Greetings friends! It’s time for antennas!
Today, aliexpress is replete with cheap antennas for the 868 band, but 99% of the time they are not tuned or tuned very poorly or very poorly performed.
I modeled and manufactured several practical structures, which, despite their simplicity, are easy to make with your own hands and configure (if you have a NanoVNA V2 device). Moreover, such antennas will have characteristics many times better than stock antennas. This means that the communication range between your Meshtasick-devices can theoretically increase from 2 to 20 times!

Note: Unfortunately, I do not have a 3D printer and I do not have the skills to make complex mechanical products, but I can model and prototype antennas. Therefore, all my designs look like “kitchen-table effort” or “quick-and-dirty”. I invite everyone who knows how to create mechanical structures to discuss and supplement the material presented below with their practical implementations of antenna products.

All related materials can be found on my github page: GitHub - NanoVHF/Meshtastic-DIY

In a neighboring branch, a colleague @j2db1k presented his designs for the 433MHz band:

Let’s start …


Construction 1.

Simple 3-element Yagi antenna.

It has an honest gain of 7dBi (4.8dBd). I designed the antenna so that it would have the maximum possible radiation pattern for 3 elements - about 120 … 130 degrees. This makes it possible to achieve the shortest possible traverse length. It is technologically convenient!
A practical application of this design is the location of the antenna on the window or on the balcony of the house. The back lobe of the antenna pattern is suppressed as much as possible and the energy of a low-power Meshtastic-device is not wasted in absorbing energy into your home. In a wide view, the front of the house is free space. Plus - we have a small antenna gain.

Further development of this construct - we place the LoRa and ESP32 boards on the same board with the traverse. The result is a long-range design. If the 7dBi gain in your application is not enough, the board provides for an extension of up to 5 … 9 elements. 9 elements are obtained with a length of about 50 … 80 cm (depending on the desired gain). All in one package 3 elements fits into the size 190 x 120mm (with a margin, you can still shrink it if you want)

SWR, LogMag, Smith

1st prototip

Ring with a diameter of 2 … 3cm

PCB FR-4 laminate 25x5x1mm with 5mm copper removal in the middle of the workpiece. The gap is included in the total length of the active element 164mm. Coaxial cable: Radiolab RG-316U or similar.


A little later I will show the model for the American audience at 915MHz. I didn’t do it live, but in theory the model should work. (My model at 868MHz above matched the mechanical settings 100%)
If anyone has a desire for other frequencies, then write here, I will adjust the model to the frequencies of your country.

Note 1:
When repeating the design, the main condition for high-quality tuning and operation of the antenna should be the absence of metal components between the antenna elements! If you are going to make a 3D construction for mounting on a printer, then make sure that your material does not contain graphite or metal powder!

Note 2:
Structural elements must be made strictly from tubes with a diameter of 4mm. They must be located strictly in the same plane (+/- 1mm tolerance). If you use pipes of a different diameter, then the parameters can deteriorate very much.
If you want other diameters of elements, but write here, I will make the correct model for your diameters.

Note 3.
You can safely repeat the design of the antenna and it will work without tuning. If you have a measuring device such as NanoVNA V2 or similar, then you can do the ideal tuning. To do this, you may only need to move the active element closer / further a couple of mm from its calculated value.
Depending on your technology for making the active element, you may need to adjust the length of the active element slightly. All this makes sense only if you have a SWR panorama measuring device.


Antenna design for our American colleagues on 915MHz.

Construction 2.
Greetings friends! Time for antennas again!

Today I am laying out the construct of one of the simplest antennas - this is the canonical GP antenna 1/4. You can easily find more than one similar design on the Internet, but here I will show my construct.
The design is actually simple. In the detailed photo, you will see all the elements.

The design is based on the correct microwave connector. This is an N-type flanged connector. It is easy to mount and fits in a variety of designs, which I will describe a little later.

For a clean GP antenna, as shown in the photo above, an important element is the absence of a metal mounting cup. If you plan to mount the antenna on a mast, then it must be dielectric or the glass must be no more than 50mm in length.

The emitting element is made of a brass tube with a diameter of 4mm. The length of the radiating element is 82mm. The tube is neatly sealed on the piping of the connector.

I have a counterweight made of soft copper wire with a diameter of 2mm.
The length of the counterweights that are attached to the connector is 75mm.

When you make counterweights, cut the total length of the elements 100 … 120mm. Counterweight fasteners can be made from the same piece of wire or can be soldered onto special petals (see photo).

To minimize the influence of the power cable, a small loop with a diameter of 2 … 3 cm is made below the connector 7 … 10 cm.

Antenna parameters: VSWR, LogMag, Smith.

If you mount the antenna on a metal glass (not recommended), then the length of the radiating element will need to be slightly increased. I used a regular screw for this.

Perform the setting with the NanoVNA V2 device. We achieve a minimum SWR by rotating the screw and changing the angle of the counterweights (about 45 degrees)


Construction 3.
Hello friends! Time for antennas again!

Today I am laying out the design of another, also, almost the simplest antenna - this is the canonical J-type antenna. On the Internet, you can easily find more than one similar design for other bands, but here I will show my design at 868MHz. I have used a few simple but not obvious solutions to achieve superior gain and VSWR stability. Before fabrication, the antenna was optimized in an electromagnetic antenna simulator.

One interesting solution is the antenna shank. It is also an implicit part of the radiating system. A dielectric mast can be attached to it.

Note 1: With relatively simple mechanics, I do not recommend the antenna to be repeated for novice radio amateurs.
Note 2: When tuning the antenna, I strongly recommend using a NanoVNA V2 meter or a similar panoramic SWR meter.

The whole antenna is actually extremely simple! In detailed photos, everything is clearly visible.
An antenna is made from one piece of a brass tube with a diameter of 4mm and a length of 500mm and a PCB made of FR-4 material with a size of 10x15x1.5mm

The tube must be carefully cut according to the drawing. Only 2 pieces of tube length 370mm and 88mm.

Note: Please note that the short piece needs to be cut 10mm longer for the attachment point.

Note: the coaxial cable Radiolab RG-316/U is sealed exactly as shown in the photo.

An attempt to bring the cable simply downward, as is often shown in various Internet designs, shows a strong deterioration in the VSWR parameter. One gets the impression that numerous Internet authors did not connect their designs to measuring instruments.
When choosing the connection point for the central core of the cable, you need to achieve a minimum SWR. This process can be quite painstaking and complicated. It is difficult enough to achieve ideal VSWR.

To minimize the influence of the cable on the antenna parameters, it is recommended to make a ring of the cable based on a diameter of 2 … 3 cm. The use of a ferrite latch further reduces the effect of the cable below the antenna.

Antenna parameters: VSWR, LogMag, Smith.

Antenna radiation pattern and gain. (Modeled)

A few additional photos for understanding.

Construction 4.
Hello friends! Time for antennas again!

Today I am laying out the design of probably the most-most simple antenna possible, but with good gain - this is not a canonical J-type antenna.
There is no such similar constructions on the Internet!
This design is also proprietary at 868MHz. The same can be done for the 433MHz band. It will be slightly larger.
The idea for this design was born from experiments and modeling of the canonical J-antenna. I asked myself a question, “Why make a complex structure? Can you make a simple one?”
The only thing that is difficult in this antenna is to find a high-quality microwave connector - N-type.

The whole antenna is actually extremely simple! In detailed photos, everything is clearly visible.
Antenna is made from one piece of brass tube with a diameter of 4mm and a length of 500mm.

The tube must be carefully cut according to the drawing. Only 2 pieces of tube length 307mm and 69mm.

Note: For this design, it is very important to use a metal sleeve on a dielectric mast. The length of the glass is not critical - 70 … 100mm

To minimize the influence of the cable on the antenna parameters, it is recommended to make a ring of the cable based on a diameter of 2 … 3 cm. The use of a ferrite latch further reduces the effect of the cable below the antenna.

Antenna parameters: VSWR, LogMag, Smith.

Antenna radiation pattern and gain.

A few additional photos for understanding.

To fine-tune the antenna, I installed a 20mm long M3 screw as the fastening of the second element. It is necessary to slightly squeeze the base of the tube with pliers (carefully!), So that it is tightly screwed onto the screw.

Important note: when using power and mechanical tools and soldering equipment, be sure to follow the safety rules !!!


Workpiece: collinear antenna and housing for radio relay unit.


Awesome job, did a few myself 433mhz Antennas Diy - #20 by ccc

1 Like

I can see this being useful for many people in the future.
Thank you for taking the time to design them and write out this tutorial.

1 Like

Great job of our colleague! Thanks for the link, I’ll add it to the topic header.

1 Like

When mounting antennas outside you should consider antennas that are DC-grounded. For example the J-pole antenna. So that there is no build up of static electricity.
And by the way, if you use a dipole to feed a yagi-antenna you should consider adding a “proper” balun, otherwise the radiation pattern may not be as good as it could be.
Building baluns is pretty easy, if there is any interest I can write a short documentation. All you need is a small piece of 1/2*λ coax cable and you can use a NanoVNA to measure the velocity factor and calculate the required length. I am currently working on a small yagi antenna with folded dipole feeder (-> DC grounded antenna), if everything works as expected I can release the 3d models.

1 Like

Yes, colleague, you are undoubtedly right about the baluns. It would be interesting to get acquainted. Be sure to write the manual for beginner radio amateurs in as much detail as possible.

Today I made a design for a short 5 element yagi. Size 170x170mm. Gain 9.5dBi (7.4dBd).
A little later I will make a full description.

The opening angle of the diagram is 80 degrees.

Antenna parameters: VSWR, LogMag, Smith.

1 Like

Nine steps to improve the stock stubby t-beam antenna:

  1. remove the outer plastic cover (pull hard)
  2. desolder the helical antenna from the center pin of the connector
  3. solder a small piece of thin coax cable on the connector, the kind you can salvaged from broken laptop wifi antennas (shield to shield, center conductor to center pin)
  4. drill two perpendicular holes in the plastic cover, and one in the top.pull the coax cable through the bottom perpendicular hole and wind about 6-7 turns around the plastic cover. Pull the coax through the top perpendicular hole back inside of the plastic housing and through the top hole. Use only clear heatshrink tubing as black heatshrink contains carbon that will mess with the RF properties of the choke coil you just formed!
  5. leave ~8.5cm coax cable, measured from the bend of the top hole. strip the first 5mm including the shield and expose the center conductor
  6. solder a piece of wire to the center conductor, trim to about 8cm total length measured from the end of the coax shield
  7. slip on a small piece of rc antenna protection tubing, press fit in the top hole of the plastic cover
  8. attach to NanoVNA, tune for best SWR by cutting the top piece of wire and/or pushing the antenna tubing down into the plastic cover below the top perpendicular hole to shorten the 8cm piece of coax cable. Do not forget to slip on the antenna tubing after adjustment !
    as it will shift the SWR down a litte bit!
  9. Enjoy the huge range boost! I forgot to do a before/after range test, but I promise you will be amazed. Before I was barely able to go around the street corner but now I can easily cover half of my small town! Pretty impressive considering all the obstructions!

the end result should look something like this:

The design is based around the “flower pot” antenna, you can easily modify the design for the desired frequency range by adjusting the coax and wire length. There is even an online calculator:

This is the kind of protection tubing I used:


Out of curiosity, are you modelling the radiation patterns or are those reuploads based off of previous test data? I just got a NanoVNA and I built a GP antenna and it seems to have the main lobe coming out about 10 degrees up from the perpendicular plane. Id prefer it to have a main lobe parallel to the perpendicular plane of the active element like J-type antenna haha

Also! I’ll have to send you a pic of what I’ve done using an SMA connector hehe. I figured out if you have a spare 2x4 and a hammer, before you place the active element and the ground elements, tap out those kinks with small, light taps and you can get the elements SUPER straight!


Edit: Here is that GP that I was telling you about that I hammered out the elements on a 2x4 with some light taps.

1 Like

I made an attempt to diy a quarter wave for 915mhz. I did use an antenna calculator to get the lengths of the elements. I dont have an analyzer to tune it unfortunately.

Cable came from a disassembled laptop, it has the ufl connector so it plugs right into my boards. Its also fairly long so I imagine there’s a loss in that. The grounding was soldered to a small metal tab through a hole, with the center core poking through. The wires used are thin and flimsy, and the element soldered to the coax is even more so as there’s no support.
Its more of a test antenna for if it was even going to work and from what I can tell, it does work better than the stock antennas that came with my modules. I had a slightly better SNR, even though buildings. By no means though is it an ideal antenna, I still can’t tune it properly and it’s very fragile. I do plan on doing a better version in the near future, with a shorter cable and hopefully using an SMA connector to mount the elements on. And maybe something to tune with…

Feedback is much appreciated, there might be something that I can learn here.


Looking good! The biggest thing when tuning when you get around to it will be to make sure you oversize your antennas so you can trim. It is way easier to take away than it is to rebuild parts or add stuff back later.

Those wires do look pretty thin but that isn’t too much of a concern except when it comes to deforming. What may be a bigger concern for you is the ground plane antenna is more sensitive to geometry and symmetry than other antennas. You will find when tuning that the angles of the ground elements will not only alter gain in a direction but also the pitch of the radiation pattern. Use a small hammer and a 2x4 and roll the wire on the 2x4 with light taps from the hammer and you’ll straighten those elements out too.

You’ll also learn really quick that tuning it inside is way different from tuning it when it is outside. What I recommend when you get to that point is find out where you want it permanently and then tune it. You’ll want to take a reading when it’s up there, bring it down, make an adjustment and try again. There is math you can use to get a more approximate idea of how it needs to be modified but ultimately, it comes down to time and whether it is worth getting it to the absolute frequency or if you are ok with it being a megahertz off