Electronics > RF, Microwave, Ham Radio

Is antenna matching important for receiver with LNA?

(1/2) > >>

Mad ID:
Hi,
I'm trying to improve the performance of my GPS receiver based on Quectel L70-R. I have included an external LNA as the datasheet recommends to boost sensitivity by 3dB.

Since my board is really small I have metal parts near the antenna which detune it. A matching network is present on the PCB to match the antenna.

My S11 was around -2dB at the GPS frequency, and after matching its now -10dB. The problem is that when measuring satelite CN0 (carrier-to-noise) with two receiver simultaneously (with and without matching) I don't see any conclusive performance benefits.

I have found in the u-blox application note that LNA after the antenna makes the matching loss less significant.

An LNA placed very close to the antenna can help to relieve the matching requirements. If the interconnect
length between antenna and LNA is much shorter than the wavelength (9.5 cm on FR-4), the matching losses
become less important. Under these conditions the matching of the input to the LNA becomes more important.
Within a reasonable mismatch range, integrated LNAs can show a gain decrease in the order of a few dBs versus
an increase of noise figure in the order of several tenths of a dB. If your application requires a very small
antenna, an LNA can help to match the impedance of the antenna to a 50 Ohms cable. This effect is indeed
beneficial if the antenna cable between the antenna and the receiver is only short. In this case, there’s no need
for the gain of the LNA to exceed 10-15 dB. In this environment the sole purpose of the LNA is to provide
impedance matching and not signal amplification.

My question is, does this mean that I don't have to bother with tuning the antenna at all? What the theory behind it? Can you give me an explanation or point me to some literature?

Thank you very much.

Earendil:
I'm sorry, but I really don't understand how you can talk about S parameters and impedance matching and then ask a question like this.
The quoted appnote explains it quite well. There are two pieces of background information you should probably know:
- At 1/10 of wavelength or less you don't  need to worry about transmission line effects.
- Amplifiers provide large reverse isolation.
You can find these basically in any of your electronics textbooks.
Obviously this is only enough for an intuitive understanding of what's going on. For properly quantifying the effect you need to do EM simulations or make some measurements. I mean if you can measure the S11 of your antenna then why you can't measure LNA+antenna?

hamdi.tn:

--- Quote ---I'm sorry, but I really don't understand how you can talk about S parameters and impedance matching and then ask a question like this.

--- End quote ---

sorry for the attitude, he's just asking a question for the things he don't understand.


--- Quote ---You can find these basically in any of your electronics textbooks.

--- End quote ---

so everything else ... suddenly no need for this forum anymore  :-DD

KJDS:
If you match an RF FET, or pHEMT for best match then you'll usually miss out on best noise figure. Most RF transistors S parameters also include noise parameters and usually a compromise between best noise figure and best match will be used. If you can get the antenna within a fraction of a wavelength all it means is that you've got less rotation round the Smith chart to account for in the matching. The antenna won't be a perfect 50 ohms and you don't want the transistor to see that either.

Earendil:

--- Quote from: hamdi.tn on May 28, 2016, 01:27:31 pm ---
--- Quote ---I'm sorry, but I really don't understand how you can talk about S parameters and impedance matching and then ask a question like this.

--- End quote ---

sorry for the attitude, he's just asking a question for the things he don't understand.


--- Quote ---You can find these basically in any of your electronics textbooks.

--- End quote ---

so everything else ... suddenly no need for this forum anymore  :-DD

--- End quote ---

Yeah, I'm sorry. But the question really seemed weird to me.

Navigation

[0] Message Index

[#] Next page