The BIG and SMALL ANTENNA topic part 1


(Tim Everitt) #391

What happens to the RSSI at a gateway when a BIG antenna is replaced with a SMALL antenna?

I have a test gateway in my worskshop. It has been using a Taoglas OMB.868.B05F21 collinear dipole array which is a typical BIG 5dBi high-gain antenna, 1m long, in a white fibreglass tube. This has been replaced with a SMALL Taoglas MA120.A.QP.001 -5dBi gain “puck” omni antenna mounted on a 25cm x 25cm steel plate ground-plane. No other changes.

I have a test “thing” outside, about 25m away and through 2 masonry walls sending uplinks every 2 minutes.

The average gateway RSSI for uplinks from the test thing with the BIG antenna on the gateway has been -65dBm. Range is -62dBm to -67dBm.

Here’s a general picture of the gateway with the new SMALL Taoglas puck antenna.

Here’s a close-up of the new SMALL antenna - under 5cm in diameter.

My estimate of the new average RSSI was that I would lose between 10dB and 15dB. 10dB as the difference between a +5 gain antenna and a -5 gain antenna and then up to another -5 because my ground-plane is smaller than Taoglas recommended 50cm x 50cm ground-plane and because I’m an “antenna pessimist” and because Taoglas bluntly calls this a “short range antenna”.

I now have the new average RSSI. What does the collected wisdom of the forum think the new RSSI will be? Please make your estimates by replying to this topic. A well-done will be awarded to the nearest response(s).


#392

The sensor is at the same height as the antenna?


#393

He used an analog widget, he changed the icon to battery and it’s name :wink:


(Duineuk) #394

Yeah, have since figured it out.

A great “Aha, that’s sneaky!” moment :slight_smile:


#395

sorry… missed your question, but @Charles is always here to help :wink:


(Bluesensing) #396

https://www.thethingsnetwork.org/community/berlin/post/optimizing-your-868-mhz-self-made-lorawan-antennas-with-n1201sa

If you are really interested in optimizing your self-made 868 MHz antennas, or all those, you received from your China seller, whom you trust, you should inhale the content of this extraordinary video from Andreas Spiess (the famous “guy with the swiss accent”) very thoroughfull.

This video is really very helpful for non-specialists and you may check and trim your self-made 3 $ groundplane (if necessary) within seconds. Each IoT (FAB) - Lab and each TTN Community should have at least one available.


(LoRaTracker) #397

Yes very interesting. I ordered the exact same device a couple of weeks ago, specifically because I wanted a device for measuring antennas out doors in real world situations, rather than in the Lab. Using bridges, noise sources and PCs outdoors is often not practical.

You can of course trim\adjust antennas using just RSSI measurements with a couple of LoRa devices and some code.


#398


(Keptenkurk) #399

This weekend i finished a Collinear antenna and wrote a small blog at https://keptenkurk.wordpress.com/2018/03/26/a-lora-868mhz-collinear-antenna/
Couldn’t have done it without (a borrowed) N1201SA.


(Amedee) #400

Interesting…
How is the reception under the antenna?

I did some colinears for 2.4Ghz in the past, but with that many half wave segments it is high gain but also quite directional (flat disc).
With the 8 segments 2.4GHz on my roof, I have almost nothing on the ground under the antenna…

(Memories, memories… my 2.4GHz colinear howto 14 years ago…)


(Keptenkurk) #401

That could indeed be a problem but then again, the range will be very close and reflections will help. Less of a problem at 868Mhz than at 2.4GHz i guess. But i will specifically test that and report back!
Another issue i am worried about is that the main lobe could not be horizontal but face down (bad for the range) or even up (even worse) due to inaccuracy of the segment sizes. I even read designs where this was deliberately done for the antenna of a gateway on a hill top.


(LoRaTracker) #402

Nice, so you built a tuned circuit.

How effective is it as an antenna ?


#403

I didn’t build it … it’s just a nice picture :sunglasses:


(LoRaTracker) #404

A coke can is a very useful object, as a size referance.

Just about everyone in the World will recognise how big a coke can is.


(Keptenkurk) #405

That looks extremely nice. Real professional craftsmanship. Need to polish my French though :slight_smile:
Page bookmarked!


(Kylix) #406

I built long time ago one diapason antenna for 1090 Mhz following the plans found here: http://f5ann.pagesperso-orange.fr/AntennePCB-1090MHz/index.html

Now I see they sell an 868Mhz active version: https://shop.jetvision.de/epages/64807909.sf/en_GB/?ObjectPath=/Shops/64807909/Products/69200


(LoRaTracker) #407

Receive only looks like.


(Remko) #408

Acrice is receive only


(Keptenkurk) #409

The collinear works out very well. See https://keptenkurk.wordpress.com/2018/04/02/lora-868-mhz-collinear-results/ for the results and map-plotting.

DSC03759_anno

Coverage for groundplane at rooftop level (Old LoRa):
coverage_old

Coverage for collinear at 2m over rooftop level (New LoRa):
coverage_new

RSSI levels directly under the collinear (inhouse) dropped from about -65 to -72, so nothing to worry about.


(Schneider Da) #410

I made a 3D printable jig to help you make your own 868MHz antenna from an SMA connecter and some 2mm brass tubing.
The instructions and the resulats are on my blog
Compared to the small monopole antenna that comes with the ESP32 LoRa node I measured an increase in the SNR of over 4dB!