The BIG and SMALL ANTENNA topic part 1


#431

I did some measurements with a N1201SA antenna analyser on two small antenna´s about 50mm long. I got those antenna´s with my Pluto SDR receiver from Digikey. Those are specified as general purpose UHF antenna´s. Inside, connected to the center pin is a straight wire of about 30mm surrounded by a coiled wire 20mm long soldered to the outside of the connector.

noplastic_2
antenna without plastic cover.

I measured the antenna directly on the N1201SA and also with several lengths of 50 ohm coax with a clamp_on ferrite choke close to the antenna side of the cable.
Directly on the analyser the antenna resonates at 820 to 825MHz with a VSWR of 1.1 to 1.2. Dependent on the place of the analyser.
Then I inserted some cablelengths, 15cm, 50cm, 65cm and 100cm with ferrite choke. Holding the antenna by the ferrite clamp so there is no handeffect. With all cable lengths the antenna´s had a resonance frequency between 865 and 870 MHz and a VSWR between 1.05 and 1.35.
Best result I had with a 50cm plus a 15cm cable which including the connectors was 68,5cm. Because cable_velocity factor = 0.66 the electrical length for the cable is 103cm which happens to be 3 wavelengths at 868MHz. When cable_length is a exact number of electrical (1/2) waves, the analyser just sees the antenna and ignores the cable. With this length of cable the antenna resonates at 867MHz with a VSWR of 1.05.
So it seems to be a good 868MHz antenna.

cable1
antenna with 65cm of cable.

Conclusion for this type of antenna´s is that the working is strongly dependant on the device it is mounted on. Meaning the amount of metal that functions as a groundplane. The antenna analyser is, compared to the antenna, a large chunk of metal. I think that is the reason that the resonance frequency was lowered by 45MHz when directly connected to the N1201SA.
Also I wonder how effective this short 30mm wire surrounded by a grounded coil can radiate. I believe that a single straight wire of 85-86 mm long directly soldered to a device is a more effective radiator than this short antenna. But sometimes you need a small antenna for a small device.

Note: without the plastic cover the antenna resonates above 900MHz


#432

I’be been playing with http://martybugs.net/wireless/collinear.cgi


#433

There is no overall gain in the antenna. The first task is to ensure that the output impedance is matched to the transmission line and then the antenna. With no reflections the power is just radiated out, as per the pattern, but integrated - there is no gain. Your lack of signal under the antenna is an achievement! No light comes from the end of a fluorescent tube. (Although that is a poor analogy).


#434

and keep 10 lambda away during any tests.


(LoRaTracker) #435

Indeed.

I have just modified my ‘field strength meter’ software (for a LoRa device) such that the device that is measuring the field strength produced by the antenna under test relays out the RSSI measurement as a LoRa packet. Thus you can pick up the measurement when standing some distance away from the antenna under test.


#436

Got the N1201SA last week and tried it on some chinese antenna from eb*y (supposed to be 868 Mhz).
2 had a broken connection inside somewhere and the rest all seemed to resonate better for 915Mhz… They did work OK on a Heltec (868Mhz).

Any one found a source for cheap antenna that really are best for Lorawan (863-870Mhz) ? :slight_smile:


#437

For a larger antenna this 19,5cm one is relatively cheap and not too bad: The BIG and SMALL ANTENNA topic part 1 (I don’t have an analyzer for 868MHz to do an objective test.)
I have several of these and use them on some gateways.

For the smaller ones I have tried several antennas from different sellers on AliExpress but their construction and performance are (very) different and difficult to test (objectively) without a good analyzer.

(I open each of my antennas to check what I received and what is under the hood.)


(LoRaTracker) #438

Its probably never going to happen.

By ‘cheap’ we assume you mean someone on eBay or Aliexpress or other ?

Appreciate that those selling on these platforms almost certainly know zilch about RF and antennas.

So whilst on Monday a ‘source’ might be selling a ‘good’ antenna, by Tuesday thier supplier has changed, so the antenna is different and could well be a ‘bad’ one. Thus quoting ‘cheap’ suppliers can be a waste of time.


#439

That can even be on the same day, in the same batch of antennas.


#440

Not 100% guaranteed but you should try. Look on aliexpress for the brandname ´allishop´ , they have a lot of RF things.
They have a antenna called 5dBi omni antenna, i measured a few of them and they were exact on 868MHz.
Also they give specifications about the gain, vswr, BW, impedance and radiation pattern, nobrand name antenna´s usually do not do that.
Allishop antenna´s are a little bit more expensive than nobrand name antenna´s.
If you order one, look for the brandname ´allishop´ in the advertisement, I think you have a good chance of getting a real one.

There are many sellers who have an antenna that looks exact like that one, same length and they always claim it is on 868MHz and has 5dBi gain, but often the seller doesn´t know anything about RF. So chances are great you get something what is not as advertised.

A few days ago I tested for someone this same (lookalike) antenna, it was (according to the gateway owner) sold together with the 868MHz TTN-gateway by Farnell. The N1201SA analyser told us the antenna was spot on 915MHz. So even Farnell may be fooled by the chinese and sell you the wrong antenna.


#441

That is exactly the antenna that I mentioned some posts above and earlier in this thread. With a link, photos of its internals and a drawing with measures: The ‘5dBi’ 868MHz 19.5cm SMA antenna from ALLISHOP.


#442

Indeed bluejedi, it is the same antenna. Your drawing shows it is a full dipole. So it should be advertised as 2.15 dBi and not as 5dBi antenna.


(Remko) #443

I did a quickscan on 12 dBi antennas on internet:

It is tempting to believe the first two but I think that the chance for a real 12 dBi antenna is with Taoglas.


#444

How about this to make a cheap 868 Mhz antenna.

  • Buy a cheap one off Ebay
  • Snap the top off
  • strip a bit of insulation off, solder a small piece of single strand wire
  • Repeat the following until satisfied - Put top back on, measure with N1201SA and cut until happy (see pics for difference if top is on or off)

Pic 1 shows stock Antenna which resonates at 915Mhz, Pic 4 shows final result - not too shabby (Measurements all at 868 Mhz) :slight_smile:
.
.
.
Pic 1 - shows stock antenna
before%20mod
.
.
.
Pic 2 - showing how much extra wire to solder on to tip
tip
.
.
.
Pic 3 - With top off
without%20cap
.
.
.
Pic 4 - Final result - with top on
with%20cap
.
.
.
.
edit Did some tests with a Heltec.
I sent a regular message within the house (> 10 lambda) and checked the RSSI after alternating between a stock and an altered antenna - did the altered antenna make a difference?
Ignoring spurious results, it did seem that the modified antenna was better - but only by a few dB. The test certainly wasn’t scientific so maybe Loratracker can give a better test?


(Leondegeling) #445

Combining recreational activities is always nice. I decided to put a battery-powered node on the mast of my small sailing boat. I basically modified the rak811-trackerboard firmware to drastically reduce the number of transmissions and to reduce the size of the payload. I added a small rubber duck antenna and wrapped the lot in a pvc tube which is fixed to the mast with a couple of tiewraps.
Then I started sailing, all around the lake. Which is for about 50% covered if I check TTN mapper. However when I checked Cayenne after 3 hours of sailing i only found some datapoint from the morning before when I tested the node after building it into the tube, and same datapoints from when I returned home. No coverage on the lake at all. I am now thinking of replacing the antenna for something better.I was wondering if anyone here has positive experiences using lora on a small boat. I would figure a big flat body of water would be positive for the range, but my RF knowledge is non existent. If anyone can recommend an antenna that doesn’t mind getting the occasional beating and can be put in a mast that would be great.


(Ezequiel Falcón) #446

Hi guys!
This week we´re installing this antenna in a 30mts tower:
https://www.aliexpress.com/store/product/Glass-Fiber-Antenna-Gain-8dbm-Transmission-range-is-further-LoRa-Gateway-Antenna-with-433-470-868/2805180_32852314761.html?spm=2114.12010608.0.0.392d35880xcJvd

With a IP-65 enclousure using RAK813 gateway:

Sin%20t%C3%ADtulo

Our goal is to cover several end devices (street lamp modules we´ve made) and this is our first field test!
The ideal coverage: 10km. The nodes are going to be 9mts above ground so I think that it will work

I will let you now how it goes! If you have any advice, it would be great!!


(Tbalon) #447

Finally found a use for one of my empty containers … :grin:

I added my newly constructed ground plane tuned for 915 Mhz to my gateway. Tremendous improvement in signal quality. Was easy to build but took a bit of tuning to get it just right. I used my NS1201 VNA.

gateway


(Remko) #448

Could be a topic for on te work bench thread.
I have measured the antenna I found in a device for elderly care.
It was told to me to be om 868 MHz. bu the device seems tuned for the US market eg. 915 MHz.

VNA_180421_095745

afbeelding

afbeelding


#449

Nice. “Aged 12 years” :grin:

I had the impression that this antenna type already provides it’s own ‘ground plane’
(because it is some variation of a di-pole).

What was the effect of adding the container? Is it all metal or only a metal top?


(Tbalon) #450

The container is made of compressed cardboard… so no real impact from the container itself. That being said, the top is metallic. Some sort of tin or aluminum …

To get a good match I had to tune the antenna once mounted on the top of the canister. The additional ground formed by the metallic top changed its impedance. All that was required to compensate was to change the angle on the ground radials so that the impedance at 915 mhz was 50 ohms. There was a small hole in the back of the container and I ran the feed line out that and connected it to the VNA. You can see the radials are not quite 45 deg.