Short Dipole for 1.9 GHz

If I would like to design a printed version of an electrically short dipole antenna, is it possible?

I read some paper about the printed version of half-wave dipole and their co-planar feed line is contain the impedance around 50 ohms
because the radiation resistance of have-wave dipole is around 72 ohms.

In case of short dipole the radiation resistance very low.
Is that means no printed version of short dipole?

Hi!
I'm not understanding what you're asking. Can you give additional informations please?
What software are you using (HFSS, CST, others...)? What do you mean with "short" dipole, in terms of wavelength?

I presume "electrically short" means eletrically small, which is a clear term in antenna theory.

But I also don't understand what you are exactly asking. The input impedance of a short dipole will be capacitive in series with a small radiation resistance. There's no basic difference between printed and discrete wire design. In a printed dipole design, you may want to integrate a matching network on the PCB. Because the room is apparently limited, this should be imagined as a lumped network.

bepi004 said:

Hi!
I'm not understanding what you're asking. Can you give additional informations please?
What software are you using (HFSS, CST, others...)? What do you mean with "short" dipole, in terms of wavelength?

Click to expand...

Yes, in terms of wavelength.
For the half-wave, dipole has a radiation resistance around 80 ohms (for printed version) so it very easy to make a matching component.
In case of, short dipole (one-tenth wavelength) has a very small impedance at the terminal which I don't know how to make a matching.

- - - Updated - - -

FvM said:

I presume "electrically short" means eletrically small, which is a clear term in antenna theory.

But I also don't understand what you are exactly asking. The input impedance of a short dipole will be capacitive in series with a small radiation resistance. There's no basic difference between printed and discrete wire design. In a printed dipole design, you may want to integrate a matching network on the PCB. Because the room is apparently limited, this should be imagined as a lumped network.

Click to expand...

That meant I need lumped element to match the impedance, right?
However, I read in Jasik's book, he said that lumped element matching is not good for high frequency. Is my understanding correct?

However, I read in Jasik's book, he said that lumped element matching is not good for high frequency. Is my understanding correct?

Click to expand...

The statement is only relative correct. If the circuit dimensions are < λ/10, even transmission lines can be considered as lumped circuit. Recent RF components in small SMT package (e.g. 0402) can be used up to several GHz without problems. You don't have much realistic alternatives anyway.

A patch antenna on a high er substrate might be considered as well, but the velocity factor of 1/√er will be probably not sufficient. Or some kind of meandering structure achieving resonance without additional lumped components. In any case, an electrical small antenna will expose small bandwidth respectively high Q by nature.

FvM said:

The statement is only relative correct. If the circuit dimensions are < λ/10, even transmission lines can be considered as lumped circuit. Recent RF components in small SMT package (e.g. 0402) can be used up to several GHz without problems. You don't have much realistic alternatives anyway.

A patch antenna on a high er substrate might be considered as well, but the velocity factor of 1/√er will be probably not sufficient. Or some kind of meandering structure achieving resonance without additional lumped components. In any case, an electrical small antenna will expose small bandwidth respectively high Q by nature.

Click to expand...

Thank you for your answer, but are there the other ways to match an impedance for the electrically short dipole to 50 ohms system?

One-tenth wavelength dipole, 1.9 GHz on FR4 PCB. About 10 mm long antenna?
Yes it is possible to design and impedance match a such antenna.
Anyone can do it. Very few can do it and get reasonable performance. Good skills how to set up a VNA is recommended as short antennas and extreme impedances both requires good calibration to be able to get useable results.

are there the other ways to match an impedance

Click to expand...

Assume required matching circuit should cost less space on PCB then the antenna if it is somewhat balanced design? A few 0402 or 0201 inductors and capacitors is then all that you can fit in available space.
Make it simple and all you need is a 50 Ohm resistor for matching across antenna terminals. It would work just as good as a somewhat miscalculated design of a reactive network.

Do not know anything about this design concept except that it should have a small dipole antenna. Is it any reasons to prefer a dipole design over other alternative in this limited space?
Whole PCB size, is a bit bigger then 20 mm? For same antenna space will then a monopole perform much better and will be much easier to impedance match with discreet components. A monopole will probably also require fewer matching components for same result.
When it comes to very small compact PCB designs with printed antennas can even a loop antenna be to prefer from antenna efficiency view, as it can use else unused PCB space, such as edges around PCB and by this cost less PCB area and deliver better antenna efficiency.

As a reference, I did design a printed monopole 2.4 GHz, 2*9 mm occupied PCB area, relative wavelength a bit bigger then your antenna (relative wavelength) but still in the range of what I call a small antenna.
It was a real quickly designed antenna, easily done in 10 minutes and documented here: https://www.antune.net/demo/bluetooth/index.html
This antenna was easy for me to design because I have the tools and know how to use them.
I can design a dipole, fitting in same space, but it would result in a less good antenna. Vast of time in my opinion.

At 1.9 GHz it is nearly impossible to get 70 ohms for a half wave dipole unless you are using micro thread for rods. As diameter of radiator rods increases the antenna shortens and impedance goes down. I used 1.6 mm rods for a free space dipole @ 2 GHz and it shortens to about 90% and has a balanced center point feed close to 50 ohms. Embedding the same antenna on FR4 with other circuitry a 15 mm. away would shorten it even more.

I am not sure what you are considering 'short' in terms of wavelengths in your application. Realize on a PCB antenna layout there is a lot of opportunity of stray capacitance to shorten the antenna and everything in the 'system' (area within a wavelength around the antenna) is part of the antenna system and will impact results.

If you can live with something in the 8-10 mm total length you might try a bow-tie dipole configuration. It will give you good performance. You have to deal with balance input feed point.