Вставка
- Опубліковано 17 чер 2024
- In this episode Shahriar demonstrates the fundamentals of Phase Noise. The theory behind phase noise is presented both from a time-domain and frequency domain points of view. Various mathematical equations which represent phase noise behavior is also presented.
Phase noise measurement challenges and techniques are covered. Capabilities and limitations of different instruments for phase noise characterization is explained. The experiments show AM/PM modulations both in time and frequency domain and their relationships to phase noise measurements. Using a spectrum analyzer the phase noise of several synthesizers are shown. However, it is clear that the phase noise of an ultra-clean 1-GHz reference signal cannot be measured using traditional methods.
The concept of cross-correlation is introduced as a means of measuring phase noise below the phase noise of the measurement instrument. The block diagram of the an Agilent Phase Noise Analyzer is presented followed by a short teardown. The measured phase nose of the ultra-clean 1-GHz source is shown using correlation techniques which matches the datasheet.
www.TheSignalPath.com
/ thesignalpath
www.Patreon.com/TheSignalPath - Наука та технологія
I really don't understand how you have less than 100k subscribers , while others very basic electronic channel have a way more ,every Real electronic engineer in this planet should subscribe to this channel
You deserve more subs. Going to need to rewatch this to get a better understanding afterwards.
Due to the nature of the advanced topics I cover, my subscription will always be limited. In fact, my most popular videos are the introductory ones.
The Signal Path ... nevertheless, thanks for spending time producing the content 👍👌😉
@@Thesignalpath you should make playlist, that will be good for us
@@Thesignalpath This demonstrates the human being more concerned with beliefs than with technology.
@@Thesignalpath OG Shahriar, We all love you.
At 43:00, so cool, the two OCXOs are shock-mounted on different axes! This must be to ensure that the vibration they encounter is not correlated. Thank you so much for the video, it was most informative.
I like these tutorials! Really helps when you visualize these things on your equipment.
Any possibility of an EMC pre-compliance testing video in the future? 😁
This is actually worth my time
Another excellent video! Thanks!
Thank you very much Dr. Shahriar , please we need more videos like this
Thanks for bringing in tutorials on fundamentals. You're the best to learn them from
amazing as usual...thanks for the super knowledge that only exists on your channel .
Thank you for this video - I really enjoyed the video, especially after noticing the time and date on some equipment was today. The tidbit about the rubber feet on the two LOs was very interesting. I happened to be doing some LTE measurements with my instrumentation atop a rattling temperature chamber - I don't think I'll be doing that again
You have a knack for doing videos on *exactly* what I want to learn! Some aspects of understanding phase noise have been kicking my butt and it's very clear now. THANK YOU! I saved this video in my future-reference list.
I'd learn more sitting next to Shahriar on a long flight than a month in the library. Great work and thank you for sharing!!
Thank you for this video you have given me a greater understanding of phase noise and to see the measurment instruments is fantastic.
Thank you Shahriar! A great class! The way I see it, and as you have pointed out, the nature of more advanced topics is such that the deeper you got, the fewer the likes. The usual "like" measuring system fails to your neat, careful and great content! I hope the feedback from the (technical) community and from us, other engineers, may somehow convey our liking and gratitude to your work!
Brilliant tutorial, thanks for your hard work preparing this :)
This is a POWERFUL vid as many coming from you. THANKS a lot make & share that information.
Blessings..🎩!!!
Congrats! As an Electronic Engineer I LOVED your video and the clarity on the explanation as you go along. Keep up the good work!
Thank you for yet another superb video. Just wanted to let you know that I greatly appreciate your content! Take care.
Just thanks shahrihar.
Professors can't explain it simpler than you did.
Hoping to see PLL, jitter, noise figure and other tutorials.
Thanks again.
I really like this format. I think that the video i enjoyed the most on your channel is the one on delta sigma :)
Glad I stuck around till the end both for the cat and the explanation of how phase noise is relevant to real world applications. Thank you!
This was wonderful. Its pure gold. Thank you so much .
Thank you.
Really, really good work. The mix between theory and practical example is excellent. Thanks so much for your work!
Haven't watched it yet, but already know it'll be good. :)
Ohhhh YES. Pre-emptive thumbs up for this undoubtedly awesome and educational video. You sir, are a god. Thank you
I have been going through a lot of contents and videos on phase noise and this video has the best explanations. I will be watching this video a few more times as there is a lot of information you passed on and to make sure I have not missed any of them. Thank you for sharing!
Excellent tutorial and demonstration with the right balance of theory and hands-on demo. I finally learned about PN concepts that had tried to grasp earlier but fell short. Really appreciate your high quality tutorial videos which tackle explaining difficult electrical engineering topics. Thank you.
You are a genius in explaining this topic in a very concise manner, demonstrating it on actual test equipment, and adding the very important 'cat scan' to this presentation. Most excellent from start to finish. You kept my attention throughout. Thanks!!
Great tutorial! I'm going to recommend this to my (ambitious) techs! I've had most of HPAKs Phase Noise Test Sets - the HP3048, E5505, and we just took delivery of the (single channel) N5511. All exceptional instruments! This is a difficult (& very expensive!) measurement to do and understand correctly!
This is by far the best video for phase noise explanation. Thank you for sharing these information.
Really impressed by the amount of knowledge that you have shared here. Learned a lot of things being a post-silicon validation engineer. Thank you!
Thank you for very detailed and informative tutorial on phase noise. I intuitively sort of knew what it is about, but this really helped with my understanding.
Excellent tutorial ! Also, cross correlation is a technique used extensively in radio astronomy. The noise from individual LNAs following antennas are uncorrelated but the sky signals picked up by antennas are correlated. In fact almost all modern synthesis radio telescopes have massive correlators to process huge volumes of data !
happy new year , Shahriar:-)
Thank you very much. Your videos are very helpful. Thank you for sharing the knowledge.
Nice to see a video from you!
Great video. Looking forward to the ones about jitter and pll!
Today is UA-cam day :D Nice timing!
Great tutorial! Thank you!
Great video. I really enjoyed this and found it very useful. Excited for jitter! Another video that could be cool is symbol generator to IQ mixer to constellation plot. Then showing how different properties of the system effect the bit error rate.
I am RF engineer involved in testing some oscillators. Hope to get a chance to learn abt PLL from you in the future :) Definitely lucky to have you sharing knowledge in this field.
wow, new TSP's video : ))
Totally reccomend it for understanding Phase modulation
Thanks for the excellent tutorial.
That was amazing bro!
Awesome video, I quite enjoyed it. Kinda makes me feel smarter than I really am. I mean, as I follow along everything seems crystal clear, but if I had to explain this to someone else I would probably have some difficulty ;-)
You have fixed the audio level differences you had in some of your previous videos (bench vs. PC), which is nice.
Great video! Thanks!
Thanks for this video. I am looking for a 10 Mhz reference for the lab and this was exactly what I needed to get my head around dBc/Hz metric so I know what I am looking at.
Please keep making tutorials!
Helpful video. I like it
Nice job 👍 Thank you sir 🙏
So cool! Thank you!
Thank You for the knowledge. After 2nd watch I got it :)
Exquisite!
After watching your videos, I picked up a broken R&S VNA at a gov auction...and I fixed it (bad attenuator switch) ! However now, I am finding out that I do not know how to use it. e.g. I did not know that S12 should be read backwards (meaning from port 2 to port1), I am murky on Return Loss measurements, its relation to SWR and why does a shorted coax give me a decaying wavy trace (I still do not understand how to measure the loss of my coax cable) ...and many many more concepts. w2aew's video explaining the dBC and dBm helped me a lot ...but there is still so much more VNA jargon and concepts to be learned !
Instrument with 7GHz bandwidth:
"That is OK."
Great video and explanation, thank you very much for this basics tutorial!
Thanks and happy new year
Hell yes. Thank you! Do you still plan to share a tutorial on PLLs?
Yes, but I needed to do this one first in order for the PLL one to have some solid background.
I wish this video was here when I did my PhD!
Thanks Shariar...added some good notes to the PN section of my DSA-1030A-TG (good enough for what I do!)...still would like to find that 3-pin probe port connector that appears on the front of just about every instrument but no source to date. Keep up the great work!
Thank you for that.
One of these days I'm hoping you'll explain to us why QAM constellations are sometimes "rotated" - apparently there are some benefits from doing this? Great video by the way!
happy new year！
Perfect.
Measuring phase noise to a greater degree than ones own oscillators are capable of alone is an interesting concept.
Always interesting to see how one pushes the boundaries of measurement capability.
Also for a second I thought to myself, "well, if one uses two different oscillator brands for the two sections, then why not just intentionally make one section a bit worse spec wise?" But the phase noise spec of each side technically still matters, so it needs to be exceptionally good, otherwise one will need more correlation cycles to reclaim that intentionally lost performance.
But it also makes me wonder, what if we add a third stage? So that we can correlate 3 pairs, instead of just one pair of signals? Logically, it could make the unit 3 times faster, at only a roughly 50% increase in cost. (but this is ballpark estimates.)
Adding a fourth stage would give us a total of 6 pairs, at only an additional 33% cost increase for 2x more performance. But there is likely other reasons that makes this impractical.
Once again a brilliant video. So, are there modulation schemes which take phase noise into account, and put more constellation points in inner radii, where they are more stable?
Great tutorial. Are there standard or commonly used offset frequency's (f[sub]m) that are typical for a given center freq when stating the phase noise of a sig gen? i.e Is it common that a 1 GHz carrier will typically have its phase noise given at say 100kHz. And are the bins always 1Hz?
Thank you for another great video.
If you have a dut and you want to measure its phase noise (e.g., an amplifier), is it enough to connect the amp’s output to the signal analyzer and measure as you explained? Don’t we have to correlate amp’s output to its input?
The reason why your fundamental drops at 27:30 because You are measuring RMS power in 1001 points over 30 Hz RBW, in a span of 1 MHz. 1001 points * 30 Hz is not enough bins to actual see the fundamental signal. If you decrease your span, increase RBW, or increase your number of points, the signal will go back to the actual level.
Thank you for this explanation! It was exactly what I was confused about.
Thank you for this video, everything I've learnt about RF (and RF equipment repair) has been from UA-cam and your videos are 95% of that. I'm well on my way to having my own cheap imitation of your lab, so thank you for the new passion.
You say that the reason for the spurious points on the constellation is due to your 10MHz ref not being common between the instruments. Does this mean you're not doing clock recovery but relying on the refs directly for sync? I think I'm missing something here.
Finally, I think someone else has already said it, but the drop off in phase noise above 10MHz away from the carrier is probably due to filtering in the source to clean up the spurious tones created while generating the 1 GHz signal also attenuating the noise.
Was looking forward to seeing the signal hounds phase noise performance using the cross-correlation technique. I may have missed it.
Justin Richards it’s there. Towards the end.
Nice vid. What would be the best way to measure the pn of a digital differential clock source? How would the characteristic of the balun change the measurement results? Can the 5052 separate the AM and PM noise?
Awesome video. Does anyone know how to choose the integration boundaries (fstart and fstop @ 13:26) in case one needs to know the RMS jitter for a given oscillator with a known log plot of the phase noise?
Yeah, finally I got one right. However, the second question I am guessing relates to the ref freq of 10MHz. And the E5052B indicate a center freq of 10MHz which I think was the center of the scan freq range from fc. But thinking more likely the ref freq was 10MHz
Great video, can't wait for the PLL video! I keep making noisy sources (-80 dbc/hz @ 100K fc=24 GHz) and i think its all my poor LPF designs. I think I've read the literature 8 times at this point, sometimes I feel I'm still shooting in the dark. Bonus question: If the loop contains a N/16 divider, does that mean any phase noise experienced at N/16 circuitry gets multiplied by 16 at Fc?
Log(16)
20Log(16) is the ammount of phase noise degradation due to the mere presence of a frequency divider il the feedback loop..... obviously the degradetion is relevant to the reference. So, the higher is N in the feedback loop, the higher your reference spoil is!
Thanks. Just for fun I would like to see a measuring setup of this scope with all its 8 channels in use. A video covering the subject "Causes of noise"' would be appreciated.
Dr. Shariar, Amazing tutorial. Early waiting for PLL tutorial. Regarding unusual drop (pointed@46:34) on PNCS-1 Phase Noise - It seems due to noise floor extension in the instrument?
Mohan K Thats a good guess. But that is not the reason.
This video is so good. Thanks a lot. What do you think about Microchip 53100A? Is it good enough to characterize the phase noise of 10MHz clock? I am confused between Allan deviation and phase noise, are they related? Maybe one of your video can help me. Thanks.
I am interested in this cross correlation technique. Can you recommend any books or publications on the theory and how correlations can be used to lower noise?
The drop in noise around 10 MHz offset for the PNCS is probably due to the SAW filter that rejects the all the spurious contents from frequency conversion.
sluap Yes!
Excellent overview not an electrical engineer but I can connect the math to the experimental work. Just a note is there any way to do some of these decompositions outside of spectrum analyzer on say software it would be good for people who don’t have access to all these expensive equipment. Btw Catscan…..👍🏾
Hey, truly Great Stuff, perhaps more jingles/sound-effects/cats. The thought occurs - if the E5052E CPU motherboard is just a standard PC microATX running WinXP. Why not swap/upgrade it ? More CPU cores, more grunt, more memory, faster results. Or for an E5052E DIY open-source build, why not use several (4 or 8)cheap SDRs ? As the absolute performance of the uncorrelated LOs are less critical than the mathematical horse-power.
at around @6:50 i.e. in the introduction, you are talking about amplitude and phase noise appearing in the frequency spectrum. What about fluctuations in central frequency f0 itself? What about frequency noise itself which is not originating from phase or amplitude noise? Is it even possible to disentangle these noices?
Since this is the 100th anniversary year of Armstrong's original Superheterodyne patent, maybe a tutorial and historical review of frequency-conversion in receivers might be apropriate?
There is no complex topic in EE that The Signal Path cannot break down into practically digestible elements.
Very good. You mention the time to do a cross-correlation test but don't give examples. I gather it can be minutes to hours for very low phase noise sources. It is a statistical test, so it will need quite a bit of time to detect enough far outliers, I presume.
What is the reason for the sudden decreased in the trace after 10 MHz, at minute 46:30, I don't understand what video is mentioned as a hint, any link please?
Probably TSP # 128 at 11min10s
ua-cam.com/video/vCeSL-ehU4E/v-deo.html
What happened to Pooch ? Great video!
ua-cam.com/video/SOHjFtw0sgo/v-deo.html
🙏🌺
Hey Shahriar, or anyone if I only took CS for undergrad and didn't take calc 3 or differential equations, or physics, should I take some of those on my own before applying to grad schoo for EE or eecs?
@@georgechiporikov2297 Thanks!
@The Signal Path the slope changes around 10 MHz and reaches the thermal noise floor of about -167 dbc/Hz could this because the instruments PLL has a loop bandwidth of 10 MHz and then the phase noise falls to the noise floor of the instrument
Zarf 42 Not quiet the reason for the drop.
how can i get certification of operating a Spectrum analyser??
We need to fund a StarTrek uniform for this guy
How did you get your knowledge, engineering degree? PhD? which book you recommend? thanks! I like your video very much :)
10 years of research work at Bell Labs after my Ph.D. :)
@@Thesignalpath WOW!!
very nice video, hopefully you will speak a bit slowly one day😄
Slides: www.qsl.net/ab4oj/test/docs/20180720_KEE7_PhaseNoise.pdf
Hi i have seen many of your videos, you do briljant repairs. But where do you get the equiptment which is way too expensive. Not that i want an adress. no what i want to say is one would not sell it for very low value but first send it in for repair. and i assume you don t buy project which are very expensive knowing it could end in not working. for example the tektronix analyzer is over $20K. you tell it yourself no schematics just a block diagram. Or are you the representation of a repair company and are you repairing for customers, what i dont think?
nevertheless compliments for you approach and work.
Poor Keysight. They are contributing so much material to the community, while Mr. Shahramian sits in front of an army of Rohde&Schwarz devices and doing the measures with Tektronix instruments. Okay, some Agilents as well.
Besides this: Excellent video with clear explanations and illustrations.
This is not an advertisement channel for any specific vendor. I rotate through all manufactures on a regular basis.
@@Thesignalpath What you clearly demonstrated. Sorry, it was not meant as an accusation.
I have been watching ASP (Arkansas State Patrol) chases and the phase noise on the cops radios is really bad. They need your work with phase noise and jitter to reduce the distortion on the radios. I am sure the police can still understand the two way because they respond just like they heard it clearly. I think humans need training over time to finally understand what the person on the other end is saying. As I listen longer to these police comms I begin to understand better. The home base has a more powerful signal and is easy to understand. The further away another unit is the crappier the signal so I cannot get a read on what they are saying like the cop listening right on the spot. No microphone is as clear as the human ear right on the spot, but the degradation even through my Sennheiser headphones makes them slightly unreadable to me. Maybe if I practice I will get better, it is exciting to me to follow these chases through the night.
E5052B ebay for only $103k