fmak:
dBc is the relationship to the "carrier" amplitude. A measure that would give information about added spectral components in a frequency plot. Jitter (phase noise) are drifts in time and are easiest measured according to it's peak to peak RMS value in seconds. But the sidebands created by single-tone modulated jitter could be expressed as dBc, still this is not the way Tent does it.
dBc is the relationship to the "carrier" amplitude. A measure that would give information about added spectral components in a frequency plot. Jitter (phase noise) are drifts in time and are easiest measured according to it's peak to peak RMS value in seconds. But the sidebands created by single-tone modulated jitter could be expressed as dBc, still this is not the way Tent does it.
Jitter and phase noise are related. Jitter causes phase noise.
As asgard says, dBc is a relative measurement from the carrier (all dB measurements are relative to something) and is the defacto standard way of measuring phase noise in an RF system.
As asgard says, dBc is a relative measurement from the carrier (all dB measurements are relative to something) and is the defacto standard way of measuring phase noise in an RF system.
BlackCatSound said:
Hi Tom,
I'd be happy to see an unambiguous, matthematical relation between jitter and phase noise, because I can't correlate them.
best
BlackCatSound said:
yes, one of the many documents. the thing is that practical time domain measurements on oscillators reveal properties i don't see when measuring phase noise....
best
Guido, I have some quick questions related to the oscillators.
Do you plan to make available the frequencies 1.4112M and 5.6448M?
For these slower frequencies, and for the existing 2.8224M, is the oscillator actually operating at that frequency, or is it internally divided?
Whenever possible, I connect the osc. directly to the bck pin of the dac.
Thanks,
Alexandre
Do you plan to make available the frequencies 1.4112M and 5.6448M?
For these slower frequencies, and for the existing 2.8224M, is the oscillator actually operating at that frequency, or is it internally divided?
Whenever possible, I connect the osc. directly to the bck pin of the dac.
Thanks,
Alexandre
Asgard said:
And that is?
Cheers
spontaneous behaviour, non Gausian
If you have aTDA and you look at time versus time, you see the true behaviour of an oscillator
It is at least a bit weirrd to observe a phenomena in the frequency domain, when the time domain properties are of interest, no ?
It is like measuring the DC voltage of the mains: about zero, but still not safe to touch, nah ?
best
Alexandre said:
Hi
The answer is no. Both can e derived with a divide by 2 from stock components. It is not internally divided.
best
Guido Tent said:
I agree that a time plot could be more intuitive than a frequency plot. Still they are closely linked since one is the derivative of the other.
Are you saying that you have found clock timing jitter without a spectral density? 😕
Asgard said:
Spectrum analysers are integrators. Once energy appears outside the meaurement window, it is not seen in the measured result.
In this thread I have posted a few times that I measure in the time domain, you may want to ignore that but I won't repeat it anymore.
Errors in the time behaviour lead to jitter. Those trying to measure it in the frequency domain call it phase noise. Period.
Guido Tent said:
I have no objections to your methods of measuring nor am I ignoring them. I was only replying to your comparison with AC/DC in the wall socket. Interpreting it as a DC measurement (spectral domain) should show up empty while the voltage still swings (jitter in time domain).
Since you are an educated engineer we both know that jitter without a spectral density is pretty far out there 😀
I recently finished a survey of all available S/PDIF receivers and I intended to post the results in this thread. The mystery DAC is the reference design I used to test the receivers. I found one receiver that significantly outperformed all the others in every category but I can’t tell you which one it is. If everybody started using an S/PDIF received that provided very low jitter clocks without resorting to snake oil, it would put the clockmongers out of business and I would be permanently banned from this forum.
I learned the true purpose of this forum is to promote and protect the business interests of the clockmongers and kit peddlers. I thought Guido Tent’s post with links to his business web site violated the standing rules against advertising and I called his post to the attention of the moderators. For that, I was sin-binned.
Maybe I should become a clockmonger. Then I could post gratuitous advertising, give out bad advice, such as, insisting a ripple counter is the proper way to divide a clock, and everyone would love me.
I learned the true purpose of this forum is to promote and protect the business interests of the clockmongers and kit peddlers. I thought Guido Tent’s post with links to his business web site violated the standing rules against advertising and I called his post to the attention of the moderators. For that, I was sin-binned.
Maybe I should become a clockmonger. Then I could post gratuitous advertising, give out bad advice, such as, insisting a ripple counter is the proper way to divide a clock, and everyone would love me.
Ulas, stick to the technical stuff, the topic of this thread. Cease and desist. No further warnings.edit: some OT stuff removed.
No, Ulas, I Sinbinned you for being abusive. Despite the fact that you had annoyed me, I looked into your complaint, and as a consequence, another member received a warning. Since you like straight talking, you won't mind me being straight with you. Your post two posts up is at best incorrect, and at worst, malicious lies.
But you're quite right that a synchronous counter is what should be used as a clock divider.
- Status
- Not open for further replies.