RF Design-13: Getting Started with Load Pull Simulations

[@user-nt4nm4fb3u]

Sir, thank you very much for the full & easy explanation tutorial about Power amplifier design in ADS including load pull simlation technique. Your channel is the only channel which providing what I really need in these days. Because I'm doing my final year Engineering research project on broadband PA design in my university. This is so helpful. Thank you....🙏

[@BhargavaAnurag]

You are most welcome

[@yaoyucao9601]

Dr Anurag is really a genius, these are definitely the best ads tutorials

[@BhargavaAnurag]

I am glad you liked it....👍

[@seyedurmanghozati5343]

As always excellent video Anurag. I have learned a lot from your videos and thanks for that. Just noted that some people asking in the comments that they are unable to plot GL. To fix that they can add the same GL equation from the schematic in the display window with added "expand" syntax to multiply to the j product as GL=rhore+j*expand(rhoim), thanks!

[@user-ci3cm5rg3x]

Thank you!! I got GL from your comments, but I still can't get effective PAE_contours 😵‍💫Do you have any ideas? (I got PAE, but the equ PAE_contour_ex in data display is invalid(red))

[@user-ci3cm5rg3x]

Sorry, I add a sweep 'pin' in HB controller, so the format of PAE changed, so I can't get the right PEA_contout ues the same equ as the video shows.

[@KimJones-xd5bt]

Thanks Dr. for the presentation, Few notes: When we use this load pull design guide, we also include the gate and drain bias components, and the DC decoupling caps, because they will rotate the optimal load and sources points. The stabilization resistor, we usually shunt it with a small cap 3pF to reduce its effects at high frequency, since the instability is at lower end of the band. Personally, I only obtain the load and source pull values from he data sheet to use as a starting point, then I run tuning the load and source manually for the desired power and efficiency, which can be all done in few minutes. This way I avoid this template tool, because lots of its is unclear and not clearly presented, and brings in doubts.

[@BhargavaAnurag]

Hi Nezami, There are different approaches you can use to design a circuit and as long as it works for your specific case it should be fine. However, in a larger context these kind of analysis gives you much better insights into device's behavior and you can take educated design where you want your device to operate and come up with your circuit design topology accordingly. Just choose whatever way works for you....😊

[@ShivamMishra-nc6wz]

Hi.. After including all these components drain and gate biasing circuit and decoupling cap ,when I am running the load pull simulations using load pull template. I am getting no efficiency at low frequencies although swiping all over Smith chart .does this template not work like this, and one more point my dc biasing circuits working fine individually .Where I am going wrong.Can you tell.thanks

[@tanghidrashid3842]

Excellent video! Thanks, Anurag!

[@BhargavaAnurag]

Glad you liked it!

[@dule8695]

Thank you for your valuable contributions

[@BhargavaAnurag]

Welcome!

[@Scuba_Bro]

This is great Anurag! very helpful!

[@BhargavaAnurag]

Thanks for the feedback...👍

[@PastoriciciRobertYO4HFU]

Excellent explanation!

[@BhargavaAnurag]

Glad it was helpful!

[@ezgis7268]

Thank you so much for the video, it was really helpful! Right now I am trying loadpull on a differential PA and was wondering how we can do that using Load Pull Instrument in ADS?

[@BhargavaAnurag]

For anything differential, simply use a Balun at both ends.

[@user-ci3cm5rg3x]

@@BhargavaAnurag And doing loadpull after the Balun?

[@username-vn5eh]

Excellent!

[@BhargavaAnurag]

Thank you! Cheers!

[@shahriaraashikkhan]

amazing, just amazing!!!

[@BhargavaAnurag]

Thanks for liking the video...👍

[@hindujaicchapuram5809]

Informative !! . Anurag have you ever tried loads pull simulation with an active source injection ? Is that even possible in ADS to add an active source and sweep across different values of a2

[@BhargavaAnurag]

Hi Hinduja, I haven't tried that but it seems quite possible to do so.

[@Astartos22]

Hi Anurag, I've been really enjoying your videos, they've been extremely useful to me. Would it be possible for you to create a video on how to use ADS's impedance matching utility (not the smith chart tool, this tool appears to be a part of newer versions of ADS)? Thanks again.

[@BhargavaAnurag]

Pls see RF Design-7 video here, here is where it starts talking about Imp Match Utility: kzbin.info/www/bejne/jqG1Z6ilgL6ZfJY

[@Astartos22]

@@BhargavaAnurag Perfect, thank you!

[@nasserhussain5698]

How to use diodes instead of using transistors, as in RF rectifiers to calculate and draw efficiency vs. impedance on the Smith chart....

[@BhargavaAnurag]

Hi Hussein, I have never tried that but as long as you know what you want to do, there will be a way to accomplish your goal. For example, if you know how to calculate Rectifier's efficiency then you can use the same method explained in the initial part of the video to sweep the reflection coefficients and then compute the efficiency on every load sweep point and then plot contours on the Smith Chart.

[@balkrishnaarora8776]

I have been trying to learn rf and ads using your channel thank you so much I have a doubt that in your harmonic balance you have used input with 50ohm In load pull simulation you used 5 ohm How can we ensure that matching is perfect at the input side of mosfet

[@BhargavaAnurag]

5 Ohm used is not for perfect matching but just enough to let that terminal load up properly for proper load analysis

[@balkrishnaarora8776]

Means if my input source impedance is changing in a range where enough power gets into it(say 2-10ohm) , that will not affect the load pull result ?

[@user-nt4nm4fb3u]

Sir, what is the name of the website? To obtain that circuit which you've shown at the first. And how can I download it? Is it free to use?

[@BhargavaAnurag]

Hi Christy, I provided the links in the description box. You can download some additional templates from the Keysight Knowledge Center, you need to register on the portal and should either have a legal copy of the software or an active evaluation license.

[@nazlizargarpour1854]

Thanks alot for another imformative video. I know load pull simulation is for power amplifier. Just I am wondering, can we use it as a tricky for designing Low noise amplifier too? Thank you in advance for your answer

[@BhargavaAnurag]

I am sure what value Loadpull simulation will bring to LNA design as LNA is always matched to 50ohm for best possible matching.

[@stevegao2888]

One more question, what's the best way to do differential loadpull? if the load tuner used in the example is connected differentially, will that work? Thanks.

[@BhargavaAnurag]

I never tried it but you might want to try using ideal system balun models at input and output. For more details kindly contact local Keysight tech support team for help.

[@MR-qi7fc]

Sorry, I have a really stupid question. Why can't I find the variable GL in the data display? All other rhore, rhoim can display it.

[@BhargavaAnurag]

You need to add that variable in the Output tab of the simulation controller.

[@MR-qi7fc]

@@BhargavaAnurag Thank you very much.

[@divyanaveenan]

I am using a modified version of this schematic and (1) i get an error saying no source was found. How can i locate where this error is from? (2) the equation GL=rhoim+j*rhore is not recognzed, i get the error that is not valid for parameter "Variable Value"

[@BhargavaAnurag]

Kindly recheck the simulation setup carefully as it seems you forgot to connect the P_1Tone source at the input. Pause the video at right points and compare your setup with the one shown in video. Its pretty straightforward so should work easily.

[@soroushmoallemi774]

Thanks for this great video. about the first part of the video, may I ask if we can define the reflection coefficient in a polar format so we can sweep the whole smith-chart? If we define the reflection coefficient in a rectangular manner, by getting close to the edge of the smith-chart, the corner of the defined rectangle gets out of the smith-chart and the power values become imaginary and data display will not show any result. Could you please explain how we can define the reflection coefficient (all the black dots) in polar format? Thanks again for your amazing videos

[@BhargavaAnurag]

Hi Soroush, Sure you can define it either rectangular or polar format. You could check out the parameter on the load pull tuner. Most of templates in the load pull designguide template are of polar format. For more help, kindly contact your local Keysight Tech support team.

[@stevegao2888]

Hi Anurag, I saw in the test bench, there is no dc simulation. but in your plot window, dc curve tracer is plotted. There is an DCIV equation=Step1_DCIV_simulation..ids.i. How does this one work? Thanks

[@BhargavaAnurag]

Hi Steve, That comes from a different schematic and dataset, that's why you see syntax as ..Ids.i where Step1_DCIV_Simulation is the dataset name. Sorry for the confusion....😊

[@stevegao2888]

@@BhargavaAnurag thank you.

[@mugheesahmad1988]

How to design an RF limiter?

[@PhuNguyen-hx9qm]

Thank for your excellent video! Can you tell me why you selected Z_load_center_fund = 20+ j*0 and not 50 + j*0? What does this mean and on what basis can we choose the appropriate Z_load_center_fund and Z_source_center_fund?

[@BhargavaAnurag]

Power devices aren't internally matched hence there is no point looking around 50Ohm point. Most of the times you will find suitable point at low impedance level hence the reason.

[@muratdagdelen8163]

Sir, is it possible that you make a video about "options > expert settings" in EM MoM setup?

[@BhargavaAnurag]

Hi Murat, That option is used in very few occasions hence there is no video for that. Anything specific you are trying to achieve?

[@muratdagdelen8163]

@@BhargavaAnurag Hi Sir, I am trying to achieve reducing the simulation to simulation S21 phase variation by meshing MIM Caps densely. I could mesh the whole structure densely, but the simulation time would increase. I find that expert option has some setting about MIM Caps that would help me to mesh spesifically MIM caps densely, but help file of ADS does not explain it how it works.

[@rddinesh4]

Hello sir, What is the difference between a variable and measurement equation in ADS? I'm always confused. Could you please clarify? Thanks

[@BhargavaAnurag]

Variables are something you define to be assigned to components in place of their fixed values on your schematic so that you can sweep it or keep your circuits symmetrical etc Measurement Equations are for your measurements i.e. post processing like for example if you want to compute mag(S11) then you will define it in Meas Eqn and not in the variable block...😊 Another difference is that anything you compute in Meas Eqn can be used as an Optimization Goal to optimize your circuit component values, if needed. Hope it is clear now.

[@QuanVuVan98]

Sir, i don't know about function of res 5 Ohm in your schemetic . Can you recommend some doc or book that dicuss abour bias for transisitor RF ? There are quite a few documents about bias RF transistor used resistor, most of them are very general, its doesn't mention much about RF , thank sirr very much

[@BhargavaAnurag]

Hi Quan, 5 Ohm resistor is used for stability purposes. Kindly watch all 3 videos of PA design and 2 videos on LNA design then you will understand the complete context...👍

[@rddinesh4]

How to use load pull to find optimum second and third harmonic impedances while designing a class F PA? How to find the appropriate stability circuit for my design? why don't you choose 14+j*8 as optimum load point, where PAE = 64% and Pdel = 41 dBm? I used the load pull template from the design guide and later wanted to remove it....How can I remove all the cells related to it at one go?

[@BhargavaAnurag]

Its difficult to provide this kind of support on KZbin, kindly contact local Keysight support team for help.

[@MrRijubratapal]

Was eagerly waiting for the PA tutorial. This marks the beginning......those tough data processing stuffs are very well explained. Glad to know that you will go from scratch so that beginners like me will surely be benefited. On what basis you are setting Z_Load_Center_Fund = 20 +j*0? P.S You were muted in the recording few times.

[@BhargavaAnurag]

Hi, Thanks for your comments. Yes, my headset was having some issues resulting in mute. How did I arrive at 20+j*0 will be answered in the next video...😊

[@sanmithabs8074]

What do I do to increase my output power levels? I'm getting PAE of 66% and my Pavs is 27dBm. I'm using the same Cree device

[@BhargavaAnurag]

Hi, Its difficult to answer the query without looking at the setup. If you are not getting required power level then look at the Smith Chart region where load pull is being performed or the bias supply voltage.

[@mmh1321]

Hi thank you so much for your videos they are really helpful. I am using CGHVIF006S and my design frequency is 9.4 GHz, I did ur exact steps and yet I am getting very small max gain like -1.690 which should be 15 dB. I don't really know how to fix that. Thank you

[@BhargavaAnurag]

You need to relook at your circuit and make sure everything is set properly. Make sure right DC voltage and currents are reaching your device and you are indeed looking in the right area of Smith Chart. If you are looking at wrong area of Smith Chart then required Gain may not be there. Try different zones in Smith Chart or refer to datasheet to see recommended Load Impedance to get good starting point. If you can't resolve the issue then contact your local/regional Keysight Technical Support team for help. There has to be something wrong.

[@eeengineers3958]

Sir, why drain bias voltage is too high? will it not be a problem in current low power SOCs?

[@BhargavaAnurag]

You select bias voltage as per the device. For the device I have used, its 28V as recommended in the transistor datasheet

[@stevegao2888]

Hi Anurag, thanks for the video. For the load-pull examples in design guide, seems source-pull are done at the same time. If I want to do load-pull only and keep source impedance at 50 ohm, what should I do? Thanks.

[@BhargavaAnurag]

Hi Steve, For any load pull setup under designguide, the source impedance will always be held constant and the value can be set by user. I am not sure where the confusion is coming from? One the side note, 50Ohm for Source impedance for a bare device without any input matching network is never the best choice....😊 Kindly watch all the related videos and the PA design videos and you will understand the proper way to get started with this.

[@stevegao2888]

@@BhargavaAnurag thanks for the reply. The confusion comes from Z_s is put into the source port in stead of 50 in the schematic.

[@yaoyucao9601]

@@stevegao2888 The Z_s_in is the input impedance of the transistor(the impedance looking into the the transistor) under certain load termination in the sweep.

[@sribhargav1620]

hello Anurag, can u tell me how to design a Class AB amplifier using 2N2222A transistor for 6.78MHz.

[@BhargavaAnurag]

Kindly watch my video on Power Amp design which is also class-AB. Only difference would be that it was GaN device I used and you would probably use BJT at such frequency and all other matching would be done with lumped components instead of using transmission lines as I did in my video. Class of Amplifiers is all about selecting the proper bias condition to select conduction angle of the transistor and for AB it should be between 180deg-360deg. Hope this helps...👍

[@annguyendang8388]

Sir, Thank you for your useful video. Can you point out how to sweep harmonics with a fixed fundamental impedance, and draw the harmonic contours.

[@BhargavaAnurag]

That's possible easily. Kindly contact your local Keysight support team for help.

[@divyanaveenan]

If you share the workspace of these examples. It will be helpful!

[@BhargavaAnurag]

I try to share the workspaces on Keysight Knowledge Center which supported customers can access but if I share this workspace then lot of people will see error as it required Cree library to be installed & configured and not everyone may have this library hence I didn't share the workspace.

[@eeengineers3958]

Sir how to perform load pull simulation for a frequency Doubler? Can you briefly give some idea?

[@BhargavaAnurag]

You can define the intended frequency and load in the loadpull instrument. For more help, contact local keysight support team

[@truptimishra8040]

Sir how can I register in Cree website?.. can u please give any link of that website.. and how can I add these devices to my workspace after the registration??

[@BhargavaAnurag]

Hi Trupti, Here is the weblink to register to get nonlinear models at Cree website: www.wolfspeed.com/tools-and-support/rf/rf-portal-access/ While registering make sure to type your full name, company name/address etc clearly to expedite the approval process. After you download the kit, you can add it in ADS as shown in this video: kzbin.info/www/bejne/jJeZlYmBgd2Yaq8

[@jonathanokocha4614]

Thanks for the video. Can you show the port on GCPW

[@BhargavaAnurag]

Hi Jonathan, I couldn't understand what is GCPW and what you meant by "show the port on GCPW"?

[@user-ci3cm5rg3x]

Hello!It is a very excelent videos for learning ADS! And I have a question for video "RFdesing-13", I don't know why I can't evaluate the VAR 'GL' in data display(it had code in SCH), and I got PAE, but the equ PAE_contour_ex is invalid(red)?Do you have any ideas?Thanks a lot!!

[@user-ci3cm5rg3x]

Sorry, I add a sweep 'pin' in HB controller, so the format of PAE changed, so I can't get the right PEA_contout ues the same equ as the video shows.

[@BhargavaAnurag]

Something might be wrong with your schematic setup, make sure you select GL in the simulation controller to be sent to output. Contact your local Keysight tech support team for help if you can't solve it yourself

[@tanghidrashid3842]

@anurag I tried to run your LoadPull_basics schematics which you showed beginning of this tutorial. Unfortunately, I am getting error from measurement eqn function "Pout_W and Pout_dBm". It says "undefined variable". Can you please guide me how to resolve this error? Thanks!

[@fourierv2]

Check that you have placed the appropriate wire/pin labels in the schematic as these are used within the Pout_xxx equations. Similarly, (also true for the PAE calculation), ensure that all current probes are labelled in accordance to the PAE formulation. Also, by default the PAE formulation is given as "PAE1=pae...", the equation used in this video is just "PAE=pae...", so make sure that you rename the PAE tool in the schematic (remove the 1 from the function/formulation, not the instance name).

[@BhargavaAnurag]

Hi Rashid, Base equation here is Pout_W which is dependent on "vout" node/wire label and Iout which is the name of the current probe at the output node. You should use the syntax/names as per your circuit setup. Once you fix this error, Pout_dBm should resolve as well. Hope this helps.

[@tanghidrashid3842]

@@fourierv2 thanks!

[@tanghidrashid3842]

@@BhargavaAnurag thanks! Will try that.

[@mcie6306]

thank you sir for your great tutorials I do have a question, why did you set the impedance of the input power to 5 ohm rather than 50?

[@BhargavaAnurag]

Hi, Power transistors have low impedances for both gate and drain so setting it to 5 gives you better chance to see good output power. Setting input impedance to 50 will usually not allow you to see optimum/good output power. I explained all this in the PA tutorial video, kindly watch videos 16-18 under the playlist

[@mcie6306]

@@BhargavaAnurag Clear, Thank you sir

[@sahandgh1624]

Hi, thanks for the video. Could you also let me know how to register on Cree website to get the design kit?

[@BhargavaAnurag]

That's pretty straightforward, go to www.wolfspeed.com/tools-and-support/rf/large-signal-models and click on Request Access to Wolfspeed's RF Portal. Make sure NOT to use any acronyms during registration, enter full name of yourself and your company/institute etc

[@oluseunwikiman]

How can you carry out load pull on a 3 stage amplifier

[@BhargavaAnurag]

Load Pull is usually done on a single transistor to figure out right impedance for matching purpose for power, efficiency etc. If you want to run 2 port Load Pull for a cascaded amplifier then look at the Keysight Knowledge Center from where you can download the example & associated PDF document.

[@oluseunwikiman]

This load pull is on a single transistor what of whole 3 stage amplifier

[@BhargavaAnurag]

Classical load pull is done on a single transistor/device. If you need to perform load pull for multi stage design then you can use the 2-port load pull template available on Keysight Knowledge Center

[@m.razlandhamirhamdan5658]

Hi! I have an error that the data "...did not converge -- associated data will be identified as '' flag in the dataset. I don't understand how to solve the error that I got and I really need your help, pleasee.

[@m.razlandhamirhamdan5658]

Hi again, I already solved the above issue but now I have another 2 problems. 1. My "GL" is invalid as if it doesn't exist. Why? 2. There is an error for all of my equations in the result. It says While evaluating "PAE_contours": Function contour(), argument 1 ("surface data") : Unable to contour the surface. This may be due to an invalid surface Please help me

[@BhargavaAnurag]

I would recommend you to contact local Keysight tech support team. It seems you are doing some error in the equation syntax...

[@markyu2970]

Anurag, can you share the ads project wrk of the Load-pull constant source power. The equations and the contour plots are quite different from my ads workspace. I like better the version you have. Thanks.

[@BhargavaAnurag]

Hi Mark, I posted links to the Keysight Knowledge Center in the description of the relevant videos for anyone to download the load pull example workspaces I used in the PA design video.

[@indrajitpatra665]

Hi Anurag, Thanks for the awesome insights into the ADS tool. I have a question i.e we are trying to build a simple cascode PA, when we need to do load-pull analysis, we should do load-pull for 2 transistors present in the circuit or only for the 1 transistor to get to the optimum load impedance?

[@BhargavaAnurag]

I would recommend to contact local Keysight tech support for help on this.

[@Sarah990]

Hello Patra I am trying to do exactly what you are doing and struggling with this point would you please let me know what are the recommendations of keysight tech center? is one transistor enough to load pull it or we need to load pull both of them?

[@ummehabiba-ef6jh]

I was not able to plot GL. Can you please guide?

[@BhargavaAnurag]

Kindly contact your local Keysight tech support team for help. GL formula is provided in the video, you can rerun the video to check that out.

[@ummehabiba-ef6jh]

​@@BhargavaAnurag Thankyou for your time sir. But GL formula is not working in my schematic. That is the main problem.

[@BhargavaAnurag]

Contact local Keysight tech support team for help.

[@user-ng2qy5pe4b]

Sir the problem I am facing that is I am not getting the contours fully. How to solve this?

[@uav-pixel9046]

same here

[@BhargavaAnurag]

Try changing the center of Zload and radius/area of your load pull simulation and it will work. If you can't figure out then contact Keysight tech support team for help

[@adnannadeem6223]

Hello Anurag sir, Can you please share CREE design kit because I have applied for registration to access their design kit but did not kit any response from them so it will be very helpful if you can provide it.

[@BhargavaAnurag]

Hi, I cannot share the Cree PDK as this is given to me under specific NDA but they usually take 2-3 days to respond to you so I would advice to wait or contact their local/regional distributor for help.

[@adnannadeem6223]

@@BhargavaAnurag okay sir. I would like to say that my RF skills are increases just because of watching different RF videos on your KZbin channel and following your design rules.

[@BhargavaAnurag]

@@adnannadeem6223 That's good to know. If you happen to fabricate your circuits and do the measurements don't forget to share your results with me.

[@adnannadeem6223]

@@BhargavaAnurag I will sir.

[@aligillani7107]

This video is not working fine. It is corrupted.

[@BhargavaAnurag]

Hi Ali, There is no problem with the video and it is playing nicely for me. Kindly recheck your browser setting or use different browser to watch the video....👍