Stats

  • Locked Locked
  • Replies 8
  • Subscribers 64
  • Views 15099
  • Members are here 0
This discussion has been locked.
You can no longer post new replies to this discussion. If you have a question you can start a new discussion

Can you simulate an EM layout extraction with lumped elements in the middle of the layout with Microwave Office?

Zeke KJ7NLL

Hello,

We have a layout with layout cells for inductors and capacitors to impedance match a Qorvo LNA (they provde an S2P).

Each capacitor/inductor S2P file from muRata straddles each layout cell so we can see how EM coupling affects the match.

We would like to get the combined behavior of EM+S2P. The extraction shows ports at each component location where the iNets connect.

When we used AXIEM to simulate the EM, it seems to only simulate the metal, but ignores the S2P components that are connected to on the circuit schematic. 

Is there a way to get AXIEM (or another EM simulator) to use the S2P components at the same time that EM simulation takes place so we can see how layout coupling affects the impedance match?

Please see the attached .emz archive:

Uh, oh.  It is bigger than 1MB.  Here are some pictures, maybe you can help!

  • The schematic "LNA test" is where the extraction takes takes place, and the graphs and measurements reference this schematic as well.
  • The schematic "LNA 144MHz Component" schematic has the LNA layout.  (Note that subcircuits like "S2P_CAP_0402" combine the footprint with S2P connectivity so we can switch between EM and not-EM by toggling the EXTRACT block.  The layout cell footprints (like 0402) are applied to dummy components inside the S2P_CAP_0402 subcircuit to help with that.)
  • The extraction "EM_Extract_Doc" shows the extraction and ports at every component pad.
  • The graph "LNA" shows S11 and S21:
    • With the EXTRACT block in the "LNA test" schematic disabled, the graph shows a good S11/21. 
    • With the EXTRACT block enabled, S11/21 are horrible, presumably because AXIEM is not considering the S2P components and only simulates metal.  See the "Show Currents" output, all current is on the Port 1 MLIN, nothing gets past, everything reflects.
      • (press F8, EM takes ~1 minute to simulate)

See this video we made about this project for ideal optimization, what we are trying to do now is simulate a circuit board with real components to have it made:

https://www.youtube.com/watch?v=cMte9q1hoHM

Also, is there a way to "show currents" in the 2D extraction view?

Schematic:

Layout:

Extraction:

-Zeke & Eric

  • Here is a link to the .emz because it was to big to upload to the forum:

    drive.google.com/.../view

  • Hi Zeke,

    This looks like a really cool project. I do not have the libraries that you are using for the EM+s2p models so I cannot simulate the project you provided.

    Here is a video that might be helpful to watch on extraction of a simple microstrip structure: https://kb.awr.com/display/awrfaq/Extraction+Video+Tutorial . This video uses the built-in example named Extract_Example.emp .

    There is also an example that shows extraction used in a large amplifier simulation with an extraction block and lumped elements. The example name is called RF_PCB_AMP1900.emp.

    Give those a try. When you use the EM+s2p models you have, try enabling the MLIN's for extraction and disabling the EM+s2p models for extraction. This is what I had to do when I tried testing a simple library model with extraction. Also remember to make sure that both the linear 2D layout and the 2D EM layout match after extraction.

  • in reply to David Webster

    Hi David! Thank you for your suggestion!

    We looked at the video and RF_PCB_AMP1900.emp, which looks pretty close to what we are trying to do except the example is using TLINs and we are using iNets in the extraction. 

    This one should contain the s2p files, can you see if it will open on your system?

    drive.google.com/.../view

    Step 1: Press F8 and check "Graph 1" that has 2 points on it.  It should be S21 =~ 28dB and S11 =~ -38dB

    Step 2: On the "LNA test" schematic, enable the EXTRACT block.

    Step 3: Press F8 again and let it run 2-3 minutes. Shouldn't the graphs look similar, but with EM calculated?  You can see that the dB is way different from the s2p component model. 

    We have stared at this for hours trying different things and its not there yet. Anything you can do to help would be greatly appreciated!

    -Eric and Zeke

  • in reply to Zeke KJ7NLL

    Here is the corrected link, the first upload was 0 bytes for some reason: drive.google.com/.../view

  • in reply to David Webster

    HI David,

    It looks like the EM simulation alaways excites port 1 on the extraction, but port 1 on the extraction never seems to be port 1 on the schematic. Is this a bug, or is there a way to force the extraction port 1 to be the same as the schematic port 1?

    We tried Analyst to see if it behaves differently.  Here are the current/fields.  You can see Port 1 in the EM at the bottom just above port-12.  The excitation should be at Port 34 in the top-left.

    We already tried the "PORT1" signal source element instead of the normal "PORT" element, but that did not change anything.

    Here is the same view with currents off:

    Ideas?

  • in reply to Zeke KJ7NLL

    Hey Zeke,

    I was able to get your project to run and I found some issues that need to be addressed. I would also stress trying to get individual pieces to work rather than trying to do it all in one shot. I would start with a simple two layer stackup, get extraction to work and any other advanced features individually, and then start adding complexity by adding the advanced features together.

    1. Look at the 3D view for the EM Structure of your LNA design. It looks like the scaling is off. I was able to fix this by going into the substrate information and modifying the scaling. This issue is shown in Figure 1.

    2. There is some floating copper (not connected to ground) in the extracted 3D view. I provided a picture where teal is the ground color and the floating copper is in purple and orange. This issue is shown in Figure 2.

    3. On the LNA schematic page, you need a HB port excitation. This issue is shown in Figure 3. Note that you need to set an appropriate input power level based on the device datasheet to get the desired results.

    I also put together an example project that shows Extraction Hierarchy being used with an EM+.s2p model. Do you have an email I can send it to?

    Figure 1:

    Figure 2:

    Figure 3:

  • in reply to David Webster

    HI David,

    Thank you for looking in to the extraction issue!

    My email is: kj7nll@gmail.com

  • in reply to Zeke KJ7NLL
    Thank you David, it is really fun to learn about Microwave Office! I tried turning on currents and the currents are only showing on port 1, the other metal is cold, even though the metal segments are connected with s2p files so the current should show on the other segments too, shouldn't it?

Community Guidelines

The Cadence Design Communities support Cadence users and technologists interacting to exchange ideas, news, technical information, and best practices to solve problems and get the most from Cadence technology. The community is open to everyone, and to provide the most value, we require participants to follow our Community Guidelines that facilitate a quality exchange of ideas and information. By accessing, contributing, using or downloading any materials from the site, you agree to be bound by the full Community Guidelines.