- This topic has 27 replies, 2 voices, and was last updated 1 week, 1 day ago by
.
-
Posts
-
And doing the full dipole via 1:1 transformer, catastrophe…
I am convinced that the matching network impedances were calculated correctly (verified the method against values in the paper), and halving the dipole looks fine as per previous post…
This data looks more believable. Have you looked at S11 on a finer scale: separate scale or separate graph?
The network was designed to provide a 180-degree phase difference, NOT constant +90 and -90 degrees required when you ground the middle of the dipole. Floating the dipole with the transformer fixed the problem. The minor common-mode effects of using a physical dipole without a physical transformer should not pose a problem if the network is mounted near the dipole.
It is usually good to employ a balun at the feed point of a dipole when using an unbalanced (coaxial) transmission line. With the network, you would also need a balun between the network and any coax line to the dipole. Mini-Circuits sells baluns and isolation transformers for low power (< 1 watt).
I didn’t quite say that right. I think the problem was that the two outputs from the network are not equal and opposite voltages, so you cannot ground the center of a two-part dipole model. It is the difference between the two network outputs that is important. Difference voltage, which is highest when the phases are 180 degrees apart.
We are talking about the setup in #8943, yes ?
I don’t get it, S11 looks terrible… I guess I should then disconnect GND between P2 and P3 and call it good 180 degrees difference ? If not, the phase diff also looks terrible to me.
That is a separate problem. I question the design of the network. The inductor values look very small for 315 MHz. S11 was terrible in my simulation with the dipole equivalent circuit also.
Yes. 8943. The phasing network is a two stage L-network in each path, stepping the impedance down toward an antenna that is greater than 50 ohms. Seems backwards. What you really care about is minimum S11 at port 1 and maximum S21 if the antenna is port 2. I would replace S1P with port 2. Delete 8943 ports 2 and 3, and adjust TR1 ratio so that port 2 reflects as the antenna real impedance at resonance. Best S11 and S21 should occur when the phasing/matching network is optimum.
A simple balun or isolation transformer would give perfect 180-degree outputs over a broad bandwidth. More complex networks are sometimes used to get 90-degree splits for single-sideband mixers or circular polarization over limited bandwidths. At 315 MHz, a small quadrature hybrid coupler would give 90-degree split outputs.
Err, resonance ? No sir!
The two 47.9 nH were there to “resonate” the small antenna: 3.9 – j183.6 Ω @315 MHz
The network is supposed to actually be ultra wide band
Besides, I was able to get a more decent S11 by halving the dipole: #8928
Thank you!
The network is supposed to actually be ultra wide band
Besides, I was able to get a more decent S11 by halving the dipole: #8928
Thank you!
- You must be logged in to reply to this topic.