TSP

In this episode Shahriar presents a deep dive into mm-Wave frequency divider; in particular static frequency dividers. These blocks are ubiquities in nearly every modern integrated circuits. The principle operation, circuit topologies and limitations are presented with several publications as examples.
The 30GHz static divider is then characterized extensively in both time domain and frequency domain. The measurements includes self-oscillation frequency, output waveform shape, input/output phase-noise relationship and the impact of temperature on self-oscillation frequency.
A detailed module teardown is also presented with focus on packaging techniques as well as closeup of the die. The integrated circuit is analyzed and the relationship between the schematic and layout is established.
www.TheSignalPath.com
/ thesignalpath​​​
www.Patreon.com/TheSignalPath

[@keithlambell1970]

I really enjoyed the mix of theory, experiments and detailed tear down of the unit. It is fascinating to see how things work at such a small scale.

[@IxIVVI]

I think this format is really working very well. Speed and depth are excellent to be widely accessible, but at the same time very informative. Amazing material and really inspiring! Thank you very much!

[@maximus6884]

I truly enjoy your wireless theory videos. They, somehow, are conducted in a unique fashion. I'm not saying the other videos are not valuable but I do not watch repair videos. I just can't get enough of videos of this type. You manage to deliver theory yet keep us entertained throughout! please do more tutorials.

[@cosmolittle1395]

In spite of my approx 55 years experience in electronic design, I can still learn something new. Thank you Shariah, I regard your videos as an important technical resource.

[@tfinmoraes]

Thank you, Shahriar, this kind of deep dive is fantastic! Truly informative and enjoyable to watch!

[@Thesignalpath]

Glad you enjoyed it!

[@gordini904cc]

Absolute brilliant video. You cover the basic divider and include high level measurement technique. Then add my favorite subject, phase noise. Going to recommend this video!

[@Bertemus60]

catching up finally...another great microscopy episode....microwave design never ceases to amaze & I greatly admire your talent to break it down for tinier minds like mine . Keep it up sir!

[@mohamedlanjri]

Very nice expanation preceded with theory. Adding background knowledge before studying systems is a plus. Congrats!

[@Gigabecquerel]

This is a great video, and I really enjoy the long form content. Some of the shorter videos leave questions, but this one was great!

[@timlister177]

Really enjoyed this deeper dive into one of the fundamental "building blocks" of RF circuits (and seeing Pooch also...). Hope you can continue these (RF mixers next?) even though I suspect they are poison for the KZbin algorithm

[@Carlitto314]

Really liked the in-depth and yet broad information, thanks Shahriar!

[@Thesignalpath]

My pleasure!

[@stefanvoigt384]

Thank you so much, this is a perfect mix! I really like the hints you provide throughout the video, regarding the limits of the instuments

[@vmiguel1988]

I loved the deep dive! Keep doing it!

[@hpux735]

The format was great! I feel like it's back closer to the older Signal Path videos with a lot more theory and practice than equipment reviews.

[@fredflickinger643]

Good balance of theory, application and testing! Definitely enough to inspire self study in this area!

[@TheBananaPlug]

Excellent video, I enjoyed the mix of theory and practical techniques. Thanks.

[@Darieee]

amazing work ❤️

[@richnormand1549]

Very nice work. Loved it.

[@tommihommi1]

great video, I had no clue this is how frequency dividers are done.

[@glasslinger]

Amazing! I see experiments at 100 times the frequencies what I can do in my home lab! With this being freely available to anyone that has CASH I wonder what is in the top secret black budget labs that are rumored to be 20 years advanced from what we see now! (swoon!)

[@erezlevi5093]

Excellent video. Well done and well explained.

[@xDevscom_EE]

Nice tutorial, learned something new. Why it needs K connectors at the divided output, just to maintain wide bandwidth of signal? (Edit: Yep, as demonstrated with sampling scope, thanks!) Also I been in NTU many times in my Taiwan years. They have some cool EE labs there.

[@AF6LJSue]

Great Video Thanks.

[@wolpumba4099]

Amazing!

[@LutzSchafer]

Thank you Shariah! This was very nicely explained. It made me remember my electronic developer times at the beginning of the 1980's. Back then in east Germany the fastest logic one could get was Soviet made ECL (emitter coupled logic). That was good for up to about 300 MHz. I used it in a similar way as you described for division until TTL could be used. Its amazing how technology has progessed over time. Thanks again. BTW have you made any progress regarding your problem with the PSF of that one lens on your Axioskop?

[@analog9439]

I am still waiting for that MXO 4 review

[@dtiydr]

37:52 Fine tuning perhaps.

[@grhinson]

He actually did a video on this subject about 5g

[@WolfmanDude]

That circuit is way simpler than I expected! Do you think I could make something like this from discrete NPN transistors? I need something like this for like 800MHz

[@YuriyKrivosheyev]

Very good, thx!

[@Thesignalpath]

Glad you like it!

[@wilfredswinkels]

the good old 7805 has an even lower noise figure than the lm317

[@AndrewZonenberg]

The text "1_2_S DIV" on the die is mirrored. Is your camera/microscope flipping this, or is the text actually backwards on the die? It looks like it's using gold wedge or ribbon bonding, rather than the more conventional ball bonding. I guess this makes sense from the perspective of minimizing parasitics. These days I'm used to low-cost stuff using ball bonding and high performance stuff using flip chip solder bumping, although I'm not sure what the state of the art these days is for mm-wave packaging. I've personally worked on projects using NRZ SERDES out to 32 Gbaud with flip chip bumping and I'm pretty sure you can do that OK at 56/112G as well. Maybe you could do a video at some point just on packaging and die fanout for high frequency stuff? Also interesting to see straight, any-angle metal 2 interconnect in the first stage latches. This is the exact opposite of what I'm used to seeing in modern deep submicron silicon CMOS tech. I assume this die is all hand laid out, but probably still with modern CAD technology rather than something like rubylith since it's not THAT old. Is this just done to reduce wire length to the absolute minimum to reduce parasitics? If so, why not go all the way to curved lines - software limitations? EDIT: If this thing does fail completely, and you're able to remove the die/substrate from the big metal package without damaging it, I may be able to try repairing it. We have a dual beam SEM/FIB at work with gas injectors for (among other things) platinum metallization, so reconnecting a damaged wire on top metal - especially on such a large process node - is very doable for us. Would have to get approval from higher-ups but I think odds are good they'd see the PR value of helping you out as being worth the instrument time.

[@Thesignalpath]

Video was mirrored. :)

[@janedoe6182]

Why is it necessary to divide by 2 such high frequencies? What is the real use of these dividers?

[@Thesignalpath]

I showed a few applications in the video.

[@Qyonek]

Pretty please hang that Enterprise picture to be visible again

[@R2AUK]

🔥👍

[@donbradley9150]

The IQ divider should be a Div/4 ring counter to get 0,90,180,270. The Div/2 only has 0,180

[@Thesignalpath]

No, the quadrature signals only require a DIV/2. The signals come from the output of the two latches. You don’t need a DIV/4.

[@donbradley9150]

@@Thesignalpath can you provide a simple circuit ? I could consider using both edges of the input clock, but it would need a symmetrical square wave input.

[@Chriva]

That thumbnail lol. Did you just go Homer Simpson mm wave dividers? :D

[@grhinson]

What's that mean? In lost...

[@ashave9100]

Have a bit of respect- If you listen to the "Good Doctor" you Will learn something !!!

[@trevorcrowley5748]

Interesting. "Mmm, donuts" , where mm is the input wavelength and the donuts represents the ring oscillator?

[@emwave100]

Do you know anything about any health effects to high power high frequency electromagnetic radiation? This is not a joke, I am experiencing several unexplainable symptoms including my ears ringing similar to extreme tinnitus, and temporary hearing loss that can last up to 30 seconds. What I have experienced can be so strong sometimes to make my skin feel like it is burning. I suspect that someone in my neighborhood is aiming a high power transmitter towards my house. I have no way to detect such high frequency emissions, and I have made a report to the local police department. If someone is doing this to me, this is assault, wouldn't you agree? I don't know who to call to even begin to detect something like this. Have you ever heard of any symptoms similar to what I have experienced from being around high power RF radiation? Thanks for any help.

[@bakagaijin7452]

Any radiation is susceptible to an inverse square root law. That necessitate a megawatt class transmitter to even feel something. At that level all the electronics around will be long long fried dead. Let alone cooling, power consumption and a sheer size problems. So thinking that your problems is somehow related to a rogue rf transmitter in your neighborhood sounds like paranoid delusion to me.

[@Zetex2000]

You can't just aim such high power directly to your house. In addition, it'll take a LOT of power, nothing an ordinary house can generate. I think you have something else.

[@paolomonai9511]

First check the basics. What about your blood pressure? The highest microwave power available almost to every one is the magnetron in a microwave oven. To concentrate such power (say about 1KW order of magnitude) you need an antenna. But not all the power can reach a target due to antenna efficiency and propagation losses...so targeting intentionally some one is not so simple. Regards from Italy.

[@pasikavecpruhovany7777]

You could buy a spectrum analyzer second-hand but it's extremely unlikely that someone is "microwaving" you. At higher frequencies (say above 5GHz) the transmit power would have to be really high as they are attenuated by walls. You should consult with your doctor and get a check-up. If it's only happening to you at home, I'd suspect some toxicity (heavy metal, co2) or allergy but I'm not a doctor.

[@martinmckee5333]

In the past years I have started suffering from heat triggered hives. I don't have the extreme tinnitus effect that you mention but I do, sometimes, suddenly feel as though my skin is burning off my body. It is, as I understand it, an auto-immune response where the body overreacts to an increase in body temperature. This is only to say that there are plenty of strange things that biology can do that has little - or nothing - to do with external causes.