I have a SMA connector, that I use in lab for connecting the sensor I build to the wave generator and spectrum analyzer. The result that I gave was very different with the HFSS simulation result, in the simulation the voltage that I measured in output was around 2 µV, but in the lab it was 8.5 mV. I want to that is the spectrum analyzer effect my result or not? because in the simulation I draw a rectangular in input and assigned lumped port. now I want design the exact SMA connector in the HFSS , and again check the output voltage.
I attach the model that I have in HFSS, and also the model of SMA connector that I have in lab.
epsilon is different, gaps are different. Did you ever assemble a semi-rigid cable? Depending on the construction, the reflection changes considerably whenever you make a small (+-0.25 mm cut) mistake, or overheat the dielectric, so the foam melts and changes density (and effective epsilon).
So, forget recreating the connector. Design just some more or less typical connector and tweak its dielectric epsilon until it matches the impedance.
... and I don't think the problem is in the connector.
Dear Najme,
The 3D model of the SMA Connector in HFSS is attached here. Hopefully, that will help you.
Ahmed S. I. Amar
Dear Ahmed
Thank you for your answer. Unfortunately I cannot import your model in HFSS, I do not know why?!
another thing is that I designed a SMA connector but it is not the same as the one I am using in Lab, (it is in datasheet that I attached). I want to know that is anyone designed this type of SMA connector before?
Andrey Porokhnyuk
Thank you for your response. I did not get what you mean exactly. would you please explain more? you mean in HFSS the simulation and the way I assigned excitation is correct, but in practical experiment I should use semi-rigid cable instead of SMA connector?
No. I mean it is pointless recreating some specific connector. Make just some simple coaxial junction with the metal exterior resembling your connector. Do not recreate the interior. Then tweak the epsilon until the impedance is matched. After that, attach your connector to your circuit.
I think that the device you have modelled operates at a low frequency where the exact shape of the SMA connector will not affect the results at all, and it can just be modelled as a short section of coaxial cable with the correct dielectric and inner and outer conductor diameters.
- You can find SMA connectors models on the internet and AS FAR AS I KNOW, there is this another notice that you are supposed to pay attention is that almost probably SMA connector models is not the weakness of your simulation.
- Oh and I remember when I was designing the excitation plane as lumped port in microstrip structures I chose the size of the rectangular lumped port is about [3width of the microstrip line * 3height of substrate].
- Although your question is about for 10 days ago and I hope you have found the solution of this question!
Saeid Mohammadpour Jaghargh thank you for your answer. Actually I could not find the model of my SMA connector, there is another model on the internet.
As the result of my practical experiment is very different with simulation, I want to draw everything in HFSS, like what we have in real.
Malcolm White thank you for your answer. Just two question is that when I draw the coaxial cable, how should I connect it to the input of my model as excitation port? another question is that, can I assign lumped port excitation to coaxial cable? If yes, what should be the reference?
I don't know if you can excite a coaxial cable with a lumped port in HFSS as I do not use HFSS. I expect you can. At high frequencies it is better not to because it is not usually a good match to the fields. In CST a waveguide port is a much better option. It may not matter at your frequency.
To connect to your model, remove the short vertical bar where you have the lumped excitation (above the earth symbol on the left of the picture), and connect the earth (outer) of the cable to the earth side (bottom), and the centre to the other side (top).
Dear @Najme Mohammadyahya
If you'd like to assign lumped port excitation in HFSS when using a coaxial cable (excitation), the outer conductor of the coax should be connected to the ground of the system model and can be used as reference whereas the inner conductor should be connected to the main conducting part of your system model.
How to design coaxial cable in HFSS as lump port?
See this link at 14:00 minute streaming! it shows you how to design it step by step.
https://www.youtube.com/watch?v=Me3mzL-sxO4
Dear Najme Mohammadyahya
If your design is really sensitive and you insist on simulating everything you need to get an accurate data sheet of that SMA. If you don't have some dimensions you have to measure it. And remember, after drawing the scheme and simulation, you should get the S-parameters of that SMA as an initial result. Then compare the simulation results with measurement ones using Network Analyzer.
Pierre Moukala Mpele Saeid Mohammadpour Jaghargh Malcolm White
Thank you very much for your answers.
- Badges
- Science topic
- Similar topics
- Phytochemistry
- Extracts