Blog Post Actions
Group Actions
Engagement
Author: gecoz
Date Created:
Views: 844
Likes: 8
Comments: 6
Related
Recommended

Building a poor man’s quarter-wave 433MHz antenna: Comparing Antennae

gecoz

 

In the previous blog, I have built my quarter-wave monopole antenna, and measured using NanoVNA. Since I have got some other 433MHz antennae around, in this article I will focus on measuring their Return Loss, SWR and Impedance, to compare it to my own. Below is the array of antennae that will be measured.With the exception of the Molex and my own built, the others are all cheap antennas, got from Aliexpress and eBay. The antennae used are:

 

  • Helical Antenna
  • Molex Flexible Printed Circuitboard (FPC) Antenna
  • Center Loaded Coil Antenna
  • Small FPC Antenna
  • Quarter-wave Monopole Antenna

 

 

Before starting, I want to stress again I'm taking the measurements in a small room, with probably a good amount of reflection from walls and furniture, and using only a NanoVNA, so when I say "compare" I am abusing the word a little, as to be able to properly compare such diverse antennae would require better equipment and much more time!

 

Helical Antenna

This antenna is the one I got delivered as part of the TTGO ESP32 LoRa board kit. It is basically and helical antenna, soldered to a coax and terminated with an IPEX connector (for the measurement I will need the IPEX-to-SMA adapter). Because has been sold as the antenna for the 433MHz board,From what I can see,  this antenna does not have a ground plane.From the measurement, it looks like the antenna, despite being sold as tunes for 433MHz, resonate at about 395MHz, with a return loss of -7dB.

 

 

Molex FPC Antenna

The antenna chosen from the Molex kit is the 433MHz, with IPEX connector (so I need to use the IPEX-to-SMA adapter). Measuring this antenna I have had some strange results. I tried slightly changing position, rotating the antenna slightly, but no change: it looked like this Molex antenna, at 433MHz, was less than ideal, showing a return loss of -about -1dB and SWR 21.89! Since, in its "open" position, I wasn't getting any joy out of this antenna, I thought why not try folding it, as it is an adhesive antenna, so bending it should do no arm to the structure. I tried to fold it into a cylindrical shape, secured the end with some masking tape, then hooked it up again to the NanoVNA and... with my surprise, the results now were a lot better, showing a return loss of about -6.7dB and SWR 2.717. I cannot explain why, if left flat and open, the antenna does not perform at all, while once folded, seems to behave more.

 

Below there is the measurement taken with the antenna "as is" :

 

And now measurement taken with the antenna "folded" in a cylindrical shape:

 

Center Loaded Coil Antenna

This is another antenna I got of eBay, claiming to be a 5dBi center loaded coil antenna.It comes with a 1 meter cable, terminated with a SMA connector.I wasn't expecting it to perform in any good way, as this antenna is typically advised as a "multi band antenna, but I was pleasantly surprise by the results: return loss -6dB and SWR di circa 3. It seems the  antenna appears to have a resonance at about 420MHz.

 

 

Small FPC Antenna

This antenna is "reduced" version of the Molex flexible antenna. I got those in a pack of 10 from Aliexpress for about a fiver, so no big expectations on the performance. As the Molex, it comes with IPEX connector (IPEX-to-SMA adapter needed to connect to the NanoVNA). Again, it is sold for operating at 433MHz, but from my measurement, resonance is found around 405MHz.

 

 

Quarter-wave Monopole

I have already described the antenna in my previous post, so I will not repeat myself.

 

Results

 

Here I summarise the results of the measurements, in tabular form for convenience.

 

Antenna Return Loss (dB) @ 433MHz SWR @ 433MHz Resonance Frequency (MHz) / Return Loss (dB) / SWR                              
Helical -1.541 11.304 370MHz  / -12.879  /  1.587
Molex FPC (folded)* -6.707 2.717 445MHz  /  -13.393  /  1.544
Molex FPC (flat)* -0.794 21.892 370MHz  /  -6.745  /  2.704
Center Loaded Coil -6.010 3.005 420MHz  /  -15.479  /  1.405
Small FPC -1.735 10.047 405MHz  /  -14.149  /  1.488
Quarter-wave Monopole -36.43 1.031 tuned at 433MHz

 

Looking at the data, the only antenna tuned for resonance seems to be the one I built. It is worth mentioning that resonance is not a necessary condition for an antenna: there are many cases where antenna need not to be tuned for resonance, like for the one used for multi band operation. Of all the antenna, the Molex is the one that still puzzles me, because of the odd measurement when used "flat"... (please check the update below)

 

In my next and last blog, I will set up the TTGO boards and test which antenna performs best as distance coverage.

 

 

*Update on Molex antenna performance - 14/10/2020

 

Thanks to 's comment below, I now know why the Molex antenna is performing so badly, and I have to say it is all my fault for just assuming it was a free air antenna. On the application document, it is stated clearly that it needs to be stuck on a layer of PC ABS material. After checking, I found a 4mm thick piece of clear polycarbonate, so I stuck the Molex antenna on it and measured in free space (no ground plane). The results are now consistent with the application document, as you can appreciate from the images below:

 

Note that, on the application document, the PC ABS layer thickness suggested is 2mm. I have used a 4mm thick layer, but the effect does not seem to degrade the antenna performance too much.

 

Updated results

 

Antenna Return Loss (dB) @ 433MHz SWR @ 433MHz Resonance Frequency (MHz) / Return Loss (dB) / SWR                             
Helical -1.541 11.304 370MHz  / -12.879  /  1.587
Molex FPC (on poly) -19.489 1.237 432.6MHz  /  -19.953  /  1.224
Center Loaded Coil -6.010 3.005 420MHz  /  -15.479  /  1.405
Small FPC -1.735 10.047 405MHz  /  -14.149  /  1.488
Quarter-wave Monopole -36.43 1.031 tuned at 433MHz

 

 

---------------------------------------------------------------

Other articles of this series:

Building a poor man's quarter-wave 433MHz antenna: Introduction

Building a poor man’s quarter-wave 433MHz antenna: Antenna’s construction

Building a poor man’s quarter-wave 433MHz antenna: Testing the antenna with ESP32 LoRa

Anonymous
  • in reply to Fred27

    Indeed, it is! I was pleasantly surprised by how well it performed around the 450MHz, considering that I have read as well that it works best for frequency range up to 300MHz, despite the upper limit of 900MHz. Definitely worth every penny

  • The nanoVNA is an incredible piece of equipment for the price, isn't it? I've heard it has some limitations when the frequency goes up, but I found it did an excellent job at 14.56MHz for an NFC antenna. It looks like it's good for sub-1GHz stuff too.

  • Hi Fabio,

     

    Very interesting experiments!

    The measurements can also be affected by the long cable, so if you wish to isolate that (i.e. since that cable does not form part of the antenna when deployed), ferrite cores can be placed around the cable.

    For example for use up to 1GHz, I used this one: Wurth 74270061Wurth 74270061 but the hole is too small for the SMA connector to go through, so it is only suitable for self-assembled cables. Otherwise, any typical clip-on ferrite for RFI suppression can be used, but they're slightly more expensive (ideally a few ferrites are used, to increase the impedance).

    Also as you know measurement will also be affected by the final intended enclosure too (discussed in the Cypress Antenna PDF doc example from that doc below for a different antenna and frequency).

  • in reply to gecoz

    Found some polycarbonate, but it is 4mm thick. Still, seems to work ok

     

  • in reply to jc2048

    Hi Jon,

     

    Thanks for the link, I wrongly assumed it was a free air antenna. Now all makes sense, also why folding it makes it work a lot better. I will have to go in the garage and see if I have a piece of polycarbonate to use as base, and repeat the measure. I doubt I have the exact dimension needed (in particular for the thickness), but worth a shot.

     

    Fabio