Topics

U3S FSKCW wave form 'jitter'

m0bmn
 

 

Hi all

One of my u3S as been working great since it was put into service only a month or so ago,but I have never really been happy with the waveform when I compare it to others that I see, its only a little thing and you need to look close to see it but it has little steps sometimes in the freq (very small part of a hert about 1/10th I guess but other peoples traces are nice and clean , mine isn’t I think, I have 3 different synth modules to try, one standard, one txco and one txco with a heater fault but still doesn’t drift at all, I must say on all of them I get no drift reports on WSPR after 30mins or so , its just I would like my FSKCW trace to be nice and clean like others I see, what could be the problem, I have a 40m one on the same PSU and that runs great with a very clean waveform with its standard synth module.

Anyone any ideas? In the past I have killed a few BS170’s on this board but that was my fault and they have all been swapped out and power out is fine.

73 Paul M0BMN

 

Sent from Mail for Windows 10

 

Alan G4ZFQ
 

I have never really been happy with the waveform when I compare it to others that I see, its only a little thing and you need to look close to see it but it has little steps sometimes in the freq (very small part of a hert about 1/10^th
Paul,

You mean 0.1Hz?
I wonder how you see that reliably?
Exactly when do you see the variations? On a constant carrier or as the frequency changes?
Presumably it is due to the reference crystal shifting frequency?
That is far less likely with a TCXO, separate oscillator, but possibly small shifts are normal with a standard Si5351-driven crystal.
There have been discussions about the reason, I favour tiny shifts in the crystal loading, but temperature effects can occur.

73 Alan G4ZFQ

geoff M0ORE
 

If it is the last digit changing on a frequency counter that you are seeing, then ignore it.

The final digit on a frequency counter can vary plus or minus one count depending where the signal being measured is on its cycle when the clock pulse starts the count. For example, if the counter is set to start counting on the rising edge and the signal under test is past the peak, it doesn't start the count until the signal starts its rising part of the cycle. Hence you have lost one cycle. To see the variation of 0.1Hz with amateur equipment is almost impossible.

Geoff

On 2/27/2019 4:12 PM, Alan G4ZFQ wrote:
I have never really been happy with the waveform when I compare it to others that I see, its only a little thing and you need to look close to see it but it has little steps sometimes in the freq (very small part of a hert about 1/10^th

Paul,

You mean 0.1Hz?
I wonder how you see that reliably?
Exactly when do you see the variations? On a constant carrier or as the frequency changes?
Presumably it is due to the reference crystal shifting frequency?
That is far less likely with a TCXO, separate oscillator, but possibly small shifts are normal with a standard Si5351-driven crystal.
There have been discussions about the reason, I favour tiny shifts in the crystal loading, but temperature effects can occur.

73 Alan G4ZFQ