Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

microstrip low pass filter design issues with sonnet

Status
Not open for further replies.

promach

Advanced Member level 4
Joined
Feb 22, 2016
Messages
1,199
Helped
2
Reputation
4
Reaction score
5
Trophy points
1,318
Activity points
11,636
I am trying to replicate the microstrip low-pass filter in SonnetLite from the book by **broken link removed**

However, the S12 graph shows that it is not a low pass filter at all. Could anyone advise why ?

jdoQZoL.png


box resonance warning analysis logs

Code:
Run 1:  Sat Jun 22 00:27:07 2019.  SL291418.107194.
        Em version 15.53-Lite (32-bit Windows) on DESKTOP-DG59KC2 local.

  Project:  C:\Users\kevin\Documents\sonnet\15.53-Lite\three-pole microstrip lowpass filter using open-circuited stubs\three-pole microstrip lowpass filter using open-circuited stubs.son.

    Frequency:  0.001 GHZ

      Sonnet Warning EG2680:
      Circuit has potential box resonances.
      Circuit: primary structure
      The estimated box resonance frequencies are listed below.
      The calculations assume lossless materials, with a box filled with
      only the specified dielectric stackup materials.

      4.140 GHZ  TE Mode 1,0,1
      4.673 GHZ  TE Mode 1,0,2

      Only lowest order modes considered.

      See the chapter on package resonances in the Sonnet User's Guide.
      Date: Sat Jun 22 00:27:07 2019


      Sonnet Warning EG2680:
      Circuit has potential box resonances.
      Circuit: left box wall SOC magnetic wall standard
      The estimated box resonance frequencies are listed below.
      The calculations assume lossless materials, with a box filled with
      only the specified dielectric stackup materials.

      4.167 GHZ  TE Mode 0,1,1
      4.700 GHZ  TE Mode 0,1,2

      Only lowest order modes considered.

      See the chapter on package resonances in the Sonnet User's Guide.
      Date: Sat Jun 22 00:27:07 2019


    Post-Analysis:
      Errors detected: 0    Warnings detected: 2.

Analysis completed Sat Jun 22 00:27:10 2019.

n422Rvv.png
 

Attachments

  • three-pole microstrip lowpass filter using open-circuited stubs.zip
    73.8 KB · Views: 58

There are few errors..

-Substrate dielectric coefficient has been marked being as 10.8 but you entered 4.4, why ??
-There was a an additional substrate just over-neath of the GND, why ?? Is it a simple Microstrip substrate ??
-Cell size is too coarse, decrease it to 0.1mm ( Circuit->>Box ) then check the connectivity ( Tools>>Check Connectivity)
-Box height is exaggerated, decrease -let say 10-15mm- ( It gives you Box Resonance Warning )
I have quickly corrected those errors..
curve.png
 
Last edited:
@BigBoss

Thanks for your tips. They help to make it more like a low-pass filter now.

However, I am confused with the layer configuration.

Does the following layer configuration makes sense ?

5RGY1M6.png
 

"Unnamed" dielectric layer should obviously be Air and it represents the air gap over the metallic layer.It's OK but the height must be higher to prevent the approximate effect the metal layer because Sonnet assumes that the system is in a covered metallic box.I personally use 10-20mm or a bit higher for that..
By the way, don't forget to enter loss tangent and metallic losses for associated layer, so you get more realistic response..
 
If I use 30mm for the 'Air' layer, it throws the box resonance warning. Why the same warning disappear for 20mm for 'Air' layer ?

Code:
Run 1:  Sat Jun 22 21:59:35 2019.  SL291418.107194.
        Em version 15.53-Lite (32-bit Windows) on DESKTOP-DG59KC2 local.

  Project:  C:\Users\kevin\Documents\sonnet\15.53-Lite\three-pole microstrip lowpass filter using open-circuited stubs\three-pole microstrip lowpass filter using open-circuited stubs.son.

    Frequency:  0.001 GHZ

      Sonnet Warning EG2680:
      Circuit has potential box resonances.
      Circuit: primary structure
      The estimated box resonance frequencies are listed below.
      The calculations assume lossless materials, with a box filled with
      only the specified dielectric stackup materials.

      6.241 GHZ  TE Mode 1,0,1

      Only lowest order modes considered.

      See the chapter on package resonances in the Sonnet User's Guide.
      Date: Sat Jun 22 21:59:35 2019


      Sonnet Warning EG2680:
      Circuit has potential box resonances.
      Circuit: left box wall SOC magnetic wall standard
      The estimated box resonance frequencies are listed below.
      The calculations assume lossless materials, with a box filled with
      only the specified dielectric stackup materials.

      6.249 GHZ  TE Mode 0,1,1

      Only lowest order modes considered.

      See the chapter on package resonances in the Sonnet User's Guide.
      Date: Sat Jun 22 21:59:35 2019


    Post-Analysis:
      Errors detected: 0    Warnings detected: 2.

Analysis completed Sat Jun 22 21:59:42 2019.



By the way, I do not quite understand how box size and free space cover affect box resonance



There are dielectric tangent loss and magnetic tangent loss. Should I fill in values for both ?

ZtVSdQj.png


Why does the above S11 graph of the low pass microstrip filter has a sharp notch somewhere near 2GHz ?
 
Last edited:

If the Top Cover is Free Space, no problem.You can place any height for that but the Box is small for highest frequency as my understanding.
Make the Box larger by its' free sides because the Box creates a resonance lower than your highest simulation frequency.Either increase the Box size or decrease the highest simulation frequency..
 
the Box is small for highest frequency as my understanding

Huh ? I thought the box is of the same planar direction of the microstrip filter ?

Did I misunderstand the concept of height ?
And what is the actual purpose of box size ?



By the way, I have updated my previous post with S11 graph, but it looks strange with an deep notch near 2GHz, why so ?
 

Look at the Ports that are very close to Box side.The filter should be symmetric as much as possible on X axis.So equal distances must be considered to bottom and top.
Also, top cover height changes some so you should properly adjust the height.
You see strange behavior because this filter has been designed for dielectric coefficient 10.8, not 4.4.If you do the simulation for this coefficient, the filter response is quite good..
curve.png
 
Last edited:

top cover height changes some so you should properly adjust the height.

Height changes ? Huh ? Why so ?



this filter has been designed for dielectric coefficient 10.8, not 4.4.If you do the simulation for this coefficient, the filter response is quite good..

No, using coefficient of 10.8 also gives a sharp notch on S11 curve.

lIbpyco.png
 

OK, I have repeated the simulation with original datas and scaled the axis in according within the paper to compare.Looks good..
The notch on S11 is not an issue.
 

ok, now I wish to re-design the microstrip low pass filter at lower cutoff frequency at 100MHz, so that I could fabricate the PCB and do the actual PCB testing.

Any advice ? The design equations are so complicated.
 

Any advice ? The design equations are so complicated.

Have a look at Nuhetz Filter Solutions synthesis, it's accurate and has a link to Sonnet.
And get a large PCB if you're desiging for 100MHz. 8-O
 

ok, now I wish to re-design the microstrip low pass filter at lower cutoff frequency at 100MHz, so that I could fabricate the PCB and do the actual PCB testing.
Any advice ? The design equations are so complicated.

The Filter for 100 MHz cut-off frequency will be very large on FR4 substrate.But however it's feasible if you have sufficient space..
LC Filter will be more practical with better attenuation and insertion loss..
 
How do I tell if the textbook example is using T configuration or Π configuration ?

yMgSq9N.png
 

You mean the Sonnet example shown above? That has narrow series lines (inductive) with an open ended line (capacitive) in the middle. So it is T configuration.
 
For the two ports (which are allocated two rectangle patches), I know I can use PCB calculator, but how do I decide the input and output impedance of the microstrip transmission line ?

Is VNA always 50 ohm ?

So, do I need any impedance transformer microstrip in this low-pass filter design ?
 

You design the filter for a given source and load impedance, here: 50 Ohm. No additional impedance transformers needed.

Input and output lines are designed for those 50 Ohm, which is the standard in instruments, lines and connectors. Only with that same line impedance, we can use any length of line without having to worry about impedance transformation effects. For a 50 Ohm line connected to a 50 Ohm device, we always have 50 Ohm input imepdance, regardless of the line length.
 

You design the filter for a given source and load impedance, here: 50 Ohm. No additional impedance transformers needed.

I know, but do you notice the two extra rectangles (which do not have any designated length and width dimensions) nearest to the ports ?

n422Rvv.png
 

I know, but do you notice the two extra rectangles (which do not have any designated length and width dimensions) nearest to the ports ?

Sure, those are the 50 Ohm feed lines. Very typical. No length because they are 50 Ohm and length doesn't matter.
 

    V

    Points: 2
    Helpful Answer Positive Rating
Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top