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

[imtiaz369]

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?

 

[dick_freebird]

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).

 

[BigBoss]

How GND was connected ?? There is a missing GND connection thought.

 

[imtiaz369]

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).

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...

 

[volker@muehlhaus]

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.

 

[PlanarMetamaterials]

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.

 

[imtiaz369]

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??

 

[imtiaz369]

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.

Here is the simulation file

 

[FvM]

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.

 

[imtiaz369]

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.

But I do not find any issues in my design?

 

[FvM]

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?

 

[volker@muehlhaus]

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.

 

[imtiaz369]

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?

I hope that views help you understand my situation.

--- Updated Dec 12, 2023 ---

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.

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?

 

[D.A.(Tony)Stewart]

It looks as expected now, a 2.5 cm stub with 3 GHz recursive nulls.
What was your question again with complete details of design?

 

[imtiaz369]

It looks as expected now, a 2.5 cm stub with 3 GHz recursive nulls.
What was your question again with complete details of design?


The S11 should be like that, because I get this result when used a hfss port, - so why it is behaving wrong when I just connect this SMA? It should be nearly like this, isn't it?

 

[volker@muehlhaus]

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.

 

[D.A.(Tony)Stewart]

Looks like s1 plots( reflected signal) for f/2 = 6MHz where f/4 = 3 MHz = 2.5 cm trace length and 3f/4 = 7.5 Mhz
It does not look like s11.

 

[volker@muehlhaus]

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.

... or you could also terminate the line with an absorbing PML boundary. Make sure you have no gap between PCB and PML boundary.

 

[D.A.(Tony)Stewart]

From what I read, Antenna plots always need PML boundaries.

 

[volker@muehlhaus]

From what I read, Antenna plots always need PML boundaries.

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.