Author Topic: USB WiFi Adapter Antenna  (Read 6411 times)

0 Members and 1 Guest are viewing this topic.

Offline bobcat2000Topic starter

  • Regular Contributor
  • *
  • Posts: 209
  • Country: us
USB WiFi Adapter Antenna
« on: May 02, 2021, 12:20:05 am »
Hi Guys,

I am looking at these USB WiFi adapters.  They seem to have 2 different type of antenna.  My understanding from my Physic class is that the piece of wire picks up the magnetic wave to generate electricity.  But these WiFi thingies are 2 ways.

It looks like the longer one has a bigger surface area.  Will the antenna with one piece of long wire better than the one with a small pads of copper? 

Thanks!



 

Offline ataradov

  • Super Contributor
  • ***
  • Posts: 11228
  • Country: us
    • Personal site
Re: USB WiFi Adapter Antenna
« Reply #1 on: May 02, 2021, 01:39:41 am »
The first one looks like some artist rendition, it is likely not actually that complicated on the inside. In any case it is hard to tell without either simulation or proper characterization what is better.

More complex antennas may have weird radiation patterns and performance would depend on the orientation.

Generally the bigger the antenna, the better it would be provided it is still tuned to the target band.
Alex
 

Offline bobcat2000Topic starter

  • Regular Contributor
  • *
  • Posts: 209
  • Country: us
Re: USB WiFi Adapter Antenna
« Reply #2 on: May 02, 2021, 02:02:31 am »
Something likes this.  They printed the copper on the board.

When you say "Tune", you mean size doesn't matter?  I just assume bigger is better.


 

Offline ataradov

  • Super Contributor
  • ***
  • Posts: 11228
  • Country: us
    • Personal site
Re: USB WiFi Adapter Antenna
« Reply #3 on: May 02, 2021, 03:06:18 am »
This is functionally no different than wire. It may be a bit worse, but not much. It is easier to manufacture, I guess.

The length of the wire is not arbitrary. It is calculated according to the frequency band (2.4 GHz or 5 GHz in this case). And when selecting the antenna it is important to match the used frequency band.

Note that the length of the squiggly line on the PCB is shorter than equivalent coil part on the wire one.  The coil or squiggly trace does not add anything, it is just a way to have antenna to be mechanically shorter while having the same electrical length.
« Last Edit: May 02, 2021, 05:09:02 pm by ataradov »
Alex
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16544
  • Country: us
  • DavidH
Re: USB WiFi Adapter Antenna
« Reply #4 on: May 02, 2021, 06:56:14 pm »
Given the frequency and antenna length, I do not think that is a center-loaded dipole.  More likely it is the third type of colinear array described here; two or or more dipoles are placed in series with phase matching sections between them.  This gives greater gain and capture area by flattening the radiation pattern.
 
The following users thanked this post: cdev

Offline bobcat2000Topic starter

  • Regular Contributor
  • *
  • Posts: 209
  • Country: us
Re: USB WiFi Adapter Antenna
« Reply #5 on: May 02, 2021, 10:10:15 pm »
I found the teardown for the printed antenna.

It looks exactly like the drawing.
I am sure it costs less to make compared to the 1 piece metal rod antenna.  They can melt the metal rod to make 10 or 20 of these traces.

It looks like they arrange the metal trace just like the Collinear antenna array.

Interesting!


 

Online E Kafeman

  • Regular Contributor
  • *
  • Posts: 107
  • Country: se
    • AnTune VNA software
Re: USB WiFi Adapter Antenna
« Reply #6 on: May 03, 2021, 09:49:15 am »
It is not a single antenna structure.
PCB consist of two separate dual band dipole antennas which can be seen as there are two separate coaxial cables
Each cable have separate RF input at WiFi radio chip.
Same PCB structure can also be seen as an antenna array but then are both dipoles feed from a single source via a single coaxial cable.

The both antenna are dual band dipoles. A typical dipole antenna have two symmetrical arms with a length length related to lambda/4.
In this case can we see that there are two arms of different lengths at each side of the dipole.
WiFi is most commonly operating at two frequency bands, 2.4 and 5 GHz, so the both arm-lengths are probably tuned for these bands.

Lambda is wavelength but as speed of light is slower propagating in PCB material then in free space must arm length be compensated for that.
If each arm is measured with a ruler at this PCB will it be found that it is up to 50% shorter then expected arm length in free space (~32mm).
« Last Edit: May 03, 2021, 10:15:46 am by E Kafeman »
No signature
 

Offline bobcat2000Topic starter

  • Regular Contributor
  • *
  • Posts: 209
  • Country: us
Re: USB WiFi Adapter Antenna
« Reply #7 on: May 03, 2021, 09:13:39 pm »
So, what is better?  Buy one with a piece of metal rob antenna?  or buy one with a printed antenna?
 

Offline ataradov

  • Super Contributor
  • ***
  • Posts: 11228
  • Country: us
    • Personal site
Re: USB WiFi Adapter Antenna
« Reply #8 on: May 03, 2021, 09:22:06 pm »
It does not matter most of the time. You are overthinking antennas. There is more dependency in orientation if the antennas than difference between those two types.
Alex
 

Offline bobcat2000Topic starter

  • Regular Contributor
  • *
  • Posts: 209
  • Country: us
Re: USB WiFi Adapter Antenna
« Reply #9 on: May 03, 2021, 09:43:13 pm »
So, do what Al Bundy did?

"...Assume Fox viewing position..."

 

Offline ataradov

  • Super Contributor
  • ***
  • Posts: 11228
  • Country: us
    • Personal site
Re: USB WiFi Adapter Antenna
« Reply #10 on: May 03, 2021, 10:47:34 pm »
Basically, yes. Avoid pointing pointy end of the antennas at each other, radiation pattern of a dipole antenna is a doughnut with the hole going though the antenna length. Basically vertically pointed antenna would have the best performance for common household uses. But if you need it to get better reception on the second floor with the router on the first floor, for example, then horizontal orientation may be better.

But generally, if you have to rely on the orientation for things to work, you are in a marginal scenario either way.

And the most common reason why there are two of them is for antenna diversity - attempt to cover holes in the radiation pattern of one antenna with another antenna. Most routers today have multiple antenna connectors on the outside of the case for the same reason.
« Last Edit: May 03, 2021, 10:50:51 pm by ataradov »
Alex
 

Online E Kafeman

  • Regular Contributor
  • *
  • Posts: 107
  • Country: se
    • AnTune VNA software
Re: USB WiFi Adapter Antenna
« Reply #11 on: May 03, 2021, 11:06:16 pm »
Quote
Buy one with a piece of metal rob antenna?  or buy one with a printed antenna?
Bananas or apples? A steel wire rod antenna is a single antenna structure. At actual PCB antenna you have two different antennas. A such structure is not replaceable with one steel rod structure.
I dislike this PCB version because PCB dipole design may be ok but the both antenna structures are a bit too close to each other and with same polarization which can reduce coverage area and stability for 802.11n and above as it will reduce diversity gain. How noticeable this effect is depend on a lot of factors including type of room and type house.
Best antenna is an antenna in free space, not hidden behind cables or placed close to a wall behind a computer.
If multiple antennas are supported by WiFi radio, place the antennas at decent distance from each other and with different orientation for best data speed and coverage distance.

A PCB antenna will always have a minor disadvantage relative a wire antenna due to losses in PCB material but it allows for better control over antenna tuning at production and is because of this often better tuned.
It exist antenna-rubbish at market, not designed at all for what promises.They are always "no name".

I have a load of old router WiFi antennas. They do in most cases contain a simple endfeed dualband dipole. They are a good choice to reuse for USB WiFi adapters. as they usually are well tuned and have high efficiency.
Very old router antennas are probably single band 2.4 GHz antennas which not is ideal to use if 5GHz band is going to be used but for 2.4GHz are they still fine and 2.4GHz allows often for better  coverage then 5GHz as waves at 5GHz are easier absorbed.
If long distance coverage is needed outdoor can a directive antenna be a better choice. Indoor in a very reflective environment have directivity less value.
When I use USB WiFi adapters do I often use a short USB extension cable to allow USB adapter to be placed at a more free location, and by then also a more free antenna location for best coverage.

Reduced speed and reduced coverage is not always a problem with radio or antenna for cheap USB WiFi adapters.
These adapters have a tendency to behave better if placed in free space which allows for better cooling, compared to if placed at back of a PC.
USB enclosure is often a bit to small to allow for a good air ventilation and heat dissipation.  If chip is overheated will speed be reduced or chip will die until it have cooled down somewhat.
No signature
 

Offline amishasingh

  • Contributor
  • !
  • Posts: 17
  • Country: ca
Re: USB WiFi Adapter Antenna
« Reply #12 on: May 17, 2021, 08:07:42 am »
Better to choose the one with the bigger antenna, that they will ensure to reach the suitable bands faster and better
Amisha
 

Offline cdev

  • Super Contributor
  • ***
  • !
  • Posts: 7350
  • Country: 00
Re: USB WiFi Adapter Antenna
« Reply #13 on: May 25, 2021, 10:53:44 am »
Amisha, not necessarily!

Such an antenna is going to have a pattern like a squished donut.

It needs to be perfectly vertical to be helpful. So the donut gets squished out to the horizon and is aligned with it. Otherwise the extra gain antenna will be very likely be less helpful than the reference lower gain monopole/rubber duckie as pictured  It all comes down to the packaging and the expected application. A longer antenna will have a more lumpy pattern with lobes most likely shaped like a butterfly's wings. They may not be where you need them. The col linear is a great antenna when its able to be kept perfectly vertical. Then it will concentrate more signal in a perfect belt around itself. at the horizon where the receiver is most likely to be.



I found the teardown for the printed antenna.

It looks exactly like the drawing.
I am sure it costs less to make compared to the 1 piece metal rod antenna.  They can melt the metal rod to make 10 or 20 of these traces.

It looks like they arrange the metal trace just like the Collinear antenna array.

Interesting!
« Last Edit: May 25, 2021, 11:00:25 am by cdev »
"What the large print giveth, the small print taketh away."
 

Offline NiHaoMike

  • Super Contributor
  • ***
  • Posts: 8972
  • Country: us
  • "Don't turn it on - Take it apart!"
    • Facebook Page
Re: USB WiFi Adapter Antenna
« Reply #14 on: May 26, 2021, 01:55:30 am »
In my experience, the best compact antenna for a wireless adapter is a biquad, lots of gain for the size. Put it on a stand that allows fine adjustments to get the best signal.
Cryptocurrency has taught me to love math and at the same time be baffled by it.

Cryptocurrency lesson 0: Altcoins and Bitcoin are not the same thing.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 16544
  • Country: us
  • DavidH
Re: USB WiFi Adapter Antenna
« Reply #15 on: May 26, 2021, 09:14:37 pm »
Such an antenna is going to have a pattern like a squished donut.

It needs to be perfectly vertical to be helpful.

I agree when the environment supports line-of-sight and is free of obstructions, but indoor WiFi applications have so much multipath that omni-directional antenna orientation is largely irrelevant.
 

Offline cdev

  • Super Contributor
  • ***
  • !
  • Posts: 7350
  • Country: 00
Re: USB WiFi Adapter Antenna
« Reply #16 on: July 02, 2021, 12:09:35 am »
I think both designs are attempts to be collinear antennas and likely the gains are roughly similar and perpendicular to the element.

There are dozens of good antennas for wifi. Chances are most of them would work okay for a short distance situation.

I agree that a biquad+ reflector is an excellent choice for a fairly decent directional antenna without a lot of work.
"What the large print giveth, the small print taketh away."
 

Online E Kafeman

  • Regular Contributor
  • *
  • Posts: 107
  • Country: se
    • AnTune VNA software
Re: USB WiFi Adapter Antenna
« Reply #17 on: July 02, 2021, 04:57:56 am »
I think both designs are attempts to be collinear antennas

Hm, hm and hm again. I am feeling I am repeating myself, trying to describe a very basic antenna structure in a simple way that all that knows anything about antennas should be able to understand.
Why not read previous antenna comments and look at available pictures before wrongly deciding which type of antenna it is? It is just too simple to identify these antennas so wrong conclusion is a bit worrying.

First antenna design, antena-1.png, is almost too easy to identify from its PCB pattern and by then exclude it as possible collinear design.
This antenna design is even more simple to identify as it also is shown in #msg3561532 that it have dual coaxial cable feeds.

It can be a good first clue to identify this antenna, why should a low cost none collinear antennas printed on a PCB have dual coaxial cables?

Antenna pattern for this antenna is very simple and very popular type of dual band dipole antenna.

Dipole antenna is basically an antenna with two arms (di-poles), commonly 0.25 lambda length for each arm.
At first picture can it be seen a second pair of arms at each dipole with slightly different length, which allows for resonance at both 2.4 GHz and 5 GHz for a single antenna structure.
This structure is repeated, there are two double band dipoles at PCB.
Collinear double band antennas at 2.4GHz and 5 GHz is not possible or very complex and do not need two coaxial cables, why we by now can be very sure that it not is a collinear design.

Hypothetical could this be an antenna array for 2.4 and 5GHz, consisting of the both visible antenna structures but as we know that most modern WiFi routers and as in this case an USB WiFi-adapter, need MIMO antennas to achieve its bandwidth according to 802.11n and above is it no hard guess that the both coaxial cables feeds separate transceiver ports. We can even read "AC" in the 2:nd picture of this antenna, if still in doubt.

Antenna shown in picture antena-2 is an wire antenna and antena-3 is a PCB pattern but both antennas are basically similar type of antenna. It is as best a single stage of a collinear antenna.

As what I now hope can be easily understandable, these both pictures, antena-1 and antena-2 are two very different animals.
Picture antena-1 is showing two antennas, each of these both antenna are designed as a double frequency band antenna. Each antenna have its own coaxial cable feed.
Picture antena-2 is just a single antenna and it is designed for just a single frequency band.

« Last Edit: July 02, 2021, 05:07:06 am by E Kafeman »
No signature
 

Offline cdev

  • Super Contributor
  • ***
  • !
  • Posts: 7350
  • Country: 00
Re: USB WiFi Adapter Antenna
« Reply #18 on: July 04, 2021, 10:47:11 pm »
I have a collinear design Ive built now, twice, it uses a short delay segment, its been around for at least 20 yrs, I first saw it around 1999, it uses 1 mm diameter copper wire and it is end fed. The wire is attached to an bulkhead mount N-connector and it has a respectable 5 or 6 db of gain out to the horizon. It resonates exactly at 2400 MHz. I have mine in a thin piece of PVC pipe.

There is another collinear thats very similar and it looks more like the picture 2. It is usually embedded in a black plastic radome and it has the stub section at the bottom. The coil is not a loading coil, its to introduce a time delay. e can increase the gain of this type of antenna substantially by adding more sections but one reaches a point of diminishing returns at a fairly low number of sections. You dont get much more gain from more than three sections.

https://martybugs.net/wireless/collinear.cgi
« Last Edit: July 04, 2021, 10:54:05 pm by cdev »
"What the large print giveth, the small print taketh away."
 

Online E Kafeman

  • Regular Contributor
  • *
  • Posts: 107
  • Country: se
    • AnTune VNA software
Re: USB WiFi Adapter Antenna
« Reply #19 on: July 05, 2021, 06:26:18 am »
>its been around for at least 20 yrs,
Collinear antenna design have been around in many versions since long, much longer then 20 years.
It was patented 1924 by C. S. Franklin, almost 100 years ago. and this antenna type is also known as a Franklin antenna.

>and it has a respectable 5 or 6 db of gain
No it haven't. dBi maybe. dB is relative to something you not did specify.

What gain you achieve is roughly calculated as it is a number of stacked halfwave dipoles.
Antenna type is mostly either centerfeed or endfeed.

>It is usually embedded in a black plastic radome
What enclosure an antenna is placed within is not an antenna property. Neither color, but yes many plastic tubers contains an antenna of some kind.

An alternative and sometimes better describing name of this antenna type could be co-phasing antenna.

>The coil is not a loading coil,
If it is a coil, it is a coil even if it is placed within an antenna structure and load? No it does not load anything, it adds serial reactance.
The phasing element can be designed in many ways, open or short stub, two coils winded in opposite directions close to each other is common, and the phasing pattern can in many cases be performed at PCB as well as using heavy metal designs.

> its to introduce a time delay.
At antenna-language do we call section between the halfwave elements for phasing element, not timedelay as that is something else.

>You dont get much more gain from more than three sections.
I have no such problems with my antennas and there are a lot of other designs that not have your limitation.
This antenna structure is nothing else then a kind of antenna array and stacking a rather high number of element is possible.
As best is gain increased with 3 dB for each doubling of number of sections as for any similar array, but poor design can of course reduce that number.
In most cases is antenna efficiency reduced for each added section but that is something else then gain.


This is a (Franklin) collinear antenna in 14 sections. Measured gain above 15 dBi. It is top-terminated with a resistor.Your link to Martybugs antenna, is not a working antenna array and whatever it is is it a real poor working co-phased antenna.
You can find descriptions of co-phasing in Balanis ANTENNA THEORY ANALYSIS AND DESIGN but collinear antenna as a defined structure, its theory is described relative simple at page 134 and forward in Transmission Lines Antennas and Waveguides. The book is first printed 1945 but still relevant and where following design is explained in more detail.




« Last Edit: July 05, 2021, 07:58:03 am by E Kafeman »
No signature
 

Offline cdev

  • Super Contributor
  • ***
  • !
  • Posts: 7350
  • Country: 00
Re: USB WiFi Adapter Antenna
« Reply #20 on: July 18, 2021, 10:19:09 pm »
If it has dual cables thats likely for diversity. Could you scan or make a better photo of the PCB antenna?


You can use a single rectangular 2400 Mhz panel antenna into a diversity antenna with gain by feeding the panel at two different places to utilize different modes on both receive and transmit. That is a very good antenna for a wireless AP with an angle of radiation of around 60 degrees. Well suited for a backyard or neighborhood. It needs a reflector behind it and needs to be rigid and held flat over its reflector in order to deliver consistant performance. If you fed the square at two corners separated by 90 degrees you will end up with a dual polarization antenna with  pattern that is consistent in both polarization plnes, basically, so its super useful that way.
you could use a coaxial stub to give you a circular polarization low gain directional antenna for its single frequency.

With the collinear, You know what I meant. Here is what I found with my martybugs style 3 element "collinear" made almost exactly like the one in the article. It does resonate at slightly above 2400 Mhz, good.

When used as a household wifi antenna it works well, as expected with a pattern thats not super directional but likely does have a bit of gain over the dipoles that came with the wireless AP. Works well on receive, picking up maybe 15 other APs.

SWR around 1.2.

Have no way to test the gain accurately.

For me this was a success. If I ever have a need for a 2400 Mhz omni again It will be used.

Currently don't really need it, my current AP has built in antenna. It has the pads for two u.fl connectors, that remove he internal antennas when n antenna is plugged in but it might be ugly. So I am just leaving it the way it is for now. Its a strong APas far as power goes, even deflecting my passive FS meter made with a 1n34 type diode and capacitor.


I will probably use this design again/ adapted for size, for other frequencies. (ADSB?, it seems ideal for that.)


If you want the 2400 panel dimensions I could measure it or scan it for you. That is even easier to make than the colinear. Just need some copper sheeting and tin snips.
« Last Edit: July 18, 2021, 10:46:04 pm by cdev »
"What the large print giveth, the small print taketh away."
 

Online E Kafeman

  • Regular Contributor
  • *
  • Posts: 107
  • Country: se
    • AnTune VNA software
Re: USB WiFi Adapter Antenna
« Reply #21 on: July 19, 2021, 03:04:11 am »
>It does resonate at slightly above 2400 Mhz, good.
I very much dislike when someone intentional is spreading false information in opposite to my already above written information.Both of us can not be right, one of us is lying.
I have written that it is a useless antenna design and you is writing opposite.

I state that there is no resonance related to wifi for this design.
Even with very simple antenna knowledge is it easy to understand that it is a rubbish random wire, not related to antenna design.
>Have no way to test the gain accurately. No excuse at all, especially as if you says this is an good antenna design, as I then must be wrong.
Basic checking if it is an antenna at all is very easy done if you have bent actual wire by yourself. You must have done some kind of check? VSWR value, from where have you got that if you not have done any checking?


One real simple method to check if antenna is working as intended:

1. Connect the antenna to your wifi adapter or router and use free software to read RSSI if such software not is already included.
2. Place hand at antenna above 1:st coil and there will be almost no reduction in performance, less then 2 dB, as these parts not add anything to antenna performance more then any random wire would do.
3. It proves that these sections are useless and just to minor degree affect antenna efficiency and gain.
4. Remaining antenna bottom section. is a 61 mm straight wire. No clever calculations needed gives that what now remains from above simple test is mainly an endfeed 0.5 lambda antenna at 2.4 GHz connected to a 50 Ohm port.

Just by knowing how long a resonant length is can it be proved that this design NOT is resonating at 2.4 GHz. Rather simple math skills needed.

So what is performance for this antenna?
It have obviously negative gain due to its high reflection losses.
VSWR is not even close to what normally is acceptable.


>For me this was a success.

Lets measure your design. We already expect it to behave similar to an endfeed 0.5 lambda antenna but an simple check to verify VSWR:

As can be seen, even if scaling dimensions, there is no better tuned frequency, nothing to tune for improved result.It is nothing else then a random bent wire, resulting in an unusual poor antenna.


This Smith chart for the actual antenna confirms that this design have an impedance similar to a 0.5 lambda monopole.
Just by looking at actual design could that be predicted.
Endfeeding a 0.5 lambda wire from an 50 Ohm port will never be impedance matched.It is not even close to a resonance at 50 Ohm, center point at above Smith chart.So mr cdev, have you measured a VSWR below 1.2:1 for a design for which it is impossible?  Did you misread? It was maybe 12:1  or 120:1 as that is closer to a possible truth?One of us must be wrong. You or me?
« Last Edit: July 19, 2021, 03:29:33 am by E Kafeman »
No signature
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf