logo elektroda
logo elektroda
X
logo elektroda

YAGI PCB 2.4GHz WiFi antenna for ESP-NOW

TechEkspert  28 5283 Cool? (+11)
📢 Listen (AI):

TL;DR

  • A 2.4GHz YAGI PCB Wi‑Fi antenna for ESP32 ESP-NOW links is tested as a range booster between two buildings.
  • The receiver uses the PCB YAGI instead of an omnidirectional antenna, with a u.FL-to-SMA adapter, and the board includes three directors, a dipole, and a reflector.
  • The antenna measures about 50mmx100mm and enabled one-way ESP-NOW transmission over approximately 430m.
  • Standard omnidirectional antennas worked at about 300m in one setup, but failed at 430m until the YAGI was fitted.
  • Urban 2.4GHz conditions and many Wi‑Fi APs can severely hamper propagation, so the antenna is not a guaranteed fix.
Generated by the language model.
Two green PCB YAGI antennas on a white background.
There are YAGI Wi-Fi PCB antennas available on Aliexpress for the 2.4GHz band. However, I didn't need such an antenna for a typical Wi-Fi card. I'm using ESP32 modules with omni-directional antennas, which transmit information between two buildings about 300m apart after ESP-NOW. One module sends information, the other receives information, it is a one-way transmission. There is no visibility between the buildings, which are obstructed by trees and other buildings. The distance of 300m is not a special achievement, as one can find information about successful attempts with ESP-NOW at 1km with optical visibility.



Such a simple set-up allowed data transmission over a distance of 300m:

Two ESP32 modules with attached omnidirectional antennas.
.

However, sometimes all that is needed is a reinforced concrete building with a large slab in the communication path and the range barely reaches 50m or less. I had a chance to check the included omnidirectional antennas and you can see the alignment for the 2.4GHz band:

Smith chart showing antenna characteristics for the 2.4 GHz band.
.

What did I need the PCB YAGI antenna for?

I wanted to use similar connectivity between two buildings about 430m apart. This time the omnidirectional antennas did not provide transmission. Unfortunately I did not know whether the problem was the distance or the terrain obstacles. It remained to do a practical test and the PCB antenna went into the receiver.

Also, I had the opportunity to test whether the antenna responds to frequencies in the 2.4GHz band:

Image of a spectrum analyzer screen with frequency band graphs at 2.4 GHz.
.

Interestingly, the PCB antenna detached when placed on a clapboard desk. Perhaps the moisture in the wood affected the antenna's performance?

The antenna connector required a suitable u.FL - SMA adapter:

Side view of a PCB YAGI antenna connector with visible mounting details.
.

After using a YAGI antenna on the receiver, transmission over approximately 430m was possible!


The dimensions of the antenna are approximately 50mmx100mm so place greater mounting requirements than an omnidirectional antenna:

PCB YAGI antenna board on green background with rulers on the side.
.

The PCB shows three directors, a dipole and a reflector:

PCB YAGI Wi-Fi antenna with three directors, a dipole, and a reflector.


On the other side there is some magic with the paths concerning the w.cz ;)
Request for a hint as to what is going on here, is this some kind of impedance matching?

PCB YAGI Wi-Fi antenna on a green board with three metal strips and a connector.


The antenna has proved successful in providing adequate range for transmission, but it is not always possible to transmit data using ESP-NOW in the 2.4GHz band,
urban conditions severely hamper propagation in this band. The numerous Wi-Fi APs operating in this band also do not make the task any easier.

Have you used PCB antennas or ESP-NOW, what kind of ranges have you been able to achieve? .

About Author
TechEkspert
TechEkspert wrote 7016 posts with rating 5443 , helped 16 times. Been with us since 2014 year.

Comments

sq3evp 08 Aug 2024 08:42

A Yagi antenna is always a gain because of its directionality. For the same power, much more power radiated in the concurrent direction, also in reception increased gain. [Read more]

ArturAVS 08 Aug 2024 09:47

. It is rather not clinker but furniture grade, like MDF, HDF or particleboard. In the furniture industry it is not used, if only because the surface is not very even (coarse chip). You made me curious... [Read more]

efi222 08 Aug 2024 10:22

The magic mentioned may be the symmetrizer matching the impedance of the dipole to the coaxial cable. [Read more]

sq3evp 08 Aug 2024 11:42

. You've hit it - a Yagi antenna is a directional antenna as a rule. [Read more]

TechEkspert 08 Aug 2024 15:11

It has a directional characteristic, the transmitter had a vertical omnidirectional antenna so I vertically aligned the dipole in the receiver, then with the directors in the direction of the transmitter... [Read more]

sq3evp 08 Aug 2024 15:52

. Similarly a very good price/effort/effect ratio. I haven't tested it, I once made such an overlay for an omnidirectional antenna pretending to be a Yagi (i.e. reflector and directors, the antenna did... [Read more]

ArturAVS 08 Aug 2024 17:34

Mine made from workshop junk and painted with some spray paint tip after 7 years; https://obrazki.elektroda.pl/2480108200_1723131005_bigthumb.jpg I already have new cabling laid and it's... [Read more]

efi222 08 Aug 2024 17:44

If ESP Now packets are being lost, you could try switching the receiver and transmitter modems rigidly to the 802.11b standard. https://obrazki.elektroda.pl/5992936000_1723131706_thumb.jpg . For... [Read more]

TechEkspert 08 Aug 2024 17:50

The omnidirectional one should be quite simple: https://mbharris.co.uk/articles/antenna/omni/quarter/ There are also those made from several sections of coax: http://wireless.gumph.org/articles/homemadeomni.html Omnidirectional... [Read more]

efi222 08 Aug 2024 18:35

I have little experience with ESP32, but Espressif reports that up to 1km straight line range can be achieved in LR mode (ESP32 only) at the expense of speed. The gain over 802.11b is 4dB, which translates... [Read more]

austin007 08 Aug 2024 19:14

. These canned Pringles are steamy. Apart from social media fame and ease of making, they have no merit. I made carefully, measured and was hugely disappointed. I made mine from a tin can with converted... [Read more]

sq3evp 09 Aug 2024 08:13

Pringles would work if they were metal like a real can. They are so-called horn antennas, very directional. And something like this, a combination of a yagi in a tube: https://www.youtube.com/watch?v=Thpdfw2b9uM... [Read more]

Anonymous 10 Aug 2024 17:20

Unfortunately 802.11 LR mode is not supported in ESP-NOW. [Read more]

austin007 10 Aug 2024 17:29

That's not the problem. I hadn't even considered non-metallic Prngels packaging. In a forum like this, making aerials out of these non-metal or non-metal cardboard ones should only inspire pity. The tubes... [Read more]

efi222 10 Aug 2024 17:49

Here are tests of this can and measurements. It appears that the diameter is too small. https://www.speedtest.pl/wiadomosci/sprzet/antena-z-puszki-do-pringlesow-prawda-czy-mit/ [Read more]

Anonymous 10 Aug 2024 17:55

The topic of WiFi strip antennas (and more) has been covered on this forum in the distant past: https://www.elektroda.pl/rtvforum/topic127400.html [Read more]

efi222 10 Aug 2024 18:09

Still coming back to the 802.11 modes of ESP operation. I can't find what the default mode is e.g. When running ESP as a soft AP. Maybe someone has read up? [Read more]

Anonymous 11 Aug 2024 13:32

. You probably won't find this information in the official Espressif documentation. However, there is nothing to prevent you from checking it. For the ESP8266 you can use the function wifi_station_get_ap_info... [Read more]

Anonymous 13 Aug 2024 23:01

HDF or MDF boards are clapboards precisely.... and particleboards are MFP boards (most commonly used in the furniture industry), coarse chipboards are OSB boards which have recently been used like furniture... [Read more]

FAQ

TL;DR: At about 430 m, one user made ESP-NOW work by replacing the ESP32 receiver’s omni antenna with a directional 2.4 GHz PCB Yagi; as one reply put it, "a Yagi antenna is always a gain because of its directionality." This FAQ helps ESP32, Wi‑Fi, and ESP‑NOW builders decide when a cheap PCB Yagi solves weak links between buildings. [#21183948]

Why it matters: In obstructed 2.4 GHz paths, antenna pattern, polarization, and mounting often matter more than transmitter power.

Antenna / option Pattern Thread result or takeaway Best fit
Stock omnidirectional antenna Omnidirectional Worked over about 300 m in one obstructed path, but sometimes failed below 50 m through reinforced concrete [#21182906] General coverage
2.4 GHz PCB Yagi Directional Enabled ESP-NOW reception over about 430 m; >30° misalignment caused signal loss [#21184362] Point-to-point links
Biquad Directional Repeatedly recommended for DIY because of strong price/effort/effect ratio [#21184397] DIY Wi‑Fi links
Panel / microstrip / log-periodic Directional Considered as alternatives, but not chosen in the test setup [#21184362] When you need other form factors

Key insight: The PCB Yagi worked because it concentrated receive gain in one direction and had to match the transmitter’s vertical polarization. Even a small aiming error or nearby object could shift performance enough to break the link.

Quick Facts

  • A simple ESP32-to-ESP32 ESP-NOW link with omnidirectional antennas worked across about 300 m between buildings, despite trees and other buildings in the path. [#21182906]
  • The same user could not pass data at about 430 m with omnidirectional antennas, but reception succeeded after fitting a 50 mm × 100 mm PCB Yagi on the receiver. [#21182906]
  • The tested PCB Yagi had 3 directors, 1 dipole, and 1 reflector, and the user reported signal loss when pointing drifted by more than about 30°. [#21184362]
  • Independent forum measurements put this small PCB antenna at about 7 dB gain over a 1/4λ ground-plane, 8.1 dB over a 1.2λ dipole, about 14 dB front-to-back, and roughly 12–13 dB return-loss minimum. [#21199071]
  • The antenna’s tuning shifted by tens of MHz when placed on a table, held in a hand, or brought near objects, and two such antennas needed at least 40 cm spacing to avoid interference. [#21199071]

How did swapping an ESP32 receiver from an omnidirectional antenna to a 2.4GHz PCB Yagi make ESP-NOW work over about 430 m between buildings?

It worked because the Yagi added directional receive gain where the omnidirectional antenna spread sensitivity everywhere. The original omni setup failed on a roughly 430 m path with obstacles, but the same link started working when the PCB Yagi was installed on the receiver. A similar omni setup had already worked over about 300 m, so the change was not ESP-NOW itself but the antenna pattern and receive margin. [#21182906]

What is ESP-NOW, and how does it differ from regular Wi‑Fi communication on ESP32 modules?

ESP-NOW is a direct ESP module radio link used here for one-way data transfer, not a normal Wi‑Fi client-to-access-point session. In the thread, one ESP32 sent data and the other received it between buildings without discussion of a router or regular AP traffic. That makes it useful for simple telemetry links, but it still shares the crowded 2.4 GHz band and suffers from urban blockage and Wi‑Fi interference. [#21182906]

Why does a 2.4GHz PCB Yagi antenna lose signal when it is misaligned by more than about 30 degrees?

It loses signal because a Yagi concentrates energy into a narrow forward lobe instead of covering all directions. The user reported that a deviation greater than about 30° caused loss of signal on the 430 m link. Another reply summarized the reason plainly: “a Yagi antenna is always a gain because of its directionality.” High gain and narrow pointing come together. [#21184362]

What is the RF trace "magic" on the back of a PCB Yagi antenna, and is it used for impedance matching or a symmetrizer?

It is most likely feed-network copper used for impedance matching and a balun-like function. One reply identified the back-side RF traces as a symmetrizer that matches the dipole impedance to the coaxial cable. The front side held the visible directors, dipole, and reflector, while the back side handled the less obvious RF feed details. [#21184056]

How should a 2.4GHz PCB Yagi be oriented relative to an ESP32 transmitter with a vertical omnidirectional antenna?

Match polarization first, then point the directors at the transmitter. The user kept the receiver dipole vertical because the transmitter used a vertical omnidirectional antenna, then aimed the Yagi directors toward the transmitter. That orientation produced the working 430 m link; larger pointing errors quickly reduced signal. [#21184362]

Why can nearby objects like a desk, wood, your hand, or metal parts shift the tuning and SWR of a 2.4GHz PCB antenna?

They detune the antenna because 2.4 GHz structures are small and very sensitive to nearby dielectric and conductive materials. One tester found that a table, a hand, wood, or metal could move center frequency and SWR by tens of MHz. In extreme cases, the antenna could stop working properly until it was remounted in freer space. [#21199071]

802.11b vs WIFI_PROTOCOL_LR on ESP32: which mode gives better range for long-distance links, and how does that relate to ESP-NOW support?

WIFI_PROTOCOL_LR offers longer range in theory, but the thread states that ESP-NOW does not support 802.11 LR mode. One poster cited Espressif claims of up to 1 km line-of-sight in LR mode and about 4 dB advantage over 802.11b, but a later reply added the critical limit: “802.11 LR mode is not supported in ESP-NOW.” For ESP-NOW links, that removes LR as a usable fix. [#21186603]

What is a symmetrizer in an antenna feed, and why would a dipole on a PCB Yagi need one when connected to coax?

“Symmetrizer” is an antenna-feed network that connects a balanced radiator to an unbalanced line, while controlling impedance and feed symmetry. A PCB Yagi dipole needs one because the radiator is balanced, but the coax feed is not. Without that transition, current imbalance can disturb matching, pattern, and efficiency in the 2.4 GHz feed section. [#21184056]

How do PCB Yagi, panel/microstrip, logarithmic, and biquad antennas compare for 2.4GHz Wi‑Fi or ESP-NOW links?

PCB Yagi and biquad came out as the most practical directional choices in this thread. The user considered panel or microstrip antennas and even a broadband log-periodic type, but finally chose the PCB Yagi for the 430 m test. Other posters repeatedly recommended the biquad for DIY, especially for its price/effort/effect ratio, while the Yagi won when narrow pointing and extra gain mattered most. [#21184362]

What steps help test whether a cheap AliExpress 2.4GHz PCB Yagi is genuine and not a badly wired or fake design?

Use a simple 3-step check. 1. Inspect the layout for a real reflector, dipole, and directors. 2. Verify the connector does not short center conductor to shield in a bogus way. 3. Measure return loss or SWR near 2.45 GHz and confirm directional behavior. The thread warned that some PCB “Yagi” products are cheated designs with only an open dipole or bad connector wiring. [#21184362]

How can I build a simple DIY omnidirectional 2.4GHz antenna for Wi‑Fi if a directional Yagi is too narrow for my setup?

Build a simple quarter-wave vertical or a coax-section omnidirectional antenna. The thread linked a quarter-wave omni design, mentioned coax-section omnidirectional builds, and also pointed to omnidirectional biquad examples. That matches the need described by a user whose phone lost signal after moving only a few meters from the workshop. [#21184575]

Why is FR4 considered a poor substrate for 2.4GHz antennas, and when is PTFE/Teflon laminate worth the extra cost?

FR4 is considered poor at RF because its properties are less stable for microwave work, especially compared with PTFE. Posters said PTFE would be ideal because it offers more constant high-frequency parameters, but its price discourages use. Even so, the FR4 PCB Yagi still delivered over 400 m in practice, so PTFE becomes worth it mainly when you need better parameter stability than low-cost FR4 can give. [#21190497]

What real-world gain, front-to-back ratio, and SWR should I expect from a small 2.4GHz PCB Yagi compared with a dipole or 1/4-wave antenna?

Expect roughly 7–8 dB more gain than simple reference antennas, plus modest rear rejection. One tester measured about 7 dB over a 1/4λ ground-plane, 8.1 dB over a 1.2λ dipole, and about 14 dB front-to-back attenuation. Return-loss minimum was only around 12–13 dB, so this was usable but not an exceptional match. [#21199071]

How do you weatherproof a PCB Yagi for outdoor use without badly affecting its 2.4GHz performance?

Keep the cover thin, non-metallic, and spaced so the antenna still sees mostly free space. One poster said the indoor PCB antenna became suitable outdoors after enclosure in a yogurt cup, and tests with a plastic or foam lid changed parameters only slightly. The same post stressed that less material around and especially in front of the antenna gives better results. [#21199071]

What is a biquad antenna, and why do so many DIY Wi‑Fi builders recommend it over popular 'Pringles can' designs?

A biquad is a simple DIY directional Wi‑Fi antenna that many builders favor because it gives a strong price/effort/effect ratio. Several posters recommended biquads, while canned “Pringles” designs were criticized as overhyped or dimensionally wrong for 2.4 GHz. One reply stated that the tube diameter should be about 82.7 mm internal for proper waveguide behavior, making casual snack-can builds suboptimal. [#21186615]
Generated by the language model.
%}