TSP #26 - Tutorial on Microwave and mm-Wave Components and Modules

[@Thesignalpath]

After a decade, there is a mistake in the video I'd like to address: The subtle description of the doubler operation at 45 minute mark is not correct. The top and bottom of the board form a broadband balun and the tapered lower layer transforms the signal into a balanced mode into a quad pair of diodes. The signal then mixes with itself producing the doubled frequency. At the same time we don't want to self-mix down to DC and create a DC offset so the little wire is there to keep the DC voltage at 0V while acting almost like an open at higher bands - kind of like a bias-T but with a lower cut-off at >1GHz. This doubler is fairly broadband, operating at the input from 1GHz to 32.5GHz. The fundamental rejection is about 15dBc worst case, likely at the edges of the band where the balun does not have very good phase and amplitude response.

[@RandyLott]

You're not only incredibly knowledgeable and intelligent, you're a phenomenal teacher. Those qualities typically lack synergism. Not only that, you are willing to take valuable time away from your life to enlighten an effectively anonymous audience for no significant monetary benefit. I'm an electrical engineer. I'm still young in my career (I'm 27), but my hunger for discovery has been present my whole life. I just wanted to thank you, on behalf of all of us.

[@T3Beatz]

Couldn't have said it better myself, coming from a 1st year electrical engineering student (telecommunications).

[@Thesignalpath]

Thank you for your kind words.

[@stevefriedl3983]

This was super useful, but I'm sure many of us are wondering how much this stuff costs, so I made notes throughout and looked up prices for everything I could identify. If no manufacturer specified, it's HP/Agilent/Keysight, and prices are either new, or taken from eBay in Jan 2021, all in US dollars. I'm not an expert on any of this, I just looked stuff up. @14:20 - 8710-1765 Torque Wrench, 8 lb-in, $300 (eBay) @17:29 - N5520B, 1.85mm - 1.85mm F-F coupler, DC-67GHz, instrument grade, $1649 @20:14 - 85138A terminator, DC-50GHz, metrology grade, $1036 @21:28 - 33340C fixed 20dB attenuator, DC-26.5 GHz. (obsolete) 8493C replacement part for 33340C, $506 @22:29 - 11636B power divider, DC-26.5 GHz, $1725 @24:46 - 11667C power splitter, DC-50GHz, $2572 @27:16 - 87304C Hybrid Power Divider, 2-26.5Ghz $1144 @28:10 - MACOM 180deg stripline hybrid, $2483 @28:57 - 5086-7518 50 GHz directional coupler, $1400 (eBay) @29:23 - MACOM 3020-4099-00 directional coupler, $75 (eBay) @31:20 - 5086-7489 50 ohm R-Channel coupler, $575 (eBay) @32:17 - Picosecond Pulse Labs 5500A DC Block, $199 (eBay) @33:05 - Picosecond Pulse Labs 5575A 12GHz bias tee, $250 (eBay) @37:32 - SHF 123B bias tee @39:54 - 11612B Bias Network, 45MHz - 50GHz, $2033 @40:32 - MACOM PS20-0001-K DC-30GHz phase shifter @41:51 - Maury Microwave 1819C trible-stub tuner 2-18GHz @44:20 - Marki Microwave D0265IJVV passive doubler @46:33 - Anaren 74125 1-4GHz double balanced mixer, $60 (eBay) @48:23 - Mini-Circuits ZFJ-1000VH amplifier, $100 (eBay) @49:40 - Anritsu A7HB3102 wideband amplifier, $400 (eBay) @50:40 - W281C waveguide adapter, 75-110GHz, $4144 @51:45 - Hitachi WR28(?) directional coupler waveguide, $150 (eBay) @52:26 - millitech 45168H-1000 circulator, CF 140 GHz @53:23 - WR10 horn antenna, $775 (eBay) @54:26 - Cascade 167-A-GSG-100 probe @57:40 - GGB PicoProbe MCW-23-137205

[@Afrotechmods]

Next week on the Signal Path Blog: How to build a communications satellite using household materials.

[@AissaAzzaz]

I always wished this this to happen

[@Thesignalpath]

100GHz circuits are rather common these days. Radars, imaging circuits, ultra high-speed data converters and data communication systems, short reach and backhaul radios...

[@DDas-yi6jr]

Nothing but respect and love from India sir

[@artrock8175]

I've used your video library as a reference for years now! Thank you, Shahriar! 🙏

[@WondersOfTheUniverse]

This is a very good review of connectors, couplers, amplifiers, waveguides, and other microwave components! It's also very interesting getting to see inside some of these components......something you don't get to do very often in the lab..... It shows that many of these devices, such as splitters, amplifiers, and multipliers, can be designed in-house if you're on a budget......

[@reidbyron7662]

26:00 Did you ever happen to put together that tutorial on microwave design? You would do a fantastic job and I hope to see that video in the future

[@haxensalat]

any news?

[@BruceNitroxpro]

Since this was made, the state of the art has changed, also. Just not in the areas he dealt with!

[@Afrotechmods]

Fantastic topic, fantastic video. Another home run.

[@Scuba_Bro]

Great video! As someone who works in microwave and mm wave R&D I have to say you do a great job of explaining these things in layman's terms. Good job!

[@BoutinMathieu]

Very nice presentation, and very interesting. I am not familiar with this topic but you explained it the right way. Thanks a lot and please continue making videos because you know how to make them very useful.

[@JLK89]

Fantastic video, thanks. It's nice to have someone go through the real world aspects of high frequency systems like connectors and components. As an electronic engineer I have a huge amount of respect for microwave engineers. Nevertheless it's happened to me a few times that I go to a lab to test something and look like a moron because I use the wrong connector terminology or idly steak my sticky fingers over their $1000 precision load...

[@Thesignalpath]

As long as you are careful, patient and familiar with the components, everything will be Ok!

[@pooyamz]

Thank you so much Shahriar. This was one of kind video which you can not find in other places.

[@JABloch]

This was a very good overview of the common microwave lab components. This would have been really useful to see when I was starting my grad research.

[@FurkanBahadr]

This video is great! I had already decided that I will contiunue my study in EE in Microwave Engineering, but of course I hadn't known that much about these things Ive never touched one except my laboratory courses because they are very expensive ! But watching you demonstrate is truely fun. Thanks, now I will go through all of your microwave related videos :D

[@-vermin-]

Physics experiments, THz imaging (in airports) and spectroscopy are a few that spring to mind.

[@deebsboy]

I have worked with waveguides in the Navy, but never made measurements. Make sense seeing as how I would have been sent to mast if I broke one of those meters. Awesome video.

[@PowderMill]

👍🏻🇺🇸🇦🇲👍🏻. THANK YOU Shahriar ! Yet ANOTHER OUTSTANDING VIDEO !! I honestly think that I’ve obtained more practical information from this KZbin channel than EE schooling and 30+ years in the field designing, installing and servicing public-safety communications systems. (and tuition was wayyyy cheaper compared to the new “WOKE” Institutions of Debatable 🤪 Learning rates are) Between TSP and IMSAI Guy (& a few others), KZbin is a phenomenal resource that I really could have benefited from in years past. Keep up the great work. &. 73 from an NJ Armo 😷😀

[@Thesignalpath]

The geometry is almost identical. But the coupling will depends on the launch port (i.e. the input signal propagation direction).

[@Thesignalpath]

These connectors are precision components. But the main reason they are expensive is volume. An Intel processor is more expensive to produce in small quantities, but you can buy them at a lower price than these connectors because they manufacture millions of them. The price of >50GHz components have already dropped significantly in the past decade.

[@scottvance74]

At 32:00 the coupler looks very much like the one in HP VNAs such as the 8753 d/e/es series. The resistors on the right hand side have a tendency to burn out due to ESD as they're built onto the trace/board. They can be fixed (basically) by replacing them with a 265 ohm resistor to make up for the burnt out component. Extremely common problem with HP VNAs. The screw positions appear to align and the internal structure is extremely similar, so if this isn't an HP part, it's a knock-off.

[@yasirshafiullah3016]

Excellent Video, Really appreciate your efforts showing these expensive components. Keep it up 👍. I learned alot.. 😊

[@G0HZU]

One minor point... The two step Wilkinson splitter shown at 26:05 probably only works over 700MHz-1500MHz and not 100MHz to 1500MHz. I don't think you can get 15:1 bandwidth from a 2 stage splitter like this.

[@mazenl7139]

Incredibly informative and definitely learnt a great deal from watching this video. You are a wealth of knowledge and greatly appreciate you taking the time to share this invaluable insight. Should've been my professor when I did my EE :). Frankly, addictive watching your videos. Keep up the great work, and will nominate you for Noble prize, cheers!

[@gamccoy]

Another stellar video. I learned a lot. My experience and education is pretty much limited to freq below 1GHz. Higher RF is a bit of a mystery to me. I wish more people would discover your videos.

[@tehcovmeme]

The more varied tutorials there are the better, they will compliment each other. Some one might cover some points while another tutorial will cover another aspect. That's why there are hundreds of books covering each topic, otherwise we would have a single book for each topic. :-)

[@Soundfactory24]

cool RF-Stuff and very good Explanation ! the Radio Amateur,s heart is bumping

[@oriole8789]

This was very interesting to watch. I didn't really realize that simple adapters and connectors could make this much of a difference in terms of attenuation and frequency response, but I have personally seen those differences now. Some cables and connector types simply block the signal entirely, whereas others introduce lots of spikes etc. I will prove this in videos very soon (not on this account), unless Shahriar brings a 65GHz VNA home to make another killer video haha. ;) ;)

[@fpliuzzi]

Now that was certainly an interesting hour of video. Thank you.

[@Thesignalpath]

Can you please clarify what you mean by "MOSFET" RLC? Are you talking about a tutorial on MOS based VCOs that use LC tanks?

[@gerrysweeney]

Hi Shahriar, great introduction into microwave components. Way beyond my understanding but nice to have an explanation I can understand. Gerry

[@user-uw1wq9rj8g]

Thanks for the explanation, that's really informative

[@davekendall9749]

Truly amazing and there was me searching around for a 1GHz Dc block. Thanks for taking your time to make interesting videos. I can only guess why satellites cost so much.

[@Alexander_Sannikov]

26:10 >tutorial on microwave design wait wait wait, where's that? that's exactly what i was looking for.

[@PyroShim]

Thanks for making videos like this. They are the best.

[@tiredfingers99]

You want your coupler to be directional because these are mostly used in networks and you want to see where in your network the signal is getting through and not getting through and which direction the signal is coming from, so there fore you can trace where in your network there may be a problem. The signal coming in through the out put port is usually terminated through the coupling to a terminator, unless of course your using a dual directional coupler.

[@tuberlook1]

a couple of Gigs is almost DC these days. Love the vids, looking forward to see where you are going with RF topic.

[@Plutonion2]

that was really interesting looking in those microwave modules.Thank You for your videos .:)

[@valerianpantsulia1730]

It's cool... When you know...you know...Thanks for honest information & Great video.

[@MuhmmadShahAlam-i9k]

Wondeful Lecture!

[@moshet842]

Do you account for the capacitance between the tips of the probes you showed at the end? I would imagine it would be significant at high frequencies.

[@waverleyrocker]

Small thing - you should only tighten with the torque wrench until its starts to move. If it 'snaps' over you may well put extra torque on the connector.

[@md.sazzadhossain3559]

Really Informative!! Thank You. Hope more other new device clarification

[@bkupiec]

Amazing microwave show and tell. Thanks!

[@ChenxiangHsu]

you are my hero. bro. let me take a seat and watch the tutorials.

[@ooibenzm]

Hi,thank you for the video. Could you explain more on the tapered grounding of the frequency doubler? Thank you.

[@Thesignalpath]

I'd love to do a full VNA measurement tutorial, but I would need a VNA for that. :) If someone wants to donate a portable VNA to The Signal Path Blog, let me know!

[@douro20]

You can buy surplus waveguide components quite cheaply (about $150-400) but it's when you have to couple back to coax where it gets really expensive...

[@joes5669]

Inspection under a microscope is critical. Some metrology labs have bizarre criteria for acceptance of these connectors. For example, I've had calibration kits returned as "okay" when upon inspection, most of the spring fingers inside the connector were missing! It turns out, the only criterion for acceptance was pin depth!

[@deefdeefdeef]

I have to ask. What was the 325GHz waveguide for? What have you used it for in the past? I can't even fathom an application that requires a frequency that high. Wow.

[@kd1s]

SMA in amateur radio handhelds too. Of course for good stuff at 70cm and above you use N connectors.

[@EEEAmin]

I did my BSc in EEE and ya i have microwave in my theory and lab....didnt learn this stuff there.learned abt waveguide and radiation pattern of antenna mostly. what do you think teachers should teach in undergrad level microwave lab which is very helpful in the industry?

[@sic-]

Absolutely superb video, I learnt huge amount. I find RF fascinating nowadays and I wish I could remember more from the comms lectures at university.... Off-topic and random, has any said you sound a bit like the character Dr Nick off the Simpsons? :)

[@rrb6544]

I love this tiny +300 GHz horn hahaha lovely!

[@DoRC]

do you have a list of your tutorials in order they should be watched?

[@nickgiannotti3775]

Question: When using right angle connectors, will there be reflections and return losses? Do they have same performance as inline connectors?

[@manse990]

excellent educational video.can you give more detailed analysis of phase noise when you can

[@Thesignalpath]

Yes, I did! Those connectors have already been misused and are quiet old.

[@nickpelov]

if it's so critical to connect the right type of connector why there are no clear markings on them? if it's not possible to make a different shape I'm sure there could be some text or markings without changing the characteristics of the connector

[@charleyfan1908]

Because it is good business to sell you another $500 cable, and $2000 to service your instrument.

[@Thesignalpath]

That is actually what that is... I'll look into fixing it.

[@KeenanTims]

Very nice to see some of this stuff, I've *heard* of most of the components on display, but never seen them, let alone with their clothes off ;) What causes the directional coupler to be directional, and why do the forward and reverse ports select power going that direction, the geometery seems nearly identical?

[@radoinc]

Did you put together a metrology grade connector with an instrument grade cable in the first half of the video? I thought that was bad.

[@assadij]

Shahriyar Jaan, You know what would be good is step by step VNA measurement of S11/S21 with calibration OSL. I know how to do it, but I haven't seen any videos on youtube. So it would be the first and very popular. Also how to measure balanced components or antennas with a 2 port VNA, which is a tricky process. Love your work. Mersi ;)

[@qwaqwa1960]

Cool. What ever happened to "plumbing"? Seems that was all one used to see for even moderately high frequencies. Now, it's copper to >100GHz! Can't remember the last time I saw plumbing...

[@at0977]

Can you provide a link to the “connector types” document?

[@Megabean]

Amazing, if I may ask, how do you have access to such expensive hardware?

[@Thesignalpath]

Of course you can measure the difference! It is more than just raw performance, it is also performance over time, repeatability and life span of the metrology grade connectors. These things are not the latest consumer product where you can 'fool' the buyer through exaggerated advertisement. They are scientific products. They are either objectively better, or they are not. For example, device modeling with production grade components is essentially not possible.

[@Thesignalpath]

You should not take my word for it! You should try it for yourself. Setup a 67GHz VNA with -20dBm launch power and do a two port calibration. Then measure the SP of a DUT with better than -40dB return loss. Disassemble your setup, and re-assemble it. Recalibrate and measure again. Do this once with metrology grade and once with production grade connectors that have been used at least 10 times. Compare the four measurements you just did. Guess which curves will be on top of each other?

[@douro20]

Don't they use solid gold or silver center pins in some of those connectors?

[@cthree87]

Your latest video is so pornographic it should not be legal. By far the most interesting part of my day. Thanks! Again!

[@fpliuzzi]

Hello, maybe you could demonstrate some basic fundamentals of using a VNA by using the Diligent Analog Discovery's basic network analyzer function. You could design your examples to fall within its limited 10MHz bandwidth. Many EE students have this $99.00 device and they would be able to follow along with your demonstration. In the video you could also point out the benefits of using a lab grade VNA too so their advantages would be fully appreciated. Just a thought. Keep up the good work.

[@slvtx]

Very nice and helpful, thank you!!!!

[@koggism]

The electronics engineer of the future would do well to take a class in plumbing/ mechanical engineering, with these pipes and cavities. I believe at these GHz freq's the signals/current travel on the surface of the metals due to skin effect. Very interesting video.

[@siliconmcleod]

Love the videos, keep up the good work.

[@98505177229850590818]

Great video I wish you had detailed schematics draw for better understanding thanks

[@BalticLab]

Man, I wanted to make a video like that! Well, I guess I was too slow. Now I gotta come up with something better. Good job!

[@lucasng3330]

Why the freq ways end at 18GHz, 26.5, 40, 53 etc?

[@Daniel-ib5bx]

That.. Was..AWESOME!!!

[@visionofwellboyofficial]

Can i salvage a (old, defunct) spectrum analyzer and get these components for free? If not, please explain why. Thank you

[@qwaqwa1960]

Ah, you show some basics. Hope for more...

[@combin8or]

At the end of your video you said that the probe was rated to 40 GHz, but you didn’t describe how impedance could physically work with such a dynamic system. :( Also, now that you have access to a decent/fancy X-ray imager, it would be fascinating to see the actual RF traces that you’ve glossed over throughout your videos 🙏

[@sparkyy0007]

Microwave, the progression of electronics.... to plumbing.

[@matseriksson8177]

14:10 When you use a torque wrench of that type it is important to apply the force (with your finger/hand) at the intended distance from the tip pivoting point, othe wise the torque will be wrong, especially if you are using a wrench with a long tip like that. Compare the effekt of "L" in this page www.engineersedge.com/manufacturing_spec/torque_wrench_1.htm

[@matseriksson8177]

@Dusty 99 Yes it sure is.

[@matseriksson8177]

@Dusty 99 Yes The type of torque wrench that is used in the video just after 14:10 has a built in "crow foot" (as required since the pivot point, which clicks/releases at a defined torque, is offset from the center/rotation point of the nut which is tightened, as that is where the cable is). And if you use a crow foot the actual torque depends on where you apply the force with your hand. See en.wikipedia.org/wiki/Torque_wrench#Extensions Imagine you apply the force exactly at the pivot point, then the wrench will not release/fold at all. Imagine you apply the force 0.1mm away from the pivot point, then you need to apply a considerable force to make the wrench to release/fold. Imagine you apply the force at the far end of the wrench, as far as possible away from the pivot point, then the force needed to make the wrench release/fold is much lower. The effective torque is constant at the PIVOT/FOLDING point when the wrench releases, no matter where you apply the force, but NOT at the nut center point, as the force-to-nut distance is NOT proptional to the force-to-pivot point distance.

[@cowshittt]

Isnt the taper on the frequency doubler a balun, not Z matching

[@cogoid]

Yes. It is a microstrip-to-balanced stripline balun driving a full bridge rectifier, then another balun to get back to single ended output. Marki has an app note on their doublers.

[@vaualbus]

ha I don't know that. thank. Whivh is the maximum frequency use today? among the tera hertz?

[@vaualbus]

But how you make this higher frequency? On agilent site I've see signal generator up 63ghz and scope. How you can make this higher frequence? There are some Spectrum analyzer to go up to 1THz what ridicolous way you should use to carrying this frequency?

[@cogoid]

Look up "Virginia Diodes" they make frequency multipliers that go into THz region.

[@songhaiwang1154]

could anyone give some introduction of phase shifter for millimeter wave?

[@vaualbus]

and you should ti use a63ghz oscilloscope why on that frequency you don't use a spectrum analyzer?

[@dipi71]

I just can’t visualize all those moving charges doing their stuff within these components: electrons and electron-holes wiggling to-and-fro in the field within those components billions of times every second. Is this even a good way of thinking about this, or is it more like oscillating electromagnetic fields?

[@joblessalex]

Thanks for showing me all that black Magic. I don't even begin to understand this!

[@utkusevincli]

Subscribed thanks to this video

[@douro20]

That HP bias network sells for about $950 on Ebay...

[@pitot1988]

For those want a pdf version of this table, www.thesignalpath.com/uploads/TheSignalPath_EP26.pdf

[@norbs]

Really great video.

[@visionofwellboyofficial]

Thank you for the explanation

[@dtiydr]

18:29 "A few thousand dollar" Holy shit.. but it is the best you can get though.

[@Thesignalpath]

I don't think you could manufacture a 1mm connector in the 18th century. You couldn't even begin to model its behavior, let alone to test it. The issue really is just volume. And as I said, many more people are working in the >50GHz space compared to a decade ago. As a result these components have become cheaper.

[@ToniT800]

Somewhere in a lab:"Where the fuck all connectors gone!?""