, the FFT mode of the Rigol scope, and the TinySA Ultra and both clocks are all fine and on frequency with no stray harmonics.? ?The results don't make sense but they are what they are.? There is something deeper going on. 73, Dean KK4DAS ?
Dean Souleles All Messages By This Member #108960 On Mon, Feb 12, 2024 at 02:22 PM, Gordon Gibby wrote: Can you command the audio board to generate a 4 kHz tone and measure it on the waterfall? That might allow you to check the crystal.?? ? Or do the reverse ? Hi Gordon - I'm trying to understand what you are suggesting. Something like generate a tone through the DAC to line-out and loop it back to line-in and look at it on the waterfall? .??
Gordon Gibby KX4Z All Messages By This Member #108961 开云体育 After getting that board to generate the tone I would measure its frequency with a different computer. ?Just measure the audio frequency toggle quoted message Show quoted text On Feb 12, 2024, at 20:38, Dean Souleles <dsouleles@...> wrote: ?On Mon, Feb 12, 2024 at 02:22 PM, Gordon Gibby wrote: Can you command the audio board to generate a 4 kHz tone and measure it on the waterfall? That might allow you to check the crystal.?? ? Or do the reverse ? Hi Gordon - I'm trying to understand what you are suggesting. Something like generate a tone through the DAC to line-out and loop it back to line-in and look at it on the waterfall? .??

Previous Topic Next Topic