Is impedance matching important if the components are in low distance to each other?

[Dr. von Rosenstein]

Hi colleagues,

simple question:
Is impedance matching important if the components are in low distance to each other?

What I mean is: lets say I want to design a RF transceiver as an integrated circuit. Should the LO and the mixer are impedance matched? In the layout they are very close to each other. The distance is much smaller than the wavelength. Therefore I think, impedance matching can be ignored.

I can not explain it with words (maybe because of my bad english), but in my imagination impedance matching is only important if the distance is greater than a certain fraction of the wavelength. Is this right? Can anyone verify this statement with an explanation?

Thank you for all help!
Rosenstein

 

[PlanarMetamaterials]

One of the critical metrics for impedance matching isn't necessarily the impedance of a certain device, but rather, the input impedance seen after the output of a certain device.

For electrically short TLs, this input impedance approaches the termination impedance, meaning that the impedance of the TL matters less and less as the line becomes electrically shorter.

In your example, so long as the input impedance into the mixer as seen from the LO is close to the expect value, the power will make it into the mixer instead of being reflected.

 

[FvM]

Impedance matching is relevant in two regards, avoiding reflections on transmission lines and maximum power transfer.

As a rule of thumb, we can consider a transmission line as lumped component, if the length is below 1/10 wavelength. Related to impedance matching, you can conclude that you can either ignore a connecting transmission line, or model it as series L respectively shunt C if the characteristic impedance is considerably different from load or source impedance.

The other aspect of impedance matching is maximum power transmission, or more generally, terminate a source or load with the designed impedance (which may also involve an intentional mismatching).

The second point still matters on short distance.

 

[BigBoss]

Impedance Matching is not a fact of "necessity" in electrically small circuits, Impedance Matching is a must for max. power transfer as explained before.( There is no presumption in max. power transfer theorem )
But there is no enough room to put them , therefore it's compulsorily "omitted"

 

[Dr. von Rosenstein]

Thank you for your answers!

Let me reword my thoughts. It is hard for me to find the right words to describe what I mean due to my bad english (sorry).

Consider: I still talk about integrated circuits! The components are in direct neighborhood. No long TLs are needed.

I can simulate the output impedance of the LO and the input impedance of the mixer. Now one could have the idea to put an impedance matching network between the LO and the mixer. But I haven’t seen a publication where somebody did this. I always thought, this is due to the close distance and that in this case impedance matching is not necessary.

Other thought: If impedance matching is useful even at close distances, then it should be also considered within a component. For example between two connected transistors in every RF component (LO, mixer, LNA ...). But this I haven’t seen also yet.

Thank you for all help!
Rosenstein

 

[BMR]

Other thought: If impedance matching is useful even at close distances, then it should be also considered within a component. For example between two connected transistors in every RF component (LO, mixer, LNA ...). But this I haven’t seen also yet.

Thank you for all help!
Rosenstein

Impedance matching concepts comes from the analog world and it has been there forever. Perhaps the attachment would help.

 

[mtwieg]

I can simulate the output impedance of the LO and the input impedance of the mixer. Now one could have the idea to put an impedance matching network between the LO and the mixer. But I haven’t seen a publication where somebody did this. I always thought, this is due to the close distance and that in this case impedance matching is not necessary.

Other thought: If impedance matching is useful even at close distances, then it should be also considered within a component. For example between two connected transistors in every RF component (LO, mixer, LNA ...). But this I haven’t seen also yet.

Consider a connection between a source with an output impedance of 1ohm and a load with impedance of 100ohm. If I connect them directly (no transmission lines or parasitics at all) then I will get very poor power transfer. On the plus side I will get very broadband performance, which may be nice. For digital circuits, this is great, job is done. For many analog circuits, especially where noise is a concern, this is terrible. It really depends on what want to get out of it.

 

[FvM]

Other thought: If impedance matching is useful even at close distances, then it should be also considered within a component. For example between two connected transistors in every RF component (LO, mixer, LNA ...). But this I haven’t seen also yet.

Impedance matching is implemented to serve a purpose, as explained either maximum power transfer or transmission line reflection avoidance, in some cases both. Maximum power transfer can be an objective for inter- and intra chip connections in some cases, e.g. power amplifiers. If so, you'll see it there. In most cases it's not, or other concurrent objectives like noise minimization are priority.

Putting the question too general like "do we need impedance matching in RF design" is pointless. Better ask about the purpose of impedance matching and where it's appropriate.

 

[biff44]

three things happen if there is a poor impedance match:
1) amplitude and phase ripple if there is a long distance between the two components, which you do NOT have to wrorry about
2) high gain...obvously the gain of a device is dependent on what system impedance you measured it in. A mixer with 7 dB conversion loss in a 50 ohm system might have say 12 dB of loss in a 20 ohm system...for instance.
3) there CAN be instabilities due to a poor impedance, such as a compnent breaking into oscillations, or added spurios signals that normally would be too small to worry about

 

[vfone]

Is impedance matching important if the components are in low distance to each other?

Depends who you ask.
If you ask an analog mixed/signal engineer who use only voltages, he will give you an answer, and if you ask an RF engineer who is using power levels, he will give you a different answer. Is not hard to guess what answers they pick.