Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.
Somehow radiating more RF power than being put into lumped port?
Posted 9 gen 2022, 22:12 GMT-5 RF & Microwave Engineering Version 5.6 4 Replies
Please login with a confirmed email address before reporting spam
Hello,
I am trying to simulate an RF antenna that excites a tip-sample junction in an STM (Scanning Tunneling Microscope). Before I get to that, I want to have a proper understanding of COMSOL. I feel like I am still lacking that.
I modeled a coaxial cable that is stripped at the end, with a metal object nearby. I've got the feeling that the standing waves inside the antenna kind of make sense. What does not make sense yet is the conservation of energy.
I have my model do several different frequencies (same order of magnitude), and several things seem off to me:
[1] The lumped port power (emw.Pport_1) changes significantly from 0.363W @ 1.50GHz to 0.721W @ 1.95GHz. (stripped part is 10cm, so 1.50GHz)
[2] The power outflow (emw.nPoav) changes significantly from -0.212W @ 1.05GHz, to 0.862W @ 1.50GHz to 11.955 W @ 1.95GHz.
[3] A second way to calculate the power outflow (emw.Poavx X nx+emw.Poavy X ny+ emw.Poavz X nz) (where X is multiplication, but this interface makes it italics if I use the asterisk) gives similar results, but not so similar it seems like a numerical issue (e.g. 10.09W @ 1.95GHz). This is integrated over the scattering boundary condition (outer boundary of PML),
I do find resistive losses (emw.Qrh integrated over the objects) to be fairly consistent ranging from 2.05E-4 W @ 1.95GHz to 2.88E-4W @ 1.50GHz.
I've attached the file. Hopefully someone (Robert?) can help me :).
Attachments:
Please login with a confirmed email address before reporting spam
Caveat: I've only had time to take a brief look at your model and I haven't attempted to execute it or examine the output. That said, I've noticed that: (1) your meshing is rather coarse (especially in regard to the coax line and other pointy objects; mesh the coax specifically so you have at least 7 segments ringing around the surface of the center conductor - the default of 4 is not sufficient) and (2) that the surfaces in your "Surface Integration 3" differ from "Surface Integration 4" under Derived values. That said, some of your calculations also don't immediately make sense to me. (For example, why would you integrate emw.EIRP over a surface? Note the resulting units!). And wIth frequencies in GHz, iron's magnetic properties are irrelevant. I may take another look at this later when I have more time. Hope that helps to give you some things to think about, for now. Added: You are also using a "lumped port." Personally, for RF, I would use a "port" instead, set it to "Coaxial," then arrange that port excitation face to be an exterior boundary. But that's just my style, which I use because I know it works properly. I'm not sure how doing it differently could impact the power calculation.
-------------------Scientific Applications & Research Associates (SARA) Inc.
www.comsol.com/partners-consultants/certified-consultants/sara
Please login with a confirmed email address before reporting spam
Hello Robbie, To augment the points that Robert is making, see: https://www.comsol.com/model/dipole-antenna-with-a-quarter-wave-coaxial-balun-12313 This shows a similar case, of how you would model the excitation and the metal parts of an antenna.
Please login with a confirmed email address before reporting spam
Hello Robbie, To augment the points that Robert is making, see: https://www.comsol.com/model/dipole-antenna-with-a-quarter-wave-coaxial-balun-12313 This shows a similar case, of how you would model the excitation and the metal parts of an antenna.
Hello Walter,
Thank you very much. I'm now playing with that example, it's a nice one. I'm removing pieces of it to get to the coaxial antenna.
One thing that I don't understand is that when I tried to evaluate emw.EIRP (to compare it with emw.Pport1 and see if this time it does work) I get an error:
This happens already at the base file (the one you can download through the Application Library)... Is this normal? Is it due to some step along the way?
Attachments:
Please login with a confirmed email address before reporting spam
EIRP is also not a global variable but a spatially dependent variable. You can only evaluate it at spatial coordinates. EIRP is the amount of power you'd have to transmit with an isotropic antenna to get the same signal strength in that far field direction as your current antenna. So if you have at some theta/phi coordinate an EIRP of 10W while you only transmit 1W you have a total antenna gain of 10 times at that direction.