Non 50-Ohm Antenna design

[AlexSIC]

Hi everyone,

I'm trying to match an antenna to Zl = 360 - j2600Ohms at 2.4GHz. In that case matching a typical 50Ohms antenna to Zl is extremely hard because of the large negative reactance (Ceq = 25fF at 2.4GHz).

I'm considering designing a non 50-Ohm to ease the matching but not so familiar with antenna design. I've tried some structures (circular loop; patch)but non of them can provide a high negative reactance.

Have anyone tried this kind of antenna design? I would appreciate any thoughts and advice.

Thanks!

 

[biff44]

have you heard of a thing called Fano's limit. this tells you the theoretical maximum bandwidth you can get for a passive matching network. i bet the calculated bandwidth is going to be small, so your final antenna will be easily mistuned.

--- Updated May 24, 2021 ---

is there some way to modify the antenna driving structure so it is more tightly coupled to the outside world?
is this a receive or transmit antenna?

 

[AlexSIC]

have you heard of a thing called Fano's limit. this tells you the theoretical maximum bandwidth you can get for a passive matching network. i bet the calculated bandwidth is going to be small, so your final antenna will be easily mistuned.

--- Updated May 24, 2021 ---

is there some way to modify the antenna driving structure so it is more tightly coupled to the outside world?
is this a receive or transmit antenna?

I've heard of Fano Limit, but it seems to me this Zl doesn't have a extremely high Q. I just want to cover 2.4G ISM band.
This antenna is supposed to be driven by a switching mode PA. I suppose we can bondwire the chip on PCB very close to the MN(if needed) and antenna and properly account for the impedance finally in ADS momentum. The major problem I have is how to design a antenna with high negative reactance.

 

[dick_freebird]

Perhaps an impedance-transform rather than a
mismatched-to-space antenna? That's a common
thing in Class E PAs I gather from listening to other
peoples' design chatter.

I'd bet there's a wealth of info on "high band" cell
phone and WLAN Class E PAs, how they build and
tune them. Looked tricky is all I can say. So find the
tricks?

 

[AlexSIC]

Perhaps an impedance-transform rather than a
mismatched-to-space antenna? That's a common
thing in Class E PAs I gather from listening to other
peoples' design chatter.

I'd bet there's a wealth of info on "high band" cell
phone and WLAN Class E PAs, how they build and
tune them. Looked tricky is all I can say. So find the
tricks?

Thanks for your thoughts. I've found some papers about codesign of PA and antenna, but they're high output power (~20dBm), which requires a small load impedance.
I'm trying to design a ultra low power PA, which is the opposite requiring a high impedance..

 

[FvM]

Either if you match the reactive PA impedance with an antenna of conjugate impedance or by a lumped or distributed matching network, you get high Q and small matching bandwidth according to Fano rule. Presuming you are working in a real environment, e.g. wearable transmitter, it will hardly work without adaptive tuning network.

 

[AlexSIC]

Either if you match the reactive PA impedance with an antenna of conjugate impedance or by a lumped or distributed matching network, you get high Q and small matching bandwidth according to Fano rule. Presuming you are working in a real environment, e.g. wearable transmitter, it will hardly work without adaptive tuning network.

For the switching mode PA, I don't think conjugate match is a must, the required load impedance is determined by the output power and Vdd. This system is working in 2.4G ISM band, I don't know if this network Q is already too high. Thanks for your reply!

 

[FvM]

For the switching mode PA, I don't think conjugate match is a must, the required load impedance is determined by the output power and Vdd.

How do you determine the reactive impedance told in post #1 then? Do you say the specification is erroneous?

 

[BigBoss]

Forget about reactive parts of the Impedance, match simply real-to-real by \[ \frac{\lambda}{4} \] Transmission Line.

 

[Azulykit]

50 Ohms is convenient in a typical coaxial transmission line based system. If I was not using "typical" coax I would not be constrained to a match to 50 Ohms and could consider other antenna input impedances. Your question does not provide enough of a picture to give a really helpful response.

 

[mtwieg]

For the switching mode PA, I don't think conjugate match is a must, the required load impedance is determined by the output power and Vdd.

Conjugate matching is never a must, even with linear amplifiers. With switchmode amplifiers, output impedance is ill-defined though.

How do you determine the reactive impedance told in post #1 then? Do you say the specification is erroneous?

Yeah this is the real question.