1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

ASUS RT-N16 replace stock antenas

Discussion in 'Tomato Firmware' started by Elfew, Apr 11, 2012.

  1. Elfew

    Elfew Addicted to LI Member

    Hi all,

    I have ASUS RT-16N with Toastman firmware...

    I wanna increase signal from my router...

    Are that antennas good?

    Or can you provide me some help or other products?
    Last question - I live in block of flat - I have bad signal in bedroom, so I want to replace the stock antenas by stronger ones. I hope that i will be better with 12dBi antenas... but I am not sure.

    Thank you for you advises :)
  2. Washu-Chan

    Washu-Chan Networkin' Nut Member

    On my RT-N16, I replaced the left and right antennas with Super Cantennas.

    I don't know if it makes any difference, though.
  3. Planiwa

    Planiwa LI Guru Member

    I bought similar antennas (although for more than $1) on eBay, and found them to be no better than the originals. They are probably fakes -- just a piece of wire inside a plastic tube.

    Did you notice that the specifications state:
    Power : 50W

    I would first find a way of measuring antenna signal, so you will be able to see if the new antennas are "better" than the old.

    Even if the signal is stronger, this may make the problem worse, by introducing more noise or distortion.

    In my case, I am now pretty sure that it is the wireless interface itself that is the problem. Its fancy measurements and "interference mitigation" are bogus. Turning it all off results in the most stable wireless network.
  4. Gaius

    Gaius Networkin' Nut Member

    These Chinese antennae that go for "rock bottom prices" on Ebay are typically garbage or fakes that give you a worse signal than the OEM antennae.

    Perhaps somebody with good hardware experience will come into this thread and guide you towards something legitimate.
  5. mvsgeek

    mvsgeek Addicted to LI Member

    I bought those same antennae on ebay, and have found them to be worthless. Compared with the stock rubber duckies on the RT-N16, they actually reduced the signal strength as reported by Inssider from -67 to -75, at a distance of about 30', passing through 2 walls and an insulated floor/ceiling. I achieved the best improvement (-67 to -63) by using the Windsurfer template.
  6. Gaius

    Gaius Networkin' Nut Member

    Not bad considering you didn't even have to use a new antenna.
  7. Elfew

    Elfew Addicted to LI Member

    OK, thank you for your time. Someone tells new antenas will have better performance than stock ones. I am not sure about that.
  8. koitsu

    koitsu Network Guru Member

  9. mito

    mito LI Guru Member

    Hi Elfew, i had an issue similar to yours, in Koitsu's provided upper link there is also a a post of mine with my experiences on my Asus avatar pick with three 9dbi antennas, those are working nice for me, but do not expect to have a signal increase more than 6-8% that for me is very important, also testing a lot of hours and checking with Inssider, i use Tx at 68. http://www.linksysinfo.org/index.ph...ng-some-asus-routers.37296/page-2#post-181791
    Off topic. By the way, does your avatar pick comes from a windsurfing forum?
  10. Planiwa

    Planiwa LI Guru Member

    What is the meaning of 6-8% in this context?

    I understand that 9dBi means 500% of 2dBi.
    7-8% is less than 2% of that. Well within the measurement error.

    If I replace 3x2dBi antennas with 3x9dBi antennas, should there not be a clear , consistent, and demonstrable increase in RSSI of at least 7dBi?

    What is meant by 7-8%? Is that 7-8% increased range?
    In other words same RSSI at 7-8% greater distance from AP?

    Are there measurements that show consistently greater RSSI at the same distance?
    7 dBi difference?
  11. mito

    mito LI Guru Member

    Radio frecuency does'nt understand about percertages, it would be very easy then to get a 18dbi antenna and get a 1000% increase.
    After a lot of try and error and several tests with 2dbi, 5dbi, 7dbi and 9dbi antennas at my most far office at my warehouse , i got a very small 6-8% increase from 2dbi stock antennas to 9dbi antennas i bought at http://www.fab-corp.com/product.php?productid=3069&cat=250&page=1
    I got that increase but most people does'nt get any significant measurement increase when changing antennas.
    Then yes, as you say in other words, "same RSSI at 6-8% greater distance" , that is the reason they work very well for me.
  12. Toastman

    Toastman Super Moderator Staff Member Member

    The real reason that changing antennas usually doesn't do much is simple. Most of the stuff on the market is absolute junk.

    The ASUS RT-N16 antenna internally is a "proper" coaxial dipole. It is actually 2" long and it's nicely made, almost certainly tuned to the correct band.


    In the local shops here, I have seen around a dozen different add on antennas from various sources, but in essence, they are all made in China. I do not need to say more about the likelihood of them having been designed properly. Since it's possible to crack them apart relatively easily, if the things were not in sealed packages I've often done so in the corner of the shops when nobody was looking. Much of the time there's a straight bit of wire inside of no particular length. Sometimes an attempt has been made at constructing a coaxial collinear array, but the lengths are hopelessly wrong and have no chance of providing any gain. But they look good :) Some have fancy spirals of wire, and one even had a 75 ohm resistor soldered half way up!

    In one shop I found supposed "Linksys" antennas with marked 10dB of gain. Inside was a bit of wire about 9" long, completely useless. The giveaway - there was the usual spelling mistake on the packaging!

    The cost to make one of these would have been about 70 cents for the housing and a few cents for the bit of wire and labour costs. Yet the cost was almost that of a WRT54GL !

    Many if not most of the articles on the internet about home made antennas are complete crap written by people who have no knowledge of the subject. Making antennas work at these frequencies is not easy. The slightest mistake in length, the proximity of the plastic housing, all detune the antenna and make it worthless. You need professional test equipment to tune and evaluate them. You can't simply make something that "looks" vaguely like a scaled down version of a taxi antenna and expect it to work.

    BTW - InSSIDer, while being useful, is not really up to the task of measuring anything. Especially, it can't be used as to compare results when used on different hardware.

  13. Elfew

    Elfew Addicted to LI Member

    Thank you Toastman!
  14. mito

    mito LI Guru Member

    Hi Toast, agree with everything, my entennas were bought from the link probided before, and the definitory test -between other tests- was made as simple as using an Iphone at the farthest office at my warehouse, with the stock 2dbi antennas no signal, with the 9 dbi antennas, conected to wifi.
    Just my 2 cts.
  15. Monk E. Boy

    Monk E. Boy Network Guru Member

    As a general rule... if it's listed as a new product and it's on eBay, it's almost assuredly fake/counterfeit/knockoff product. This holds true for most items on eBay.

    Used products are usually legitimate but are, well, used. The ones that aren't legitimate are people who bought something "new" off eBay, realized their mistake, and don't care about ruining someone else's day.
  16. Toastman

    Toastman Super Moderator Staff Member Member

    Yep. All tests must be done over a period of time and under controlled conditions.

    Let me illustrate a point about the RSSI - Max value used in Wifi systems by different chip manufacturers. We assume a stable link, with no changing path variations.

    We have a router on a table. We have a client on another table 25 meters away. It is receiving a signal of -55. That reading is JUST on the threshold of changing to the next "step" in the RSSI table.

    OK, so now we add a little antenna with say 0.1dB extra gain. Now the distant client receives a signal 0.1dB stronger, but because of the crappy RSSI system, its display now jumps to the next "step" in its measurement parameters, which could be 4dB higher. So it now tells you that that antenna gave you an extra 4 dB - instead of almost nothing! However, since it's only an error in the measurement and display, the actual throughput of the client does not change.

    Now, from minute to minute, hour to hour, day to day, the signal path is always changing, maybe someone moved in the room. Maybe someone put a teapot on a table somewhere that wasn't there before, or moved a mirror. That perhaps also pushes the apparent signal up 4dB, it could do the opposite, of course.

    So, for example, someone experimenting with placing a lump of putty on top of the antenna (!!) might conclude that his "mod" had increased the gain by 4dB, when in fact it didn't do anything at all.

    This is how people are able to get away with selling junk.

    Smoke and Mirrors...
  17. Planiwa

    Planiwa LI Guru Member

    When all we have is Tomato running on an RT-N16, the wl command is our measuring instrument.

    I have been taking measurements for several weeks.

    After setting up Tomato router as a client, and leaving it alone, I have seen the RSSI, as measured by the main access point, vary from -11dBm to -50dBm.

    That's without changing anything at all!

    The difference between -11dBm and -50 dBm is a factor of 10^3.9, which is 8000.

    That's a measuring error of of 800,000%.
  18. Toastman

    Toastman Super Moderator Staff Member Member

    Somewhat inaccurate, no? :D

    Usually here, the readings are relatively stable, they do vary by several dB, which is however pretty normal for wireless signals due to path changes, cancellation, multipath, etc. - but as you say there can be times when the measuring process appears to have died altogether. If the wl driver actually took some action as a result of that failed measurement, then who knows what might happen.

    The KISS principle was never apparently applied to wifi. It is the most convoluted mess I have ever seen.
  19. Monk E. Boy

    Monk E. Boy Network Guru Member

    I'd say the fact that signals aren't limited by line of sight coupled with the massive adoption rate of 2.4Ghz technology is the root cause of a huge discrepancy in signal strengths.

    I wouldn't be surprised if wide variation is caused by interference from other 2.4GHz devices, 802.11 or otherwise.. At one townhouse my wireless reception would get shot to hell moments after I (faintly) heard a neighbor's phone ring, because they picked up on their 2.4Ghz cordless handset and blew me out of the water.

    I'd also like to talk to whoever was the first one to think that setting up 2.4Ghz access points on channels other than 1, 6, or 11 was a good idea for "Auto" channel select. It's one thing when end users mistakenly do it out of misunderstanding, but the guys writing firmware should have a better grasp of the problems inherent in autoselecting a channel that effectively screws over not one but two nearby access points...
  20. Toastman

    Toastman Super Moderator Staff Member Member

    Yep. They should be hanged. BUT - whether allowed all channels or restricted to 1,6, and 11 it's still a bloody stupid idea to let access points keep jumping all over the place. It is not possible to select a decent channel if everybody elses' (and your own) AP's keep changing channels every time they think there's a better one, causing no end of interference as they do so. The goalposts are continually changing.
  21. Planiwa

    Planiwa LI Guru Member

    I was wrong when I said that one client's RSSI changed between -11 and -50. It actually changed between -11 and -58, and within a few hours. [Meanwhile, noise has fluctuated between -92 and -72.]

    The client in question is a WRT54GL, running in Client Mode, and being hidden in a kitchen cupboard, to serve as an "on-site informant".

    The idea was that without such an on-site known device, I would never know whether "no activity" meant "no one wants to connect" or "no one is able to connect", etc.

    The idea was that if this host becomes invisible, I *know* that there is a Wireless problem.

    Trouble is -- it has never become fully visible. :-(

    I seem to have misconfigured some subnet or routing setting in this Client-Mode.
    I can ssh to it, and I can arping it.
    But arp does not show it, and I cannot get ping responses from it.

    EDIT: arp visibility fluctuates.

    (I wonder if anyone else has tried to use a "router" as a Client-Host, rather than a Client-Adapter???)

    It's at a remote site, so I can't just plug in an ether connection.

    But it has certainly not moved.

    In other news, I have now incorporated most of my findings about "wl" and some about "nvram" in one command that retrieves readings and settings. It is for people who want to help advance the state of understanding of this wl command.

    Back to RSSI measurement -- when I have more time, I hope to take (wl) signal measurements and analyse them over time. I have measurements, but haven't had a chance to analyse them, or even do a preliminary analysis to see what might be a good cadence and format for this.

Share This Page