Radio Connections to Antennas on RT-AC66 Router

[Just Checking]

Which radio band is connected to each of the antennas on the Asus RT-AC66 Router? I looked at the board pictures in the following article but I am not able to tell.

http://www.smallnetbuilder.com/wire...-band-wireless-ac1750-gigabit-router-reviewed

The reason I want to know is because I am testing out different antennas and I find that an aftermarket antenna (TP-Link 8db High Gain whip antenna) gives a good improvement on the 2.4GHz band when in the center position. I cannot find any aftermarket antennas that perform better than the stock antennas for the 5GHz band. I just want to determine if my experimental results actually correlate to the wireless radio band for each antenna.

Hopefully, someone who has opened up these routers or has access to a circuit diagram can tell me.

Thanks

 

[Deleted member 22229]

I believe all three antennas work with both bands at the same time as long as both radios are enabled.

 

[Just Checking]

I believe all three antennas work with both bands at the same time as long as both radios are enabled.

Thanks for the reply.

Are you sure that the antennas function on two frequencies simultaneously?

What I know about radios would indicate that broadcasting and receiving two different frequencies simultaneously is difficult. I'm not doubting that it can be done, just wondering why they would do that when there seem to be three separate radios and antennas on the board (I cannot tell from the pictures). It seems like it adds a lot of complexity with no apparent gain.

If each of the antenna is broadcasting and receiving on both frequencies simultaneously, that would account for the poor performance of the 5GHz band vs. the 2.4GHz band. The antenna can only be optimized for one frequency. It would not be possible to get optimum performance on both frequencies at the same time except in certain conditions.

It also raises the question of why the router manufacturers use three antennas if they all do the same function. I can understand two antenna for a single frequency because you can get increased signal strength and better S/N with two antennas (if properly aligned and with the correct spacing) without having extra length antennas. It is much harder to optimize signal performance with three antennas. To do it simultaneously with three antennas and two different frequencies is very much harder. The problem is simplified if you make the frequencies exactly 2x or 4X each other (2.4GHz and 4.8GHz for example). I thought that the frequencies were actually 2.48GHz and 5.12GHz which are not exactly double. Again, this would lead to sub-optimal antenna performance.

It would make much more sense if there were two radios on the board. Each radio broadcasting and receiving on only one frequency. One radio could go to two antennas, the other radio to only one. Asus-Merlin firmware shows a separate temperature reading for the 2.4GHz radio and the 5GHz radio band and the firmware allows you to adjust the power to the 2.4GHz radio separate from the 5GHz radio indicating physically separate radios. Why would you connect two separate radios to one antenna when there are three antennas? There are no cost savings to offset the decreased performance. It even makes some sense to have three separate radios on the board with each one going to its own separate antenna and two of the radios going to one frequency (5GHz) which requires more power output to achieve high signal strength at distance.

There always seem to be more questions coming from each answer.

 

[jim769]

You do bring up some valid points but thats not how it works with duel band routers. First you can use one antenna for more than one frequency ham ops do it all the time in fact there are multi band antennas for almost all radio bands.

That said there is a compromise for both, 2.4 Ghz will win every time its the nature of the frequency and how it propagates through the local environment.

 

[Just Checking]

You do bring up some valid points but thats not how it works with duel band routers. First you can use one antenna for more than one frequency ham ops do it all the time in fact there are multi band antennas for almost all radio bands.

That said there is a compromise for both, 2.4 Ghz will win every time its the nature of the frequency and how it propagates through the local environment.

Yes, ham radio is where I draw most of my antenna knowledge from. In ham radio, multi-band antennas are not really a single antenna but a combo of two or more elements. Also, the multi-band portion of the system is often a sub-band or a multiple of the frequency of the main band. I don't know any ham radio operators who have multiple radios on multiple bands then put them to multiple iterations of the same antenna. They have multiple radios going to an antenna optimized for that frequency range. They tune their antennas all the time for optimum S/N when they switch frequencies. Another point is that the tuners on the ham radios are extremely sophisticated (expensive) and would tremendous overkill for a radio that was just set on one frequency (with a few sub-channels).

If they are out in the field, ham ops can throw up an antenna that is not optimized. The military used to do it all the time too before satellite uplinks. Why do it in a stationary piece of equipment with set frequencies when you have the designers, engineers, and capacity to optimize performance on both bands at no extra cost over what they have already?

Can someone tell me if there are two, or three radios on the board? From the pictures, there are three separate contact points going the three separate antennas. There are no wire connections going to all three antennas. It pictures suggest that there may be two radios of one type and one radio of another frequency separated on the edges of the board. I wish I could see the front of the board without heatsinks/shield. If the connections between the radios are in traces on the board, why do they have such long wires coming out of the board? Why not put the conductive traces on the board up to the point where the antenna connectors are? It is a multi-layer board so that wouldn't be a problem and makes more sense to do that then to run the traces in the board to the same contact points from the radio. Also, the cables coming out of the board are not the same length (that I can see). If those cables are part of the antenna (which they would be), then they should be the same length to ensure reasonable signal. Wouldn't having that cable/antenna length difference introduce another level of complexity for the radio tuners?

The only way I could explain this is if the frequency was so high that the wavelength is so short as to make those differences in antenna length insignificant to the radio signal. I need to run some calculations on that one but, my empirical experiments seem to indicate that changes in antenna length which are as gross as I see from those cables would make a significant difference (for the 2.4GHz frequency).

From a design standpoint, why go through the trouble of making a very sophisticated (expensive) radio tuner with lousy antennas, when you can spend no more money and make good antennas that require less sophisticated (cheaper) radios. Separate antennas for each radio (or a pair for the 5GHz frequency) should be the way to go. Performance would be better and costs would be lower. That is a no brainer to me. (Yes, I do work in Product Design/Development and know the pressures put on designers and engineers by management that doesn't have a clue. Still, money talks. This seems to be a loud and clear voice.) I have to be missing something if this is that obvious.

I agree that 2.4GHz has better ability to propagate through difficult medium than 5GHz. Longer wavelengths always win. That is why ham radios can reach "around the world" (with the right bounce). 900MHz phones have better signal strength at a distance than 2.4GHz or 5GHz. That is also true. If we were only concerned about propagation strength through particular media, we would use long wavelength radio waves. Data transmission rates would be miniscule but we would have great signal.

I have experimental evidence that changing the antenna design and size from the stock antennas can have significant positive and negative impact on both signal strength and uplink/downlink data transmission rates for the 2.4GHz frequency. I did tests replacing one, two, or all three antennas with different antennas and types of antennas. If either the antenna length was not important, or all the antennas are connected together, then it would be predicted that putting only one different antenna on vs changing all the antennas would make no difference. That is not true (for the 2.4GHz frequency). My tests indicated that I could make a difference in transmitted signal strength by changing out the number and position of the antenna. Unfortunately, I really could only make a minimal impact on the 5GHz radio frequency signal by changing out the antenna(s). I got much more change from changing the orientation of the transmitter to receiver than I ever got by changing the antenna.

Is there a good radio engineer on this forum who could explain this?

 

[Nerre]

It also raises the question of why the router manufacturers use three antennas if they all do the same function.

On routers today the main purpose is MIMO
http://en.wikipedia.org/wiki/MIMO

But more antennas can also be used for diversity
http://en.wikipedia.org/wiki/Antenna_diversity

 

[Just Checking]

On routers today the main purpose is MIMO
http://en.wikipedia.org/wiki/MIMO

But more antennas can also be used for diversity
http://en.wikipedia.org/wiki/Antenna_diversity

Many Thanks! I am paying too little attention to the reflection and partial polarization of the signal which appears to be a major consideration with antenna design in a cluttered environment like a building. The need for Antenna Diversity with both radio bands going to multiple antennas makes sense to me now. It is pretty obvious when presented to me. I must be getting senile in my old age.

I started a consult on an radio antenna website where the people have the background and ability to run design calculations. Hopefully, they can provide me with some avenues for improvements in antenna design to try out. Just doing quick empirical studies with the antennas at different angles (horizontal, vertical, and 45 degree angle) is insufficient. I have to find someone who can model this system for me.

 

[L&LD]

Many Thanks! I am paying too little attention to the reflection and partial polarization of the signal which appears to be a major consideration with antenna design in a cluttered environment like a building. The need for Antenna Diversity with both radio bands going to multiple antennas makes sense to me now. It is pretty obvious when presented to me. I must be getting senile in my old age.

I started a consult on an radio antenna website where the people have the background and ability to run design calculations. Hopefully, they can provide me with some avenues for improvements in antenna design to try out. Just doing quick empirical studies with the antennas at different angles (horizontal, vertical, and 45 degree angle) is insufficient. I have to find someone who can model this system for me.

Would be great to share this information when/if someone responds.

Thanking you in advance!

 

[thiggins]

The RT-AC66U is a simultaneous dual-band AC1750 class router. To get the maximum link rates of 450 Mbps in 2.4 GHz and 1300 Mbps in 5 GHz, there must be (at least) three spatial streams, which requires (at least) three Tx / Rx chains.

 

[KGB7]

Just like ac68, it uses all 3 antennas simultaneously on both bands. You can test it your self by disconnecting 2 antennas and only use one antenna at a time in each connector.

 

[TonyH]

Hi,
There is antenna design software available which you can run on a PC.
Antenna has two basic kind. Isopole(vertical) and dipole(horizontal) Basic Isopole is 1/4 Lambda(wave length) long. Dipole is 1/2 Lamda long. Vertical antenna is dominantly vertically polarized, horizontal dipole antenna is horizontally polarized. They have different radiation pattern/angle. Antennas are all Lump sum of L,C values represented in Z(impedance) TX and antenna has to be Z matched at feed point(~50 Ohm for vertical, ~70 Ohm for dipole) The better match the more erp. Degree of match is expressed as value of SWR. Perfect match being 1.0(never happens in real world) Antenna macher or tuner is fooling the TX as tho it is feeding perfectly matched antenna. If you want to test some thing like what you are trying, at least you should have something to measure erp and plot radiation pattern. What tool(s) are you using for this?

First we are dealing with 2.4 or 5.0GHz range; these are SHF, EHF range frequencies wth quite short wave length. Looked at internal antennas? See how small they are? Did you happened to cut up external antenna to see what's in it? I spent more than half a century
as a HAM and RF data/telecomm. EE working in the field. Antenna engineering is very complex, just like acoustic engineering(see how they design audio speakers) Based on very basic theory sky is the limit learning. Can you tell why basic vertical antenna is 1/4 wave long, horizontal dipole is 1/2 wave long? Antenna size, gain, radiation pattern/angle interact very much. I'd say 100% perfect antenna is impossible to produce. Also bottom line comes into play. How much
to spend on an antenna kit as part of a router. Good luck.

 

[faunus]

Which radio band is connected to each of the antennas on the Asus RT-AC66 Router? I looked at the board pictures in the following article but I am not able to tell.

http://www.smallnetbuilder.com/wire...-band-wireless-ac1750-gigabit-router-reviewed

The reason I want to know is because I am testing out different antennas and I find that an aftermarket antenna (TP-Link 8db High Gain whip antenna) gives a good improvement on the 2.4GHz band when in the center position. I cannot find any aftermarket antennas that perform better than the stock antennas for the 5GHz band. I just want to determine if my experimental results actually correlate to the wireless radio band for each antenna.

Hopefully, someone who has opened up these routers or has access to a circuit diagram can tell me.

Thanks

Actually, I am running two AC-68 routers with different antennas. I have attached an outside 2.4ghz antenna (http://www.radiolabs.com/products/antennas/2.4gig/high-gain-wifi-antenna.php) to the centre antenna connector on the router. This gives me a comfortable 75 meter outside range, with good signal strength on 2.4Ghz. 5Ghz is fine within the building (22 by 60 feet two floors and steel clad). For a building that is 500 feet away, where I have another AC-68 as a repeater, I use a tight beam antenna (http://www.radiolabs.com/products/antennas/2.4gig/backfire.php) to pick up a good signal from the outdoor antenna on the other building.

I have set the power on transmission to a max of 80mW. Transmission rates vary but I get a regular 200Mbps+ on 2.4Ghz and over 400Mbps on 5Gghz.

Sadly I have no AC cards to check out the AC speed. Temperatures in the units are normal.

Considering that for over 10 years the best I could get was 54Mbps between buildings I am quite happy with my new setup, which manages to provide reliable wifi between three buildings. I do not use the Asus firmware, it is terribly unstable and unreliable. I use Merlin's stable beta4 build which is just fabulous. The system is rock solid.

With the total cost of the system (two routers/two antennas) at about $700 I think I have a bargain!

 

[Just Checking]

Hi,
There is antenna design software available which you can run on a PC.
Antenna has two basic kind. Isopole(vertical) and dipole(horizontal) Basic Isopole is 1/4 Lambda(wave length) long. Dipole is 1/2 Lamda long. Vertical antenna is dominantly vertically polarized, horizontal dipole antenna is horizontally polarized. They have different radiation pattern/angle. Antennas are all Lump sum of L,C values represented in Z(impedance) TX and antenna has to be Z matched at feed point(~50 Ohm for vertical, ~70 Ohm for dipole) The better match the more erp. Degree of match is expressed as value of SWR. Perfect match being 1.0(never happens in real world) Antenna macher or tuner is fooling the TX as tho it is feeding perfectly matched antenna. If you want to test some thing like what you are trying, at least you should have something to measure erp and plot radiation pattern. What tool(s) are you using for this?

First we are dealing with 2.4 or 5.0GHz range; these are SHF, EHF range frequencies wth quite short wave length. Looked at internal antennas? See how small they are? Did you happened to cut up external antenna to see what's in it? I spent more than half a century
as a HAM and RF data/telecomm. EE working in the field. Antenna engineering is very complex, just like acoustic engineering(see how they design audio speakers) Based on very basic theory sky is the limit learning. Can you tell why basic vertical antenna is 1/4 wave long, horizontal dipole is 1/2 wave long? Antenna size, gain, radiation pattern/angle interact very much. I'd say 100% perfect antenna is impossible to produce. Also bottom line comes into play. How much
to spend on an antenna kit as part of a router. Good luck.

Thanks for this post. Lots of excellent points here. I don't have access to equipment for measuring signal polarization, and only basic things like software on other routers or laptops with wireless receivers to measure signal strength and properties. The best I can do is to try to make as precise measurements as possible while controlling as many variables like location, distance from the source, and the transmitter and receiver. My main goal is to try to optimize wireless performance for the networks I have in a building with multiple tenants. Partly, it is to increase my knowledge of the principles used to optimize the performance so that I can reproduce that at other locations. As my original problem statement suggested, I am more interested in the 5GHz frequency than 2.4GHz. I have lots of questions but here are a few that are the most pressing.

1. The wavelength for a2.4GHz frequency is 12.491cm. 5GHz is 5.996cm. The quarter wave is 3.122cm for 2.4GHz and 6.244cm for the half wave. Quarter wave for the 5GHz band is 1.50cm and half wave is 3.0cm. How do the router mfg. prevent antenna to antenna interference with a three antenna system and the spacing that is on the router?

2. I did some more tests with multiple different antenna lengths (whip antennas - a couple of which I took the covers off to see how they were built). The specs say that the antennas are 50 ohm impedance. I do not have access to 70 ohm impedance antennas. I don't have the specs on the stock antennas but I know they are all the same and do not vary the impedance. If impedance makes that much of a difference, why not provide different impedance antennas stock (either texture, or marked differently to let people know which is which)?

3.I started orienting the antennas vertically, horizontally, at 45 degree angles, and finally one antenna at each of the three angles. I use a second Asus RT-AC66R router to monitor changes in the received signal strength at a fixed distance and orientation from the test router/radio source (0 degrees +/- 3 degrees). I do not have a easy way to plot vs. angle and position. What I do see is that having a vertical antenna combined with a horizontal antenna made the most dramatic increase in signal strength and sensitivity. I orient the antenna with one of the side antennas at 45 degree angle away from the center antenna which is vertical and the other side antenna in the horizonal position. I am detecting many more radio signals that way (including things like wireless printers, Roku media boxes, wireless televisions, and other low powered radio signals) that I did not find when the antennas were all in the vertical position. This is with the stock antennas. Why doesn't the router manufacturer recommend a particular antenna orientation pattern (vertical and horizontal)? I know that antenna orientation is particular to the application environment but there are no guidelines for setting up the router antenna orientation or even how to test and optimize signals.

4. By using longer antennas than stock (TP-Link 8db High Gain Antennas TL-ANT2408CL) in all three antenna positions, the number of radio signals detected by the router is increased dramatically. The signal strength, as indicated by the site survey tool of the Asus-Merlin firmware, is also significantly higher. I easily detect high signal strength from wireless radios and repeaters that are more than 100 meters away (through several concrete walls and in completely different buildings) that the Asus-Merlin Firmware says is 100% power - greater than -50dbm. Granted, this is on the 2.4GHz frequency which very populated in the area of my test (Urban Residential). I do see an increase in the 5GHz signal strength with the longer antennas and more when the antennas are oriented in the 0,45, and 90 degree angles. Why don't the router manufacturers provide these larger whip antennas to get better radio sensitivity?

5. Vertical and horizontal antenna orientation does make a difference for the 5GHz radio frequency. I do get a measurable increase in the signal strength received from one router to the other router when orienting the antennas of both routers in the 0, 45, and 90 degree angles. Since there are only three antennas there are two areas of 45 degrees in the circular orientation that are not seen swept by the antenna. Would having a fourth antenna which is oriented at -45 degrees improve signal sensitivity for polarization effects?

6. Because detected signal sensitivity is dependant on the antenna area, I would have expected a signal strength increase or transfer rate increase with the Asus WL-ANT-157 plate antenna. I found quite the opposite in fact. There are many antenna manufacturers of Yagi antennas, directional (plate antennas), and other wall mount antennas which may or may not work. Why would Asus sell aftermarket antennas that are such poor performers when they are the ones who design the radio array? I expect to get better performance from an antenna specifically sold by the router manufacturer for improved performance.

7. Is there any free software that helps to model antenna design, length, and orientation?

 

[Just Checking]

As an addition, I have no equipment to measure signal noise which is just as important as measuring signal strength.

The "real world" work around is to measure upload and download transmission rates after making changes. That is what I am really trying to improve or maximize anyway. By using the same single iso media file with the same uplink/downlink equipment, I can develop a comparative understanding of the changes. Each environment and set of equipment will provide a different absolute number. By keeping the environment and equipment the same (except the antennas) I believe that I get data points that have meaning in relation to each other.

 

[TonyH]

Hi,
Since you mentioned different feed point impedance on an antenna.
Draw a sine wave of current and voltage present at resonant to the frequency. Here
you can make current lead or lag or vice versa with voltage which will change the impedance. Whip is two kinds, one full length single conductor whip, one with coiled wire
whip which will reduce size but because of increased inductance, there should be counter balancing capacitance is need. On ARRL antenna hand book specially refer to
VHF/UHF microwave chapter and actual antenna design. Polo=arization can be expressed as vector. due to real world imperfection it is not always straight vertical or horizontal.
You can make an antenna which will receive both V-H component of EM wave, the signal. Noise is two kind external, internal. S/N ratio is two kind one signal strenth vs total noise or one excluding inherited internal noise of a RX or TX, So it comes down to
what's most important for you. That is where effort should be focused. Like if you incease gain coverage decreases benefitting only ones where beam is directed. Basic vertical coverage in horizontal plane is omni directional. dipole will have figure 8 pattern.
That's why plain vanilla DBi figure is lower than DBd. Rough demonstration on 3 whip mounted on router, think about 3 inflated baloons between your hands and squeeze them like accordion, each balloons E-W size will decrease and N-S increase. Hope you can visualize this as change in radiation patter. Next read about Yagi Uda basic 3 element design, director-radiator-reflector Here you can see radiation pattern on
polar coordinate. big forward Main node, much smaller rear node, various small spurious
side nodes. Graphical comparison of forward and rear node is gain of the antenna.

 

[Just Checking]

Hi,
Since you mentioned different feed point impedance on an antenna.
Draw a sine wave of current and voltage present at resonant to the frequency. Here
you can make current lead or lag or vice versa with voltage which will change the impedance. Whip is two kinds, one full length single conductor whip, one with coiled wire
whip which will reduce size but because of increased inductance, there should be counter balancing capacitance is need. On ARRL antenna hand book specially refer to
VHF/UHF microwave chapter and actual antenna design. Polo=arization can be expressed as vector. due to real world imperfection it is not always straight vertical or horizontal.
You can make an antenna which will receive both V-H component of EM wave, the signal. Noise is two kind external, internal. S/N ratio is two kind one signal strenth vs total noise or one excluding inherited internal noise of a RX or TX, So it comes down to
what's most important for you. That is where effort should be focused. Like if you incease gain coverage decreases benefitting only ones where beam is directed. Basic vertical coverage in horizontal plane is omni directional. dipole will have figure 8 pattern.
That's why plain vanilla DBi figure is lower than DBd. Rough demonstration on 3 whip mounted on router, think about 3 inflated baloons between your hands and squeeze them like accordion, each balloons E-W size will decrease and N-S increase. Hope you can visualize this as change in radiation patter. Next read about Yagi Uda basic 3 element design, director-radiator-reflector Here you can see radiation pattern on
polar coordinate. big forward Main node, much smaller rear node, various small spurious
side nodes. Graphical comparison of forward and rear node is gain of the antenna.

I believe that I can visualize what you are saying regarding the field interactions of antenna being distorted by the other antenna fields. It is easy to visualize the distortions of the field when the 3 antennas are parallel to each other. That causes the center antenna field to be pushed out to the front and rear thus creating a directionality and increased range in front and back. I suspect that the "beam forming" function built into these routers is done by changing the power individually into each of the three antennas to shape the EM field.

What makes me intrigued is how the router software adjusts for changes in the antenna positions. There must be a sophisticated algorithm which detects changes in the impedance for each router and changes the power to each antenna individually to compensate. My latest antenna configuration is having the router antenna positions in the -45 degree, 0 degree (vertical), and +90 degree positions in a plane. With my setup, it gives the most apparent signal output to the front and back of the router; while providing increased detection sensitivity over a wide range of angles. It is puzzling to try to think what the fields look like with the antennas in these positions. I also tried come up with a configuration where the antenna are orthogonal to each other. (just to see what would happen). If it were possible to do this, there should be the least interaction of each antenna field (I think). Is this true?

 

[TonyH]

Hi,
We can say beam forrming is sort of beam steering. We HAMs used to play arraying antennas without rotating the actual antenna, can steer the beam by making each antenna phase relation adjustment. There could be several ways doing it by software means, not hardware. I am retired dinosaur now, so current up-to-date specs. is beyond me but if I have to deal with some thing like this, I believe I can pick up things pretty quick, I am at wrong side of 70 I always want to prod my aging brain cells, LOL!