Receive Antenna & Impedance "Matching"

[Mike2459]

Reprinted from Discussion in 'Ham Radio Discussions' started by KL7AJ, Nov 30, 2009.
https://forums.qrz.com/index.php?thread ... ce.227442/

Technical topics on QRZ seem to cycle through periodically. The topic of the week seems to be impedance matching of an antenna system to a receiver. Here's a little primer on the matter. When I worked in ionospheric research, we had to get this right, since we were looking at VERY weak signals.

1) Gain is cheap!

Some of us who have been around a while remember when some H.F. receivers were "deaf" on the higher bands. QST and Handbook preamplifier and preselector projects abounded, to take care of the sometimes weak performance of receivers. Those days are LONG gone. ANY modern receiver has far more gain than you will ever need (in many cases, far TOO much gain for good performance!

2) Modern receiver input impedances only VAGUELY resemble 50 ohms! It is a RARE receiver indeed (and probably non-existent!) that exhibits a 50 ohm input impedance across the H.F. spectrum. At HIPAS Observatory, we measured the input impedance of many high end amateur (and mil-spec) recievers with a $45,000 HP network analyzer. The BEST of these receivers exhibited input impedances ranging from 10 ohms to 150 ohms! (ignoring reactance, which at some frequencies was quite significant!)

3) It is easy to misinterpret the importance of point #2 above. One might conclude that since receiver inputs are so widely varying, you will gain a lot by using a matching network on your receiver. (Again, in the olden days, there were several published "receiver tuner" articles). However, Point #2 is GREATLY overridden by Point #1. The fact is, receiver input matching doesn't matter, because ANY modern receiver has tons of gain you can throw away. In fact, receiver input matching only begins to be important at UHF frequencies.

4) Although this point has been elaborated COUNTLESS times, both on QRZ and elsewhere, nobody seems to get it. I will repeat it once more, at the risk of sounding tedious, in capital letters. IF YOU CAN HEAR A DIFFERENCE IN THE NOISE LEVEL OF YOUR RECEIVER BETWEEN AN ANTENNA BEING CONNECTED AND ONE NOT BEING CONNECTED, YOU HAVE ALL THE RECEIVER GAIN YOU CAN USE! This is an indisputable fact. There is no advantage whatsoever to having receiver gain beyond this level.

5) A receiving antenna is a signal generator. The question comes up often as to how to model a receiving antenna. The correct model for a receiving antenna is a voltage source with a fixed series resistance. This resistance value is the radiation resistance....remember the reciprocity theorem.
With this in mind, it is important to note that the standing wave ratio on an antenna system may be different in receive than in transmit. Remember SWR is determined strictly by the load impedance. In receive, the load impedance is the receiver input impedance, and your antenna is the transmitter! Ideally, the antenna impedance should be the complex conjugate of the reciever/transmission line input impedance....but again, at H.F. frequencies it doesn't really matter.


Eric


KL7AJ, Nov 30, 2009

[glovisol]

Modern receiver input impedances only VAGUELY resemble 50 ohms! It is a RARE receiver indeed (and probably non-existent!) that exhibits a 50 ohm input impedance across the H.F. spectrum. At HIPAS Observatory, we measured the input impedance of many high end amateur (and mil-spec) recievers with a $45,000 HP network analyzer. The BEST of these receivers exhibited input impedances ranging from 10 ohms to 150 ohms! (ignoring reactance, which at some frequencies was quite significant!)

Well, with RSP class receivers we seem to have found those "RARE receivers indeed". The input impedance of the RSP-1A has been measured and results can be seen here:

https://www.sdrplay.com/community/viewt ... =12&t=4414

A well defined terminating impedance as near 50 Ohm resistive as possible is obviously very important when the RSP-1A is used as a front end for the new Spectrum Analyser software by Steve Andrew. For instance how can you measure a swept filter response if it looks into an arbitrary and undefined termination?

In my opinion SDRplay did not produce designs with well defined input impedance by sheer chance, but because they introduced the concept of received power precision measurement. In fact with this class of receivers we are able to read the value of received power at the antenna terminals in dBm with a specified and measured accuracy of +/- 1 dB. This measurement would be meaningless if the input impedance "VAGUELY resembled 50 Ohm".

This is one of the features that places the RSP class receivers in a category of its own and this fact is becoming more and more recognised amongst the users at large.

I am personally very pleased that Mike2459 has opened a thread dedicated to this important subject!

glovisol

[glovisol]

The complete mapping of the RSP-1A input impedance in the range 0 - 29 MHz has now been completed, demostrating the high quality of the Receiver 50 Ohm socket termination with respect to 50 Ohm. Test results available here:

https://www.sdrplay.com/community/viewt ... =12&t=4414

glovisol

[glovisol]

4) Although this point has been elaborated COUNTLESS times, both on QRZ and elsewhere, nobody seems to get it. I will repeat it once more, at the risk of sounding tedious, in capital letters. IF YOU CAN HEAR A DIFFERENCE IN THE NOISE LEVEL OF YOUR RECEIVER BETWEEN AN ANTENNA BEING CONNECTED AND ONE NOT BEING CONNECTED, YOU HAVE ALL THE RECEIVER GAIN YOU CAN USE! This is an indisputable fact. There is no advantage whatsoever to having receiver gain beyond this level.

I hope I will be allowed to comment on the fact that generic and qualitative statements as those quoted above, do not help the reader, who, in fact, if already versed in this branch of knowledge, does not need them. If, on the contrary, the reader is not conversant with the subject, then he needs QUANTITATIVE and SPECIFIC INFORMATION on what he has to do to optimise the use of his receiver in presence of noise.

This quantitative and up to the point information on the subject has already been given in this Forum, in the following thread, a few months ago:

https://www.sdrplay.com/community/viewt ... f=5&t=3723

cheers,

glovisol

[Mike2459]

The first two Antscope screenshots you posted at the link show Z deviating from 50Ω by 42% @5.475 Mhz and 40% @5.650 Mhz. In my world a 42% deviation is significant.

Cheers

[glovisol]

Hi Mike, welcome back!

Well, the screenshots you mention show a return loss of better than 10 dB. This means that in this case, which is the worst across the entire spectrum 0 to 29 MHz I measured, less than 10% of the power received is reflected back by the 50 Ohm termination provided by the RSP-1A socket. We are talking of a reflected power of approximately 0.8 dB.

It is always a matter of standards: if for you 10% reflection is very significant and you would rather opinate (non professionally, of course) that only 1% (reflected power 0.086 dB) is acceptable, then perhaps you should look for a better receiver, provided you can find one, at any price. Additionally I draw your attention to the fact that the worst case 10% reflection you note is fully compatible with the accurate measurement in dBm of received power at the antenna terminals.

In any case, to paraphrase what you wrote not long ago on a very similar subject, nowhere (truly, this time) in the SDR play specifications is stated the tolerance for the 50 Ohm termination of the RSP-1A. SDRplay states, contrary to different opinion, that the input impedance of the RSP-1A IS 50 Ohm, but they DO NOT specify what the deviation should be across the 0 to 29 MHz frequency range.

According to my very modest (non professional?) experience, the 50 Ohm input termination (specified or not specified.....) of the RSP-1A across the frequency range 0 - 29 MHz (and probably of all the other RSP class receivers, when I shall get a chance to measure them) is the best and most accurate now available on the market.

To tell you the truth, I was simply AMAZED in finding such quality in the RSP-1A 50 Ohm input termination when I measured it.

Cheers,

glovisol

[Mike2459]

IF YOU CAN HEAR A DIFFERENCE IN THE NOISE LEVEL OF YOUR RECEIVER BETWEEN AN ANTENNA BEING CONNECTED AND ONE NOT BEING CONNECTED, YOU HAVE ALL THE RECEIVER GAIN YOU CAN USE!

Let me add to what the author of the above is saying in my own words.
"OPTIMIZE THE SIGNAL TO NOISE RATIO NOT THE RECEIVED SIGNAL POWER, BECAUSE IF YOU CAN HEAR IT YOU HAVE ALL THE GAIN YOU NEED."

In addition, in the majority of cases the reason you can't hear a signal is because it's buried in the noise. That's one of the reason loops like the Wellbrook line are so popular. I doubt that many of Wellbrook owners are concerned about it's output impedance.

[sdrom33]

Antenna and/or receive impedance is an important parameter, regardless of the concerns of Wellbrook antenna users. As far as receiving signals buried in the noise, the issue here is to use the right level of gain, or attenuation, as the case may be, but this has nothing to do with impedance. If you need gain, then an amplifier working between well defined impedance will work best. If you need an attenuator, this will give a precise (and constant with frequency) attenuation again only working between well defined and stable impedance levels.

[Mike2459]

SDROM, You never cease to amaze me. Increasing gain or adding attenuation does one of two things only: It either increases signal AND noise in the case of gain and reduces signal AND noise in the case of attenuation. There is a big advantage to maintaining a low noise floor but adding attenuation is not the way to achieve it. Your only practical option to increase signal if needed and either hold noise constant or reduce it is to:
1) Move to a new , less noisy location. This can be an expensive option.

A more practical and economic path:
1) Relocate your antenna as far away from the worst noise sources.
2) Take active steps to identify local noise sources that YOU can correct - many of them are only a few feet away. Minimize local RFI with chokes and filters. Reduce common mode noise as much as possible with chokes designed for the frequency range that you are using.
3) Install an RF isolation transformer.

By taking steps to reduce noise you may find you have all the signal you need.

[sdrom33]

I am glad to be your source of amazement and wonder, but , Mike you should not change the object of your discussion with every new post...you start by splitting hairs on receiver's input impedance and end up, on your last instalment on radio know how, with advising the obvious.....A new Monsieur de la Palisse was born today!