SMA Connector - Not work - Is there any way to solve it?

imtiaz369

Member level 2
Joined
Jul 8, 2021
Messages
45
Helped
0
Reputation
0
Reaction score
0
Trophy points
6
Activity points
484
When I used an SMA connector instead of a hfss port (lumped/wave), it returned an incorrect s11 (see attached image). Everything is fine because the pin is properly connected to the antenna's feed line.


But why it is showing this type of behavior?

Are there any additional things need to do when I try with a SMA Connector model?




 

Sort by date Sort by votes
And you know the "wrong" result is "wrong", how? Measurement?
Or just don't like the result?

SMA connectors do have their own parasitics, that's why there have
been generations of improvement since (for higher and higher bands).
 

Upvote 0 Downvote
dick_freebird said:
And you know the "wrong" result is "wrong", how? Measurement?
Or just don't like the result?

SMA connectors do have their own parasitics, that's why there have
been generations of improvement since (for higher and higher bands).
Click to expand...
Yes, it was wrong...this is s11, when used port or also in measurement it was below -10dB, but here, you see, it just starting from 0 and...
 

Upvote 0 Downvote
Your results show S11=0dB at DC, which is short or open. You can easily look at S11 in Smith chart to see if you have an OPEN at DC, or possibly a short circuit somewhere.

As BigBoss wrote, the question is how your GND is connected to the PCB. It seems that the HFSS connector model is designed for more substrate thickness than you used here, and the body ground pin doesn't touch your PCB backside.

 

Upvote 0 Downvote
You haven't shown us details on the connection, or provided any information on what kind of mode you are trying to excite. Depending on the SMA connector, there may be arms on the upper layer. These may be connected to conductors on the upper layer, if - for example - you are trying to excite a CPW. Please show us a view from multiple angles so that we can see what you are trying to do. Or post a 3D model.
 

Upvote 0 Downvote
Hi, I designed an SMA model following a video, but overall, it is not working at all. See the S11, fully reflecting. I've attached the hfss file here, anyone please help to figure out the issue??

 

Attachments

  • sma.zip
    69.4 KB · Views: 96

Upvote 0 Downvote
Here is the simulation file
 

Attachments

  • sma.zip
    69.4 KB · Views: 101

Upvote 0 Downvote
S11 magnitude plot shows that your circuit is either shorted or open. I don't have HFSS at hand right now to review your design but apparently you are expecting to see a matched load through the SMA connector.

I can just guess why the measurement doesn't show the load. Either there's no load at all or it's not correctly connected.
 

Upvote 0 Downvote
But I do not find any issues in my design?
 

Upvote 0 Downvote
Unfortunately you didn't yet answer questions in post #3 and post #5, e.g. is it open or short. As long as nobody is motivated to check the HFSS setup, we won't know.

According to post#7 plot showing something like S11 and S21 magnitude, it's a two port setup. Where are the ports located? Can you show a more informatitive view, e.g. isometric perspective?
 

Upvote 0 Downvote
I totally agree with FvM, you need to answer these questions.

I have an idea what the mistake is (ground connection missing, turning your signal line into a wire antenna that resonates at 3GHz), but you can easily check connectivity yourself. The issue exists from DC to GHz, so it will be easy to identify once you start checking step by step.
 

Upvote 0 Downvote

I hope that views help you understand my situation.






--- Updated ---

The ground is now perfectly attached, but the results remain the same. I tried several different designs, such as an antenna and a transmission line, and in all cases, the results remain the same. Do you have any ideas or suggestions on how I can determine which part is problematic or causing this error?
 

Upvote 0 Downvote


 

Upvote 0 Downvote
Your model shows a 50 Ohm port with an unterminated (!!!!!) line, and that is what we also see in results.
It seems that you forgot to add another port at the end of the line.
 

Upvote 0 Downvote
volker@muehlhaus said:
Your model shows a 50 Ohm port with an unterminated (!!!!!) line, and that is what we also see in results.
It seems that you forgot to add another port at the end of the line.
Click to expand...
... or you could also terminate the line with an absorbing PML boundary. Make sure you have no gap between PCB and PML boundary.
 

Upvote 0 Downvote
D.A.(Tony)Stewart said:
From what I read, Antenna plots always need PML boundaries.
Click to expand...
That is a different case: for antennas we place absorbing PML boundaries at some distance, with airbox between antenna and absorbing boundary.

What I meant here is a method to make sure that all signals at the end of the line are absorbed, regardless of line impedance. This can be achieved by running the line straight into a PML wall (no distance!). One situation where this is used in simulation is "measurement" of line impedance, input impedance at the port is then equal to line impedance because there is never any reflection from the PML side.

But of course, the "normal" testcase for a line is to place ports at both ends.
 

Upvote 0 Downvote
You must log in or register to reply here.

Similar threads

Menu
Log in
Register
Cookies are required to use this site. You must accept them to continue using the site. Learn more…