TSP

[@MrCarlsonsLab]

Great video Shahriar. Wishing You and Yours the best this season!

[@Thesignalpath]

Thank you my friend. You too!

[@AissaAzzaz]

No words could express the quality and value of this 1 hour wonder video That experiment setup is just fantastic even though i didn't know too much about RF but man shahrihar have done an awesome job explaining it I wish i could support you bur i don't have any online payment method Anyway this video deserves to be a put in the archive of R&S or in their first page on their website Thanks again shahrihar

[@S1naFa]

یکی از این ها رو تو ایران از نزدیک دیدم ولی اون موقع نمیدونستم چیه و به چه دردی میخوره ، ممنونم که تو این یک سالی که کانالت رو دنبال میکنم کلی چیز جدید بهم یاد دادی

[@BruceNitroxpro]

sina faghihi , برای شما خوب است که درک درستی از کاربردهای ظریف ابزار یاضی مورد استفاده وی داشته باشید. موفق باشید.

[@cakeforcat]

I'm an eee undergrad and watching your videos makes me want to consider a wireless communications specialisation. Keep up the great work!

[@xDevscom_EE]

Xmas video, yay. Would be also cool to compare performance of these modern VNAs vs something old like 8753, as the older equipment is often popular among budget R&D work stuff. Additional thumbs up for 1 hour video!

[@BruceNitroxpro]

I noticed that the power of the diversity and completeness of the instrument impressed you. Is there anything about the LACK of some of this completeness which makes the OLD instruments "useless" in the stricter areas of measurement or use today?

[@connecticutaggie]

Looks like a very nice instrument. The R&S Instrument I have been spending a lot of time on the past couple of years is the CMW100 Radio Test Head. This unit is a great unit for doing production RF testing as it can handle a wide RF range and supports a wide variety or RF protocols.

[@ahedproductions]

As always Shariar, great content, professional review. Your explanations convey all the essentials amazingly well and I have enjoyed each and every one of your great videos. Thank you for your efforts, be healthy and loved in the coming year!

[@steve_case]

Thanks for doing all the experiments. You went above and beyond on this one. Great job!

[@sergeaudenaert]

Quite a lot of info and knowledge in this video - many thanks! Wishing you and the community a joyful and safe reboot in 2021!

[@KF7JO]

The noise figure measurement application shows occasional spikes with very high noise figure when your phone, wifi router, or other nearby wireless device operating in the 2.4 GHz ISM band generate a signal that's stronger than what the noise source is putting out and leaks into the signal path of the DUT. The noise source output power is around 10 dB above the thermal noise floor, while a nearby interfering signal is significantly stronger, even with significant radiated susceptibility isolation. To put that theory to the test, put the amplifier (and possibly cabling, though the leakage here is likely in the unshielded plastic housing of the amplifier) in a faraday cage or turn off nearby emitting devices. The dropouts happen when the calculated noise figure is negative (a physical impossibility, as this requires a thermal noise temperature below 0 K). This can happen for a variety of reasons. The easiest explanation is that during the noise figure calibration step, there was constructive interference between the noise and signal into the ZNL's receiver, while during the measurement step there was destructive interference between the noise and signal, and that the difference between the constructive noise and destructive noise exceeded the noise figure of the DUT. It's easy to visualize this when looking at additive vector drawings. To avoid it, make sure you're operating the DUT and ZNL with a good signal to noise ratio. The noise power out of the DUT may still close to or below the noise floor of the DUT, so removing the 20 dB attenuators after the LNA or an additional pre-amp in front of the ZNL may be needed

[@Thesignalpath]

Excellent. Thank you!

[@bfx8185]

R&S have a really nice instruments. Great review! All the best to the new year!

[@SKARTHIKSELVAN]

Thanks for putting efforts in making these videos.

[@Factory400]

I love this setup. After looking closely at the latest USB options out there, I have decided a full size unit is better for me. The USB VNA's look nice until you realize that if it is used as a portable unit - you have to lug a laptop. Stand alone still wins in my world.

[@daveb5041]

*I also love how you explain the microwave voodoo that other tech channels just have no idea what it is*

[@BruceNitroxpro]

Dave B , And much of the "voodoo" is EXACTLY that. IF it is above the scope frequency for me, much of this doesn't exist.

[@IxIVVI]

Incredible quality!!!

[@Oliver-by1ko]

Great video, i have been wanting to know what the inside of my ZNA VNA looks like, now i have some idea.

[@TomStorey96]

When your lab is so big that you have different areas for each vendor. 😄

[@wadehsu2347]

Happy holidays Shahriar. Very nice video, I really like what R&S is doing with their UI for these instruments. As manufacturer pack more and more functions into one box, having a good UI become more and more important. I also like what the instrument makers are doing in general these days, making the box shallower so that the work bench doesn't need to be as deep to accommodate them. It would be interesting to do a comparison to Keysight's new offering of the streamline series VNA that also has a spectrum analyzer option build in. Also, It looks like R&S are using PXIe style connectors inside their instruments, do those run PCIe or they are just using the connectors?

[@KF7JO]

Instruments are becoming more powerful and usable today thanks to the software. The entry-level ZNLE ($20k) has the same interface as the flagship ZNA ($200k), and everything in between. Same SCPI remote automation commands, too. I believe it's all PCIe.

[@rfmonkey4942]

love the toys dude, have few my self but clearly i need to en-devour to acquire more love your videos dude !

[@gn_ghost4757]

Great video! The setup really shows how powerful this instrument is... I am just wandering would you still consider doing the PLL video? Have been waiting for it since the phase noise one. I have only learned the basic PLL operation and construction from my circuit's lab course, curious on how the spec is determined. Have a nice holiday!

[@Thesignalpath]

Thanks! Yes, PLL is on my list.

[@derekkozel]

At 10:55, could that line be the bias voltage for the LO amplifiers? A lot of passives, but possibly stabilization and decoupling?

[@NomenNescio99]

How is the starlink dish analysis coming along? Any ETA on when you will release the video?

[@Thesignalpath]

Right after the break. :)

[@PeterHuson]

@@Thesignalpath Super excited for this!!! Just so you're aware, @MikeOnSpace has been poking at Dishy's serial console, and it's been 2 weeks since he posted his last update. Maybe he's gotten into Dishy's operating system by now, it would be worth seeing what some poking around in there might uncover. kzbin.info/door/-M_5d0nlD2mO0qhN_J9XIg

[@KevinRegan875]

Shariar, do you have any high res photos of the RF pcb? Also can you recommend some good books on vector network anaylzer design? Very hard to find detailed information about VNA's. Thanks!

[@KF7JO]

www.rohde-schwarz.com/us/products/test-and-measurement/network-analyzers/fundamentals-of-vector-network-analysis_253352.html us.artechhouse.com/mobile/The-VNA-Applications-Handbook-P2050.aspx

[@christophscholz]

The VNA textbook is by Dr. Dunsmore (now in the 2nd edition) www.amazon.com/gp/aw/d/1119477131/ref=dbs_a_w_dp_1119477131# Two applications engineers from R&S wrote a book on VNA applications: www.amazon.com/VNA-Applications-Handbook-Gregory-Bonaguide/dp/1630816000#

[@Abt2495]

Merry Christmas Shahriar! You got me into repairing equipment from eBay and your channel will forever be one of my favorites!! any tips for an electrical engineering college student who wants to work on RF systems? Thanks!

[@KF7JO]

Join or start your college's or town's amateur radio club. I can't say enough about how many practical hands-on applications there are in amateur radio which pertain to the design of RF systems. Passive Intermod, Intermod distortion, linearity, power efficiency, gain, noise figure, sensitivity, emissions and susceptibility, filtering, mixing, harmonics, power supply noise, isolation, circulators, duplexers and diplexers, modulation, over-/under-deviating... The list goes on and on. You'll learn a lot about this stuff in amateur radio and have a much more concrete understanding of these concepts than what you'll learn from your BSEE, assuming the circiculum even teaches all these RF concepts (many EE programs focus on the DC and AC, and never really dive deep into the RF and microwave design).

[@tommihommi1]

Maybe a naive question, why would you not include the attenuattors in the calibration? Something to do with the dynamic range?

[@KF7JO]

Good question! Probably to keep the video simple and save on time so he'd have time for intermodulation, analog modulation, and noise figure measurements. 20 dB attenuation is probably too much to include in the calibration. The calibration on port 2 would measure an S22 return loss of -40 dB without connecting the open, short, or match standard. A calibration of -30 to -40 dB is considered good, which means that the open, short, and match would be nearly indistinguishable from each other. Not a good calibration! The better approach is to calibrate without the attenuators, then measure the attenuators and de-embed the result in the Offset Embed menu (load in the s2p file for the measured attenuators). If these were each 3 dB attenuators, then including the attenuators in the calibration would have made sense. Much more than 6 dB to 10 dB attenuation (12 dB to 20 dB return loss) and the de-embed approach is more favorable. Since Shahriar was only measuring scalar gain of the amplifier and the attenuation was a conveniently round 20 dB, doing the mental math was totally fine and engineers will do this on a daily basis in order to save time when the additional accuracy from measuring and deembedding the attenuators isn't strictly necessary. In fact, many engineers who are new to VNAs (probably the largest audience of this video), it's easier to understand what the VNA is doing by keeping the calibration and measurement as simple as possible. The more powee you delegate to the instrument, the more of a black box it is, and then this video isn't as effective at teaching viewers who aren't VNA experts. Of course anyone who uses a VNA on a regular basis will know about deembedding and will use this feature when it's convenient for them.

[@bvernoux]

Amazing teardown and review !

[@bvernoux]

Price 24.5KUSD for the ZNL6-06 on newmark: www.newark.com/rohde-schwarz/znl6-06/vector-n-w-analyzer-2-port-5khz/dp/43AC9678 I dream one day we could buy those instruments for less than 5KUSD ... (reaching 7.5GHz for sub-6 GHz / Wi-Fi-6 and 5G)

[@cubipak]

hi, after watching the video. on the rf board, i can't find a rf connector to input the reference signal for the DDS, does the reference signal is input from the DC connector?

[@AF6LJSue]

Thanks Great review, nice piece of gear.

[@lukasblenk3684]

Wow maybe some day this thing is on my bench but until then the nanoVNA 2 has to do its job ^^.

[@PlasmaHH]

hm, and I would have thought the sweep button in the power submenu would have done the power sweep when in cw mode...

[@R2AUK]

58:00 any chance the glitch comes from local Wi-Fi / Bluetooth devices?

[@KF7JO]

Bingo! The amplifier is unshielded on a PCB, so of course the input PCB trace would act as an antenna and pick up any nearby 2.4 GHz ISM band signals for amplification. The signal from the noise source is extremely weak, so it doesn't take much radiated power from a nearby WiFi or Bluetooth device to show up as a glitch on the noise figure measurement.

[@b4096-l5o]

Do they use 2 layer boards? Also at these frequencies, what do they use as the PCB material?

[@jorditribo94]

49:50 why is it called spectrogram if the horizontal axis is time? 52:00 are you supposed to get a sine wave in a logarithmic baseband power graph of an AM modulated signal?

[@KF7JO]

49:50: Good observation. You're correct that this isn't your usual *frequency* spectogram, which shows magnitude (z axis or color) versus frequency (x axis) and time (y axis). This spectogram is a time waterfall of the magnitude versus time (essentially a zero span measurement with history) trace above. Keep in mind that a spectogram can show us any properties of a signal over time--frequency doesnt need to be one of the axes. Besides the usual frequency spectogram, there are also phase spectograms and time spectograms (here), Perhaps the colloquial and more general term "waterfall" would work here without using the root "spect" which immediately leads us to think of power versus frequency. This spectogram shows power versus time-in-acquisition and acquisition. Essentially 2 time axes, with the IQ Power trigger aligning the time in acquisition between each acquisition.

[@KF7JO]

52:00 The RF Time Domain and AM Time Domain are slightly different. You'll see a pure sine wave on an AM Time Domain display, while the RF Time Domain will look pinched on either the low side of high side depending on whether you use linear or logarithmic scaling. So the best way to view this is with a true AM Time Domain (or AM Spectrum for AF harmonics of the low frequency modulating signal and RF Spectrum for RF harmonics of the high frequency AM-modulated carrier)

[@R2AUK]

36:00 I don't think you should account for the attenuator when measuring TOI. Doesn't TOI depend only on ratio of desired signal and IMD products? Also TOI is not a function of applied power. Maybe the reason why you see a change of TOI from 14 dBm to 18 dBm is that TOI is applicable to class A amplifiers, which DUT most likely is not?

[@Thesignalpath]

Attenuation is indeed added to the TOI. Two amplifiers with the same IM3 terms but different output powers do not have the same TOI. TOI is almost always a function of input power at some point especially when the amplifier changes class of operation (or is pushed into a different class due to input power). This amplifier is also not class A.

[@R2AUK]

@@Thesignalpath I must admit, I'm not an professional RF engineer, only an amateur radio enthusiast. The reason why I'm so surprised is that what you say strongly contradicts literally all books published by ARRL, "Experimental Methods in RF Design" for instance (or the definition given by Wikipedia if that matters). My best guess is that there might be some extended understanding of TOI for amplifiers that don't work in class A, which I'm currently unfamiliar with. Maybe you could recommend a book or two whith an alternative view on that matter?

[@KF7JO]

@@R2AUK When you add a 20 dB attenuator onto the output of an amplifier, it reduces the power of the carriers by 20 dB and the intermods by 20 dB (not 60 dB!). If the attenuated spectrum was used to calculate the TOI, the IMs would be 40 dB higher than they really were, resulting in a TOI that's 40 dB lower than the true value. Anything that doesn't keep the carrier to Intermod ratio at 3:1 must be corrected for. The correct way to calculate the TOI is to de-embed the 20 dB attenuator so that the spectrum analyzer shows the unattenuated output power, then compute the 3:1 slope to find the TOI of the unattenuated amplifier. This is all consistent with the Intermod theory taught in ARRL handbooks.

[@R2AUK]

@@KF7JO you are right, sorry for the confusion. TOI is calculated as IM_db_ratio / 2 + PowerOutput. The IM_db_ratio is not affected by the attenuator but I forgot about the last part :( Still the fact that TOI changes with the change of output power is not right.

[@jjoonathan7178]

Why do the heart of R&S instruments always have rusty crusty heat sink /shields? They usually have other machined components without the trademark crust and it's been this way for decades, so there's clearly *some* method to the madness... but what could it be?

[@JWH3]

They only thing crusty here is the copper, there are no artificial copper passivation processes so they have no choice but to let it oxidize naturally. They could probably gold or silver plate it but the area there is huge and wouldn't be cheap and it's not really necessary. Given the cost of these devices though it would certainly be a nice luxury choice

[@jjoonathan7178]

@@JWH3 @Jake Heuft I'm not convinced. If it's copper, it's doing a hell of a job cosplaying as one of copper's silvery cousins. R&S is no stranger to custom machined brass and gold-plated BeCu, which really do look quite different, so I'm pretty sure this is something else, something intentionally different. Here are some closeups from my RTO. I've seen similar construction in 3 or 4 R&S devices now.

[@jjoonathan7178]

Ah, right, my bad, I can't include a URL without youtube silently dropping my comment. Anyway, it shows a very silvery closeup with a number of nicks and dents that are also silvery. Still, even if it was copper, it wouldn't explain why some of their Au/BeCu or CuZn modules have perfect golden surface finishes while others look like crusty rust over a silvery metal.

[@RandomUser2401]

@@jjoonathan7178 I am also very curious. Maybe Shahriar can comment on that.

[@KX36]

complete guess, but maybe it's used in a similar way to "weathering steel" where there's an intentional thin layer of corrosion which protects it from deeper corrosion (not saying that it's steel).

[@erezlevi5093]

Resistive couplers??? What kind of VNA has resistive couplers on its inputs???

[@Thesignalpath]

Many kinds!

[@erezlevi5093]

@@Thesignalpath Does resistive coupler has any directivity?

[@sanmvegs1641]

mind=blown

[@Nik930714]

7:31 - I love how small the main PC is. Any idea if its ARM based or x86?

[@SalamanderDancer]

x86-64 running Windows 10 Embedded

[@testboga5991]

Beauty

[@jusbmx5515]

Is that a XLR connection for power?

[@KF7JO]

4:30 Good eye! Possibly. The ZNL datasheet unfortunately does not specify the connector type for the FPL1-B30 12/24V DC Power option. scdn.rohde-schwarz.com/ur/pws/dl_downloads/dl_common_library/dl_brochures_and_datasheets/pdf_1/ZNL_dat-sw_en_3607-1071-22_v0500.pdf page 17

[@johnwest7993]

I can only drool. But at least I have a NanoVNA and a TinySA Ultra, so I'm doing OK.

[@daveb5041]

*Love the channel but I always wonder how do you afford all this crazy gear like a thousand dollar microwave splitter/adapter and 1000$ probes* ? If we added up the total $ of all the equipment you have featured on your channel it would be well into the six figures if not more. Is your full time job being a dictator/king of an oil rich country?

[@aqib2000]

That’s alot of money on that bench

[@TheDefpom]

One of those would look great on my bench... pity it’s way out of my budget.

[@yaghiyahbrenner8902]

DVI not HDMI ?

[@Daluk0]

Very common to avoid the licencing involved with a HDMI port

[@yaghiyahbrenner8902]

@@Daluk0 ah yes that makes sense.

[@timun4493]

It is the DVI-I variant, the cross on the right side carries analog RGB signals so it can be connected to a VGA display using a passive adapter and the digital TMDS video signals are the same as with HDMI so this really is the most versatile connector they could have chosen

[@yaghiyahbrenner8902]

@@timun4493 thank you for explaining that one.

[@NiHaoMike64]

@@timun4493 Displayport is the most flexible of all - can easily be adapted to VGA or HDMI.

[@AaronHe-z6s]

the content of teardown is too short.