I have lots of antennas at the moment from my mesh-core experiments. I have no idea which are good and which are bad.
How can I test them and compare them in quality?
I have an oscilloscope with a signal generator if that helps.
1 Like
Hi Pix
Without the appropriate test instruments with great difficulty
Cheers Bob
1 Like
A VNA should help in seeing how well the antenna (and coax) is matched. You can get cheap VNA but make sure it covers the frequency you want to check (but I suspect like most things, you will get what you pay for). At the very least, you should be able to use a cheap VNA to compare the antennas to each other as a relative thing.
I also want to stress that simply because the antenna looks well tuned, it may or may not perform better or worse then others, as performance is very environmental and gain/direction etc of the signal needs to match your setup. Higher gain is not always what you are after.
But just to show… the following VSWR shows a 915Mhz antenna I tested with its supplied base, sitting on my desk (so may have had interference/reelections from my monitors) i.e. not a clean test… shows the center frequency of the antenna is a far bit higher then 915 Mhz.
Then the smith chart shows that same 939Mhz is near 50ohms, so really good at 939Mhz.
When I moved the antenna away from my desk, it did drop fo 933Mhz, so still a bit on the high side, but this was more to show that location can matter when testing and using antennas.
It is important to note, with passive antennas they will nearly always have “1” ideal frequency then be less ideal as you move away from the center frequency.
You can also see in the smith chart that the start frequency was at 2 on the center line… this means the seen “resistance” at the start frequency (909Mhz) was 2 times the nominal; i.e. 2 * 50 ohms = 100 ohms, which is also reflected in the VSWR near 2. so this mismatch will been less then ideal amount of power is getting sent at that 909Mhz
2 Likes
Hi Pix
Add on to michael post above.
These 2 displays are good to ascertain the antenna match to the cable as VSWR (top) and complex impedance (lower) in their respective display formats but just as important is the radiation pattern.
The radiation pattern will indicate the gain and directivity (front to back ratio) of a particular antenna.
For instance an omni-directional vertical dipole should be circular (horizontal pattern)indicating unity or 0db gain. Some specify gain referenced to dbi which is about +2.8db. The reference here is an imaginary dipole in free space which in reality does not or cannot exist. BUT the numbers look bigger when quoting gain so I suppose for marketing it looks a bit better. In all fairness the numbers are quoted indicating this imaginary reference by using the term dbi.
I personally prefer the “db” measurement as this to me is more realistic and can be easily measured.
When more elements are added the circular pattern is “squeezed” in one direction ant the “lobe” will become larger in this direction and smaller in the reverse. Thus an apparent “gain” is achieved. This can be measured easily as “db” by first measuring the signal strength with an omni-directional dipole, this level becomes 0db. Then replace the dipole with the directional antenna and remeasure observing antenna polarity. The increased signal can then be expressed in db which will be the effective antenna gain. By turning the antenna 180º the front to back ratio can be ascertained.
While VSWR and Complex impedance measurements are useful in measuring cable match and physical resonance of an antenna unless said antenna is home made there is little the hobbyist can do about it as alll this is physics and is determined by component physical dimensions.
All these measurement techniques would be important for comparison purposes. But for a complete picture ALL the measurements are important
Cheers Bob
2 Likes
Hi Jonny,
As Michael and Bob mentioned, a VNA is the way to go - without one, or a specialised RF test suite, the next best option is probably field tests!
A fixed-distance test would be the way to go:
- Keep both nodes and their orientation stationary
- Power down before changing antennas
- Send plenty of packets with each antenna
- Comparing packet loss, RSSI and SNR
- Repeat outdoors with clear line of sight (ideally) or, better yet, match the environment you intend to use them in
Liam
1 Like
Good test
Hi liam
Not really. This will tell you how well the antenna matches the cable and what sort of resistive and complex impedances are involved. While important it is not something you can do much about with commercially built antennas. the other characteristics could be regarded as more important in the real world.
As an interesting bit of trivia that some might be interested in.
If you have a 50Ω source and a 50Ω load such as an antenna. Now if you connect with a cable which is an exact number of half wavelengths the source will “see” 50Ω irrespective of cable impedance. That is it could be 60Ω or 75Ω or whatever, it will make no difference. The source will “see” the load as 50Ω
Handy to know if you have a 50Ω system and only have 60Ω or 75Ω cable. When measuring half wavelengths don’t forget the cable velocity factor.
Cheers Bob
1 Like
