RF Design-7: Broadband and Multi-Stage Impedance Matching Design

[@debapratimdhara1385]

After a long day of HFSS simulations , I started watching this video and watched the entire video without any break. I never knew that I could simultaneously learn and relax at the same time. Please continue making more videos.

[@BhargavaAnurag]

Great to hear!

[@nasserhussain5698]

I agree

[@hejoshua4170]

This is exactly what I feel after transforming from Hfss to ADS, fundamental circuit knowledge is always helpful to refresh your understand ing

[@akhileshverma7105]

This is the complete package for RF Designers, excellent series. Thank you sir

[@BhargavaAnurag]

You are most welcome

[@lasithathilakarathne3205]

I have been watching these tutorials too much now. I am addicted to ADS.

[@BhargavaAnurag]

I will take "addiction" in positive manner here...😊

[@bganeshp5484]

Thank you for tweaking me in the world of RF design

[@BhargavaAnurag]

You're Welcome....😊

[@kevinpan7133]

excellent video, can't wait to see more series videos about RF ciruit and ADS!

[@BhargavaAnurag]

Hi Kevin, There is plenty of them already uploaded. Hope you will find them useful for your work and learning...👍

[@rockydm4266]

Dear Sir all your video are very interesting very compleate of arguments, please do not stop. thank you very much

[@BhargavaAnurag]

Keep watching

[@HamzaKiani3]

Thanks. This will be a good point for me to start wideband matching.

[@BhargavaAnurag]

Great to hear!

[@ravidurbha3857]

Very useful tutorial and fast approach for designing matching networks

[@BhargavaAnurag]

Glad you liked it

[@emmanuelmuganza3426]

you are just amazing sir ... thank you !

[@BhargavaAnurag]

You are most welcome

[@hamidk4772]

Outstanding job 👍.

[@BhargavaAnurag]

Thank you! Cheers!

[@devanrpeters7229]

Great Videos! keep up the good work. I would like to see some tutorial videos on designing the Gilbert cell mixer as well!

[@BhargavaAnurag]

Noted!

[@ltteogoali7251]

love your work, waiting for more

[@BhargavaAnurag]

Thank you! Will do!

[@satyat7051]

Very useful video. Thank you for this. Have you also uploaded something regarding basics of load pull design techniques for power amplifier ?

[@BhargavaAnurag]

Hi Satya, Thanks for your comments. Loadpull video is not created yet and it will come once we reach PA design topic

[@ajithsrighakollapu3569]

great tutorial , which is worth

[@BhargavaAnurag]

Glad it was helpful!

[@aliesmaili3213]

Great video! Thanks alot.

[@BhargavaAnurag]

Glad you liked it!

[@brothercavil491]

Great videos, thank you. Any chance you could run through one example of doing the unified transformation (mentioned at 18:40) by hand? I'd prefer to understand the theory behind that. Thanks again!

[@BhargavaAnurag]

Unfortunately not..!! Idea is to show how to do these using utilities in the design tool directly.

[@indhujanivetha2138]

Could you please make a detailed video on Broad band rectifier and its impedance matching?

[@user-nt4nm4fb3u]

Can I do the optimisation method which is showed in this video instead of Load/source pull method to achieve the input and output matching network for the broadband Power Amplifier design? Is it OK to not to obtain the optimum source & load impedance of active device, if I use the method which has shown in this video?

[@BhargavaAnurag]

Yes, you can. However, that like using blind method which may not be always optimum but you can still work using that concept by optimizing the matching network component values and optimize for output power and efficiency.

[@359623495]

Hi, for the input matching, could you explain why on Smith chart the matching starts from Zin to 50ohm? Gonzales book says starts from 50ohm, page 165. Also, on your video 12:28, why not use conjugate match 26+j6 for Zload on Smith Chart?

[@BhargavaAnurag]

Hi, This is how the tool is designed, here we always start from the Zin or Zload and then reach 50Ohm. About conjugate match: ADS Smith Chart read out directly gives you the load impedance that you need to match and that is why you don't see * symbol on ZL terminal. Hope this helps.

[@Mohammedali-vc1rr]

Hi Mr. Anurag, thanks alot for your great support . A small question which is not exactly related to the topic of the video, if I want to separate/connect two circuits from each other (let's call them circuit 1 and 2) using a switch, while dont affect their impedance when switch between them is OFF-state and same as short circuit when switch in ON-state, could it be done using a switch? I tried to use a GaN HEMT switch but the second circuit wasn't completely off when switch in Off-state (that it wasn't appeared as an open circuit when circuit 1 is looking to circuit 2).. Can you please help which switch technology I should use and if you have its model number it will be great . really thanks sir for your videos and may success be with you always

[@BhargavaAnurag]

Kindly check this video that uses a Switch to create different topologies: kzbin.info/www/bejne/iaSUoqSkfqecr9k As long as you keep Isolation to a very high number to avoid any loading etc then you should be fine...👍 Hope this helps.

[@Mohammedali-vc1rr]

@@BhargavaAnurag Thanks a lot Sir, So do you think that SPDT are based on transistors in RF applications as what I observed in some papers which isn't a simple mechanical switch? , And for LNA application what is the minimum parameters for the SPDT (switch) you think I should consider ? Many thanks Dr. Anurag

[@BhargavaAnurag]

@@Mohammedali-vc1rr Yes, SPDT switches are based on transistors (or MEMS) and it's never a mechanical switch in RF applications...😊

[@arijitdas1884]

Hello @BhargavaAnurag Sir. I would like to have your insight or advice relating to one issue I have often faced (and I continue to face actually) in the context of broadband matching circuit design. After doing EM Cosimulation of an RF Chain layout and placing all the required components S parameter files, I do get proper whole chain S parameter response by placing interstage LC matching circuits wherever needed. Here I use practical components model files of the capacitors and inductors to form the matching circuits (making use of Murata library capacitor and inductor model files). However, later when proto testing of the same layout with the same matching circuit components takes place, the output power comes quite less than expected, and only after removing all the matching components and placing series 0E resistor instead is when the power comes atleast close to what was expected. This is found especially true when shunt inductors are used in matching circuits and some active component(s) in the chain is/are involved. Have you ever faced such thing? What can be the reason behind this? Am I missing something? Can you kindly provide your valuable feedback on this.🙏🙏🥺

[@BhargavaAnurag]

This simply means there is something wrong in your simulation setup and there is something you are missing to model in simulation e.g. PCB parasitics and you really need to perform EM/Circuit cosimulation by bring in your PCB EM model in schematic and then integrate the L & C to see the actual assembly performance. Watch EM/Circuit cosim video on my channel and contact Keysight tech support team if you need more help.

[@arijitdas1884]

@@BhargavaAnurag Thanks a lot for your response. Kindly note that I did EM cosimulation only every single time, so pcb parasitics are already taken care of I think...

[@BhargavaAnurag]

@@arijitdas1884 It means someone from our local field AE team or tech support team needs to take a closer look at your design and then offer resolution.

[@rfengineering2022]

Thank you very much sir for great explanation. In the broadband matching network design, I have noticed that you always set your Q-factor to be 1. Why is it so please? I mean considering your specs, the center freq is 6GHz; for a desired bandwidth of 2GHz, the corresponding Q-factor should be 3 instead of 1. Kindly help me to understand, I am designing a broadband M/Ns for a PA but my Q-factor is 3.5, and if I try to set it to 1, my source or load impedances obtained from loadPull simulation are no longer within the Q circle of 1. How to deal with such a case ? Thank you in advance

[@rfengineering2022]

I want to mention that I am using the distributed elements for matching

[@BhargavaAnurag]

Q is dependent on your center freq and Bandwidth. For your case it is 3 so you can go ahead with that...👍

[@hzlin4147]

Great videos! Helped me a lot as a student. One question though is that how do you practically get the inductance values the optimizer produces? Not all inductance values are available for purchase. (I know for higher frequency you can use distributed elements instead, but for ~100MHz that does not seem to be an option.)

[@BhargavaAnurag]

Optimization is only the 1st step in the design process. After you achieve the optimized results, you need to change the values to the nearest standard values and reoptimize the circuit perform if needed. If you have circuits involving both L and C's then try to change inductors first and optimize the capacitor values as you normally have more values available in capacitors. You can also run constraints based optimization as shown in Constraints based RF Circuit Design video....

[@sidhath98]

Excellent video. I have one question though. How do you perform broadband matching when the impedances to be matched lie outside the Q circle?

[@BhargavaAnurag]

Good question! Idea in your case would be to bring that impedance point inside Q-circle and then always remain inside to reach to 50 Ohm point. Broadband is the generic term and you need to 1st define what do you mean by broadband and if it is 1 octave or more, tighter your considerations become and you cannot achieve 1-octave or more bandwidth if reactive part of your impedance is more than 2.1 times of the resistive part.

[@sidhath98]

@@BhargavaAnurag Thanks a lot. Will keep it in mind. Your tutorials have been really helpful.

[@rushikeshtayade4851]

hello sir as i am simulating a 5G LNA it is really difficult to follow this tutorial without knowing the "snp" file that you have used, so could you please drop a link for it as i am unable to find it on the internet after searching a lot. And also thank you for such detailed tutorial.

[@BhargavaAnurag]

Hi, I am not sure why you need this file? You just follow the process and adapt as per the file which you have to use for impedance matching network design. Kindly spend some time in looking at S11 and S22 of your file in the desired freq range and you should be able to make it work...👍

[@prakhargupta1409]

Thank you for the tutorial. It would be very nice if you please attach links to the videos you ask the viewers to refer before watching this video.

[@BhargavaAnurag]

You need to go to the main channel page and look at the playlist and the topics are self explanatory to find the relevant topic of interest.

[@peimanparand]

awesome, thanks

[@hongpan2032]

Great video! I enjoyed every minute of it. I have a questions: What happens if I want to achieve an impedance matching within constant Q of 1 circle, but the load to be matched is outside of Q of 1 circle? can I still achieve the bandwidth based on constant Q definition? Does the Q of the load to be matched impose a limit to the bandwidth we can achieve? Thank you very much.

[@BhargavaAnurag]

Good question! Yes for ultra wideband matching requirements there is a relationship between resistive and reactive part of the load. If the imp. point is outside Q-circle, you will have to bring it inside and try to stay inside but even that your bandwidth might suffer a little bit...

[@hongpan2032]

@@BhargavaAnurag Can you please provide a couple reference material for the wideband matching that you talked about in this video? paper, text book section etc. I am interested in getting into the details and understand it more. Thank you very much!

[@divyanaveenan]

I dont see a smith chart matching option my parts. do they have to be installed separately?

[@BhargavaAnurag]

Hi, No, Smith Chart is default part of ADS for past many years so you should see that under Tools menu in schematic.

[@НиколайТ-о3з]

Hello! Tell me, please, what is your file with S-parameters? In what form are the values themselves presented? The program returns an error to my file: load file contains invalid impedance values.

[@BhargavaAnurag]

Hi, I used a generic touchstone file which wasn't fully matched to 50Ohm, nothing special. If you are having issues then contact Keysight tech support team: [e-mail]

[@saifulsyazwanyusoff9075]

Nice video and clear explanation.. If you have time, can you make video explanation how to design the impedance matching using stub that can operate for two or three frequency. For example : ZL1 @ f1 , ZL2 @ f2 Thank you very much..

[@BhargavaAnurag]

Kindly check out Part2 video of the Power Amplifier design to get an idea how you could do something like that.

[@saifulsyazwanyusoff9075]

@@BhargavaAnurag you mean this video kzbin.info/www/bejne/n3KvgHSvfLeGitE thank you very much

[@BhargavaAnurag]

@@saifulsyazwanyusoff9075 Yes, you can refer to the technique shown for the output matching network design

[@saifulsyazwanyusoff9075]

@@BhargavaAnurag Thank you for nice video. Can I need your help to make a video how to design impedance matching using transmission line (in ADS use MLIN, TL1) which is can operated for more than one frequency especially for high frequency. I try study about it from this attached article journal and not quite understand although it simple circuit use (impedance matching and single stage rectifier): - bit.ly/3eq4ywy I try follow the explanation from this video kzbin.info/www/bejne/n3KvgHSvfLeGitE to make it but I cannot reach my objective to design it. Hopefully, you can share some knowledge or the proper step to design it. Thank you and I really appreciate.

[@alirezajafarieh226]

Very nice, thank you🤩

[@BhargavaAnurag]

You’re welcome 😊

[@MR-qi7fc]

Great videos! And I have a question. How do I match the impedance of a circuit with two bjt's connected in parallel? The problem is that impedance will change as I do impedance matching in other positions. I don't know how to get multiple positions to meet impedance matching at the same time. Thank you!

[@loho6894]

I am by no means an expert, just a student working on his diploma. Copy your components and add different terminations in the same schematic. Then run an optimization by having goals for all of your configurations. If you don't know how to do optimization check the relevant videon in the ADS series of the channel.

[@BhargavaAnurag]

Please contact Keysight tech support team....

[@khalid999zzz]

33:46 Sir whenever I want to upload the S2P file in the bandpass filter domain, it shows me an error "read_datafile() failed server error"

[@BhargavaAnurag]

I am not sure what could cause that error. Can you simulate the S2P file along first to check if the format is proper and ADS can read it well. If it works then you should be able to use it. Incase that doesn't solve the issue then contact Keysight tech support team for help.

[@khalid999zzz]

@@BhargavaAnurag sir it is working in plane schematic but whenever for the purpose of impedance matching it show that error.

[@khalid999zzz]

@@BhargavaAnurag sir it is working in plane schematic but whenever for the purpose of impedance matching it show that error.

[@prakhargupta1409]

In the multi-stage impedance matching, you used the touchstone file containing S parameters of the components, to obtain the input and output impedance. But, S-parameters would help only when the other ports except input are match terminated, which is not the case in general. What if we use the SP probe to determine the impedances and put these values in the matching network design utility? Do you think we would get better matching and quicker optimization results?

[@BhargavaAnurag]

There are different ways of doing things and you can proceed with whatever you feel is appropriate for your application. All techniques will have little pros & cons and ultimately you will need to optimize the overall circuit.

[@СергейАлёхин-л2у]

Simple English in this video is good advantage. I can easily undestend the information from this video despite than English isn't my native languege.

[@BhargavaAnurag]

Glad to hear that

[@AnilKumar-ds1he]

Anurag, thanks for the nice set of tutorial videos. Really speeds up the learning. In optimization, is it possible to restrict the component values to within the standard values available ?

[@BhargavaAnurag]

Yes, definitely. Watch constraint based RF Design video in the same playlist

[@MrRijubratapal]

Very detailed explanation as usual. Any tips on types of matching network to choose from low/high/band pass network under impedance matching utility?

[@BhargavaAnurag]

Hi Rijubrata, I think I covered that in the tutorial at some point. Choice can be made depending upon the matching requirements, component values suitability and amount of components you can accomodate as per board area available etc etc. e.g. for some requirements HPF topology is more suitable as you can achieve matching with lesser components but the component values may not be so good so you can opt for LPF topology which may require 1-2 more components but component values are more manageable etc. Hope this is clear up to some extent?

[@RajeshDas-nw2cv]

can we use a antenna design from hfss and use it as s2p file here

[@BhargavaAnurag]

Yes, you can but Antenna usually will have s1p file which can act as load or source for your matching network design

[@RajeshDas-nw2cv]

How to make dual or multiband matching network operating at multiple frequency....for broadband u used Q factor method....For multiband what method should be used since in smith chart we can put only single frequency value to make a matching network..

[@eeemagic7729]

Hello, actually i wanted to practice with your tutorials. But, i did not have access to s2p files you were using. How can I access those files?

[@BhargavaAnurag]

You can take any S2P that is not already matched to 50Ohm so no magic here...😊

[@abdullahali-lc8mv]

Thanks a lot for these nice videos sir. Actually, when I apply this in a two-stage amplifier ( cascode amplifier). I have matched the input stage the same as what you did ( S11

[@BhargavaAnurag]

That's not a problem and will happen as you are dealing with bilateral circuits. You can run optimization on the overall bench to get the performance back. Another useful trick is to remember to start impedance matching from output side first, you will still have the impact on the overall bench but it would be much smaller than if you start from input side...😊

[@shwetagarg9276]

Very helpful videos 👍

[@BhargavaAnurag]

Glad it was helpful!

[@debapratimdhara1385]

Can you please share some resources on multi-antenna matching network? Great job with the videos though!

[@BhargavaAnurag]

I will try! If you have some papers/references then kindly share the links here...

[@debapratimdhara1385]

@@BhargavaAnurag ieeexplore.ieee.org/document/4718024 www.hindawi.com/journals/ijap/2019/1934726 ieeexplore.ieee.org/document/1643610 ieeexplore.ieee.org/abstract/document/7835289 ieeexplore.ieee.org/abstract/document/6748889 ieeexplore.ieee.org/abstract/document/6697539 ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=8658000

[@BhargavaAnurag]

Thanks Debapratim...I will check these out in near future...!!

[@parul426]

Sir could uou you please make video on impedance matching between a wideband antenna and rectifier.

[@ssri2308]

i am also interested in answer

[@minakshmishaw8036]

please add some video regarding rectenna design

[@BhargavaAnurag]

Hi Minakshmi, Will try to do so in near future...👍

[@dushayantkumarsharma7398]

very useful

[@BhargavaAnurag]

Glad to hear that

[@prakhargupta1409]

How to decide on whether to use Smith Chart tool or Impedance matching tool ?

[@BhargavaAnurag]

Its your preference. Smith Chart you with the interactive experience and you can manipulate the network the way you want whereas in Imp. matching tool you don't have much control. In Smith Chart you could actively handle 1 freq at a time whereas in Imp. matching tool you could do broadband matching network. So in nutshell it depends on what you want to do and chose one over other based on that.

[@gabrielgiribaldi2953]

I tried to reproduce the results on one of my touchstone files. Nevertheless, the impedance matching tool is not able to read it. It is weird, since the same file is used as SnP component in S_params simulations. Any idea? (P.S. these tutorials are amazing)

[@wk-de1jj]

change the location of the SnP file and try again, it worked for me when i used USB flash driver.

[@BhargavaAnurag]

Difficult to comment without seeing the file. Contact local Keysight tech support team for help

[@adnannadeem6223]

Please share the link of website where you get the Snp files so i might help me understanding the tutorial well.

[@BhargavaAnurag]

Hi Adnan, You can look at different vendor websites like for Analog Devices or Minicircuits etc and you would be able to get the SnP file from their website.

[@adnannadeem6223]

@@BhargavaAnurag okay sir.

[@mohitgrover1303]

Sir please explain load pull and source pull

[@BhargavaAnurag]

Hi Mohit, I will surely do that when we reach to PA design topic...

[@mohitgrover1303]

@@BhargavaAnurag OK Sir

[@KimJones-xd5bt]

The only concern here is that the S11 and S22 s of the block changes after the first stage of previous stage matching due to loading S12

[@BhargavaAnurag]

Yes, that will be the case for all bilateral devices...😊

[@santhoshkumarsinguru8531]

Please tell any one, how to get s2p file for cmd240 amplifier from qorvo

[@BhargavaAnurag]

It is available at this link: www.qorvo.com/products/p/CMD240#documents. Look for lower right side under S-Parameters

[@vicafrederick3386]

"Greed has no limit my friends" lol

[@BhargavaAnurag]

correct...😊

[@anisachour916]

Svp question monsieur?