The best video on phase noise. One thing not mentioned is that spectrum analysers can only be used to measure phase noise where the phase noise of the DUT is worse than the phase noise of the spectrum analyser. This is a problem for low-phase noise oscillators, hence the need for cross-correlation phase noise analysers.
Thanks! And yes, you're absolutely correct. I did a separate video on cross-correlation since it is, as you mention, a very important topic when measuring low-phase noise devices.
I can't thank you enough really! Every time i search for something on the internet that i don't understand it that well, no matter how long I look, i always end in one of your videos! Thank you alot!
A clear and concise presentation that I found it technically informative and easy to absorb as an introduction. Where do I find this short powerpoint presentation for future reference?
@2:36 the 2nd plot may i know what the label of the y axis ? Is it Power ? Also @2:36 2nd plot may I know mathematically how did the time changing phase cause spectral broadening in other words how did we get this skirt shape Vs Freq ? @3:57 you mentioned that it is generally easier to work with less Freq signals , may I know why ?
Those are all great questions. The y-axis of the frequency domain on slide 3 and 4 is power, usually measured in dBm. The math behind phase noise measurement is a bit too complicated to type in a TH-cam comment :) Conceptually, it might help to imagine that the sidebands or skits are a probability density function (PDF). If the oscillator were perfect, the frequency would be constant and the PDF would be a single line. A real oscillator might have small frequency variations away from the nominal frequency most of the time, and larger excursions from the carrier less often. This isn't really exactly how it works (mathematically), but it might help in explaining the shape of the phase noise sidebands or skits. There are lots of reasons why it is easier to work at lower vs. higher frequencies. Passive components (resistors, capacitors, inductors) and even wires or traces start behaving non-ideally as frequency increases. For example, wires start developing inductance at higher frequencies and capacitors also start developing a series inductance and resistance: this greatly complicates circuit design and analysis. Higher frequency components also cost more than lower frequency components :) There are many more examples, but the use of downconversion is extremely common in RF applications. Thanks again for the questions! Hope that helps.
@@pauldenisowski you don't want to get down too low in the frequency realm or you open yourself up to potential interference from DCDC power supplies. Also large power electronics can resonate between 1-100MHz, CPU clock signals, etc. Downsizing is very sensitive to the rest of the system and it's quite fascinating 😀
thank you for this video, but I have a question. how much does the power of carrying wave? If i gain the power of carrying wave, phase noise will change or fix?
Great, clear and to the point educational videos to get the basic understanding. One question if I may. Older spectrum analyzers do not go so low in RBW to 1Hz, nor have automated measurement of phase noise. When doing manual measurement with specific RBW (for example mine goes to 200Hz) do you divide the measured power (at some specific offset, like usual 10kHz) in dBm by 200 to get [dBm/Hz]? Thanks
Oh, I was right, only one problem, you need to divide the measured power in [W] by the RWB _used and not in [dBm]! dB's are (of course) all about "adding and subtracting".
It's a great video. Thank you a lot! I also see the term "phase error" in the Specifications of R&S products, so what is the difference between "phase error" and "phase noise"?
Thanks. Usually "phase error" means "deviation from ideal phase angle", for example, in digital modulation (See the video "Understanding EVM"). Basically, the received symbol's phase differs by this amount from what it should be. Phase noise is unintentional variation of a signal and, as far as I know, is what is specified for instruments. If you have a particular R&S document in mind, please send me a link and I'd be happy to take a look. Thanks again!
That's a complicated question :) There are some applications that are relatively immune to even very high levels of phase noise, and there are some applications where even very low phase noise might cause problems. Generally speaking, higher modulation orders that use phase modulation (PSK, QAM, APSK) have stricter phase noise requirements: e.g. 64QAM requires much lower phase noise than QPSK. Phase noise is usually reported as curves (like the one shown in this presentation) so that you can see what the phase noise is at the frequencies that are interesting / important for your application. If you haven't already, you might want to watch the videos "Understanding APSK and QAM", "Understanding EVM" and "Interpreting Constellation Diagrams" for a better, more visual, explanation of what phase noise can do in modern digital modulation systems. One last note: most standards, like 5GNR, 802.11 (Wi-Fi), etc. specify required EVM levels (in %, for a given modulation order) rather than permissible phase noise. Hope that helps!
Usually they are (a) residual PM in degrees, (b) residual AM or FM in hertz, and (c) jitter, in seconds. There's actually a video coming out in the next few days that shows these measurements (and their units) on our phase noise tester. Longer term, I'm also planning a video on integrated phase noise measurements (a topic that really needs a video of its own). Thanks for the question!
@@pauldenisowski I would definitely appreciate the video on integrated phase noise measurements - in particular how to derive jitter (in units of secs from phase noise in units of dBm)
The best video on phase noise. One thing not mentioned is that spectrum analysers can only be used to measure phase noise where the phase noise of the DUT is worse than the phase noise of the spectrum analyser. This is a problem for low-phase noise oscillators, hence the need for cross-correlation phase noise analysers.
Thanks! And yes, you're absolutely correct. I did a separate video on cross-correlation since it is, as you mention, a very important topic when measuring low-phase noise devices.
I can't thank you enough really!
Every time i search for something on the internet that i don't understand it that well, no matter how long I look, i always end in one of your videos!
Thank you alot!
Thanks! I'm working on a lot more videos, so please stay tuned!
@@pauldenisowski Great!, ty!
Thank you for this amazing way to present in an easy way such technical information. Keep going !
Thanks! There are other phase noise related videos coming in the next few weeks, so please stay tuned!
A big like to the video! And an even bigger thank you to their creators!!
Thank you!
A clear and concise presentation that I found it technically informative and easy to absorb as an introduction. Where do I find this short powerpoint presentation for future reference?
A very technical topic explained beautifully and in an easy to understand manner!! Thankyou!!
Thanks!
Great video, very good resume of Phase Noise!!!
Thank you!
It was great to learn phase noise in a very simple and nice way. After lot of efforts finally i got this video, Its really great.
Thanks!
Great lecture! I am also waiting on Allan variance topic. Thanks!
Thanks! Allan variance is on the to-do list :)
Amazing presentation
Thanks for the excellent explanation. Loved it and was motivated to explore more..
Thank you!
Perfect video, very easy to understand
Thank you!
@2:36 the 2nd plot may i know what the label of the y axis ? Is it Power ?
Also @2:36 2nd plot may I know mathematically how did the time changing phase cause spectral broadening in other words how did we get this skirt shape Vs Freq ?
@3:57 you mentioned that it is generally easier to work with less Freq signals , may I know why ?
Those are all great questions. The y-axis of the frequency domain on slide 3 and 4 is power, usually measured in dBm.
The math behind phase noise measurement is a bit too complicated to type in a TH-cam comment :) Conceptually, it might help to imagine that the sidebands or skits are a probability density function (PDF). If the oscillator were perfect, the frequency would be constant and the PDF would be a single line. A real oscillator might have small frequency variations away from the nominal frequency most of the time, and larger excursions from the carrier less often. This isn't really exactly how it works (mathematically), but it might help in explaining the shape of the phase noise sidebands or skits.
There are lots of reasons why it is easier to work at lower vs. higher frequencies. Passive components (resistors, capacitors, inductors) and even wires or traces start behaving non-ideally as frequency increases. For example, wires start developing inductance at higher frequencies and capacitors also start developing a series inductance and resistance: this greatly complicates circuit design and analysis. Higher frequency components also cost more than lower frequency components :) There are many more examples, but the use of downconversion is extremely common in RF applications.
Thanks again for the questions! Hope that helps.
@@pauldenisowski you don't want to get down too low in the frequency realm or you open yourself up to potential interference from DCDC power supplies. Also large power electronics can resonate between 1-100MHz, CPU clock signals, etc. Downsizing is very sensitive to the rest of the system and it's quite fascinating 😀
thank you for this video, but I have a question. how much does the power of carrying wave? If i gain the power of carrying wave, phase noise will change or fix?
Great, clear and to the point educational videos to get the basic understanding. One question if I may. Older spectrum analyzers do not go so low in RBW to 1Hz, nor have automated measurement of phase noise. When doing manual measurement with specific RBW (for example mine goes to 200Hz) do you divide the measured power (at some specific offset, like usual 10kHz) in dBm by 200 to get [dBm/Hz]? Thanks
Edit: I just found my answer in the next video 😁 It's not just diving, it is subtracting the 10log(RBW_used).
Oh, I was right, only one problem, you need to divide the measured power in [W] by the RWB _used and not in [dBm]! dB's are (of course) all about "adding and subtracting".
It's a great video. Thank you a lot! I also see the term "phase error" in the Specifications of R&S products, so what is the difference between "phase error" and "phase noise"?
Thanks. Usually "phase error" means "deviation from ideal phase angle", for example, in digital modulation (See the video "Understanding EVM"). Basically, the received symbol's phase differs by this amount from what it should be. Phase noise is unintentional variation of a signal and, as far as I know, is what is specified for instruments. If you have a particular R&S document in mind, please send me a link and I'd be happy to take a look. Thanks again!
Thank you. Finally i understood clearly.
Thanks for the feedback!
Very clear and helpful! Thanks!
Thank you!
Very nice presentation 👏 👌 👍 😀
Appreciate the feedback!
great explanation. Really well done.
Thanks for the feedback -- very much appreciated!
Do you have a video about Alan variance?
Not yet, but it's on my to-do list for 2023. I'm also hoping to do some additional content on other phase noise related topics.
Very well explained
Thanks you!
Very nice presentation 🤔👍👍🧘♂
Thank you!
question, how to determine the phase noise signal level it good or bad. what the limit.
That's a complicated question :) There are some applications that are relatively immune to even very high levels of phase noise, and there are some applications where even very low phase noise might cause problems. Generally speaking, higher modulation orders that use phase modulation (PSK, QAM, APSK) have stricter phase noise requirements: e.g. 64QAM requires much lower phase noise than QPSK. Phase noise is usually reported as curves (like the one shown in this presentation) so that you can see what the phase noise is at the frequencies that are interesting / important for your application. If you haven't already, you might want to watch the videos "Understanding APSK and QAM", "Understanding EVM" and "Interpreting Constellation Diagrams" for a better, more visual, explanation of what phase noise can do in modern digital modulation systems. One last note: most standards, like 5GNR, 802.11 (Wi-Fi), etc. specify required EVM levels (in %, for a given modulation order) rather than permissible phase noise. Hope that helps!
Perfect explanation. Thank you!
Thanks for the feedback!
On point. I am in love.
Stay tuned for many more phase noise videos in the next few weeks! :)
What are the units of integrated phase noise?
Usually they are (a) residual PM in degrees, (b) residual AM or FM in hertz, and (c) jitter, in seconds. There's actually a video coming out in the next few days that shows these measurements (and their units) on our phase noise tester. Longer term, I'm also planning a video on integrated phase noise measurements (a topic that really needs a video of its own). Thanks for the question!
@@pauldenisowski I would definitely appreciate the video on integrated phase noise measurements - in particular how to derive jitter (in units of secs from phase noise in units of dBm)
ty )