Experience report with "active antenna"

[MaiKeeReis]

Hi there,

I would like to hear your thoughts an the following report.

As mentioned sometimes, I am using in one of my sites this cheap Ten-90-clone from Ali.

I assume, the internal setup is as follows:

So this would be a kind of hybrid between a vertical dipole design and a sleeve mono-pole.
It’s not flesh nor fish but it does its job. After some discussion about Lightning Arrestor, I remebered that my antenna cable shield has no real ground connection. So I decided to connect my house ground line directly with the SMA socket - at least until the ordered lightning arrestor will get delivered finally.

But then I had the crazy idea to not only add ground-line, but to also add a full ground-plane. For mechanical and location related reasons, I decided for a full disc of tinned 0.2 mm sheet steel (perhaps “tinplate” in English?), which is in fact an infinite number of radials. Antenna calculators said the “radials” shall be approx. λ/4 + 10%, so the disc has a diameter of 14.6 cm.

Yes, this many radials are not required.
Yes, the radials shall be bent down ~45 degree, otherwise the impedance is bad.
Yes, the thickness is low.
Yes, the position of the GP should be before the LNA, not after.

… but since this modification I have notably better performance! A slightly better SNR, more messages per plane, more range, more … everything. And slightly reduced CPU utilization.

The change is around the middle of the time lines, you see the strong signal dropout when I removed and reassembled the cable.

So, what’s your idea? Just improved “by chance” or “expected behavior”?

[geckoVN]

The antenna is simply a center fed Dipole - no mystery.

I use several passive versions of this antenna for reference purposes - they are ok, but not great (for obvious reasons).
Sticking an amp on a mediocre antenna doesn’t improve the radiation pattern, which is where the real benefit is.

Adding metal below the amp doesn’t make it part of the antenna, it’s part of the feedline.
How did you connect your “ground plane” to the coax?

[MaiKeeReis]

@geckoVN

The antenna is simply a center fed Dipole - no mystery.

Nah, not really “simply this way”. The feeding through the middle of downward dipole fork has also the sleeve effect. Only in a flat dimension, but I am quite sure it has.

@geckoVN

Adding metal below the amp doesn’t make it part of the antenna, it’s part of the feedline.
How did you connect your “ground plane” to the coax?

I just drilled a 6.3 mm hole into the disc and pushed the SMA plug through the hole. So the disc is connected with the outer part of the SMA plug and the cable’s shield.

Don’t know, if any free antenna simulation program could simulate this. The last time I calculated an antenna, I had to use a slide rule.

[member91040]

@MaiKeeReis

I just drilled a 6.3 mm hole into the disc and pushed the SMA plug through the hole. So the disc is connected with the outer part of the SMA plug and the cable’s shield.

Can we see the setup picture.

[abcd567]

@MaiKeeReis

Nah, not really “simply this way”. The feeding through the middle of downward dipole fork has also the sleeve effect. Only in a flat dimension, but I am quite sure it has

This antenna seems to be a 2-dimensional (stripline) version of “Coaxial Antenna”.

 

 

[geckoVN]

@MaiKeeReis

Don’t know, if any free antenna simulation program could simulate this.

Yes and no.
NEC will do it, but it’s a GIGO situation.
Without knowing the exact proporties of the ‘FR4’ (the original FR4 hasn’t been made in years) it’s impossible to create an accurate model.

[abcd567]

@geckoVN

NEC will do it, but it’s a GIGO situation.
Without knowing the exact proporties of the ‘FR4’ (the original FR4 hasn’t been made in years) it’s impossible to create an accurate model.

 

Which one is the best antenna?

The second side strip of PCB is ready now Next is to put pieces together and run the simulation. Full View: [image]   Closeup: [image]

 

Which one is the best antenna?

Either the simulation is grossly wrong OR If Simulation is OK, then the PCB antenna is rubbish Only the field test will reveal the truth   [image]   [image]  

 

[MaiKeeReis]

@member91040

Can we see the setup picture.

It’s in the attic of my barn, maybe I’ll be able to take a photo there on Sunday.

[MaiKeeReis]

@Which one is the best antenna?

Only the field test will reveal the truth

Yes, I agree.
Bit hard to see in the picture, but since adding the disc / plane in the evening of Feb.27th, I get a better Signal to Noise Ratio, compared to the earlier five days.
The distance between “Weakest” (blue) and “Noise” (green) has also been good at some earlier nights, but since the disc noise is always far below weakest in the night and less stronger than weakest in the days.

So I would say, this disc is a slight improvement for the Ten-90-clone. Perhaps it will only work well on the clone. Perhaps because by change it get placed at nearly one wavelength below the feeding point or similar. Perhaps because the disc is working as a reflector at - by chance - good distance. Who knows.

[MaiKeeReis]

@member91040

Can we see the setup picture.

Buttom up view:

I drilled two dirty holes for mounting it - recently removed mounting woodpiece.

Side view

you see the junction of SMA outside + yellow-green PG/GND cable + the disc.

[geckoVN]

Making a secondary ground to that point runs the risk of earth loops.

[MaiKeeReis]

Generally, you are right. This time it is the only earth connection. The Raspberry Pi has no Ethernet, just WiFi.

With the same idea you had in mind, I asked myself: “is it not the disc, is it only the GND which makes the better performance?”

So I removed the ground-plane disc and only connected the PG/GND cable. Happened at 13:10, you see the 5 minutes droput.

As we can see above: Only PG/GND cable is not the root cause of the positive effect. Removing the disc while keeping the ground line does strongly increase the noise, reduces the positions per second, increases the CPU load. See also this statistics: