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

ASUS RT-N16 Transmit Power

Discussion in 'Tomato Firmware' started by Elanzer, Oct 19, 2010.

  1. Elanzer

    Elanzer Addicted to LI Member

    For the ASUS RT-N16, the default value is "17" for the transmit power in the latest version of tomatousb firmware.

    I know in dd-wrt some people say this is dBm rating so 17 is equal to 50mw and a setting of 20 equals 100mw. I assumed this was different in tomato, as past builds defaulted to 42 for me as I remember.

    Now that the default is 17, does this mean the setting is actually dBm and not mw?
     
  2. myersw

    myersw Network Guru Member

    Default has always been 17 in all builds I have used of Tomato and DD-WRT. The Asus OEM firmware also defaults to 17. There have been reports on forums of increasing transmit will actually reduce the range. Thinking being that the transmitter is over driven in to distortion. I know when I tried 42, as an experiment, running Tomato I experienced reduced distance then with the default 17.
    --bill
     
  3. Toastman

    Toastman Super Moderator Staff Member Member

    Asus RT-N16 and Linksys E3000 Transmit power versus RSSI
    (applies to most hardware using the RT drivers)


    OK, that got me interested enough to pull one of my routers and enable wireless! I just tested B and G at different power levels - country set to US and verified with Singapore (which allows more channels). Firmware was Fedor's latest RT beta23 ext. with 5.10.56.46 ND wireless driver

    I monitored the strength with 4 different WRT54GL's at different distances. The one nearest to me (5m) showed really strange results - power above 50 caused an indicated DROP in signal strength (which is of course impossible). This indicates that something in the wireless driver is creating a situation that it should not, or the receiver is overloading. The next furthest away AP, dropped off rather less, and one about 40m away showed an absolutely flat level above 50, which is what I would expect. This shows that some weird things can happen when you are rather close to a router and you can't rely on the RSSI indicator at all. And after all, 5m is not that close!!!

    So remember this - if you are too close to your router you may find you need to reduce the power, not increase it.

    Anyway, having established that the more distant router seemed to be giving sensible readings and was suitable to make measurements, the 40m GL gave the following plot. Modes B and G were tested on channels 1, 6 and 11 - all had a similar curve but mode B was always 2 - 3 dB stronger than G (which is correct).

    From this, it seems that the maximum strength is reached at a setting of about 60. No improvement was obtained changing to higher figures such as 250 or up to the (strange) 400 limit.

    Setting as Zero (said to take the default setting from the wireless driver) gave a reading which corresponded roughly with a setting of 20. So it probably took 17 as a default from the driver. Unlike some people, I have never found the default of Zero to give best range and speed. I have always found best range and speed occurred with the highest power I used. But this may be normal as the setting of 0 to set hardware defaults is known to work only with some router models and some countries.

    A subsequent download speed test at 17, 41, 60 and 100 mW (deliberately chosen as being higher than the 60 maximum) showed no change in speed whatsoever. Also interesting is that the noise floor (as measured by WRT54GL routers at various distances) did not change at any setting up to 400 on the same or adjacent channels. From this one can deduce that if stock level were 17, then about 6dB increase in strength should be obtained by increasing the level to 50 or 60, and my tests showed no adverse effects whatsoever from doing so. No significant change in noise occurred on adjacent channels at this power setting.

    Mode "N" was also tested, the signal strengths were the same.

    [​IMG]

    Similar tests previously conducted on the WRT54GL (this was part of a long discussion):
    [​IMG]

    You can see from this that the RT-N16 is pretty similar to the WRT54GL with an ND driver. The Linksys E3000 is absolutely identical, probably this applies to the other routers in the E series too.

    1) DD-WRT v24sp1 with ND driver 4.150.10.5
    2) Tomato 1.25 with standard driver 4.130.19.0
    3) Tomato 1.25 RAF with ND driver 4.158.4.0

    CAUTION

    Those who have tried doing this kind of test know full well how much the signal leaps about and how difficult it is to get a proper reading. Each point on the plot was read a minimum of 20 times, obvious random scatter eliminated, and the mean taken of the remainder. Having done this, the tests were repeatable within a few dB every time. The test took 7 hours to repeat several times and more than 2000 readings were taken. My point is - whenever you read a post that claims some rather significant change in signal strength, take it with a huge shovelfull of salt. The poster probably did not do any proper averaged tests over a long period, and is just wasting everybody's time.

    I'm very sorry if it offends anyone, but neither Teddy Bear or myself take much notice of these claims in the forums. The majority of the reports are quite clearly wrong. Let's make a point here. There have been over 450,000 downloads of Toastman-RT versions alone, mostly for the RT-N16 and E3000. On another website there have been 7500 downloads solely for the WNR3500 series. Add to this far more downloads of TomatoUSB and then add Shibby and Victek's builds and we're looking at maybe a quarter of a million or more! And just a very few have reported wireless issues. New issues have even been reported with the wireless after nothing at all has been changed in the firmware except a comment added to a QOS setup page! :)

    In fact, you should beware of statements made in forum posts especially in the DD-WRT forum. The vast majority of posts are merely repeating what someone has "heard" - and nobody can remember who started all of these myths.

    See also:

    http://www.linksysinfo.org/index.php?threads/asus-rt-n16-transmit-power.33078/#post-172726
    http://www.linksysinfo.org/forums/showpost.php?p=350183&postcount=66
     
    Elfew likes this.
  4. Elanzer

    Elanzer Addicted to LI Member

    Hmm, so it seems about 60 is the optimal value. Is that with a more recent (ie: <1month old) build Toastman? At some point 42 was the default value - I'm pretty sure of this, but I rarely do a full nvram clear between each build so I'm not sure how long it has switched to 17 as default, atleast a few months ago 42 was default.

    I'm just wondering if the actual value it's taking into consideration has changed somewhere along the line, or if it has always been the same. I don't want to find out I've been setting a ridiculous ~80 dBm for a 100000mw output on these routers all of a sudden, then again I would be surprised as to how they're still functioning..
     
  5. Toastman

    Toastman Super Moderator Staff Member Member

    Yes, I used Teddy Bear's beta 43 on a recently bought RT-N16. But any build with the recent wireless drivers will behave the same way.

    These routers wireless device is almost indestructible. The PA chip is designed to run at a maximum of 250mW and looks like setting a power figure of 50-60 might correspond roughly to that 250mW. As the specification, mode G is running at around half the power of mode B. The figure we are entering is otherwise probably totally meaningless other than as a relative number, and doesn't interest me in the slightest.

    Nobody seems to know what the strange figure of 400 for maximum power is there for. You can set any figure over 60 but the signal strength won't change, nor will the router suddenly blow up or overheat - but it may well produce a very dirty signal - so don't do it. 60 is the maximum, you can't output more than the maximum, don't try.

    Incidentally, the figure of 42mW was probably carried over initially from the old WRT compiles - but since it doesn't relate to a RT-N16's wireless drivers at all, it was probably changed to 17 somewhere along the line because some people on the forum found it to be more stable. If you read back through Teddy Bear's threads you'll probably find the posts. I never understood the logic behind that setting, as it never made anything better on any of the RT-N16's I have - it made things worse with the more remote, weaker clients, as you'd expect from reducing the power level.

    Why may they have found 17 to be better? I don't know because none of my tests have ever shown any disadvantage in running higher power, quite the opposite in fact. I conclude that maybe the posters had their routers very close to their PC's, and one or both radio receivers were overloaded. Reducing the power therefore helped them.

    So to recap, in my case at least, a setting of 60 most certainly doesn't give worse performance than 17. That has never happened here - ever - on any RT-N16 (I have quite a few) or the E3000. I would be inclined to think that persons who found this may well be too close to their router, one or both receivers may therefore be overloading, when reducing power would improve the situation. For all we know, the RT-N16's receiver may be more prone to overload than the old WRT series, and that could explain a few things. I wish that I still had access to a test lab with proper test equipment, and could make some definitive measurements on this stuff, but sadly I don't.

    ***EDIT*** N tested briefly !

    Someone just fetched their "N" laptop to do a very quick test. While they surfed the web sitting maybe 10m from my router I established that the N plots with 20 and 40MHz bandwidths followed almost exactly the same curves. Statistics showed no undue packet loss and speed was no different between 17 and 60 when we did a quick file download from my webserver. In future there's no need to use an N client, just measure and be done with it.

    Please note that I have since also tested channel 11 with no change in signal strength.

    There are several settings in Advanced/Wireless that can affect the signal strength or reliability. In particular, the "interference mitigation" setting is best turned off or at least left at "Auto". APSD mode may also be a PITA. If working, that sets "power saving" features, and wireless clients go to sleep for periods and sometimes don't wake up. That's a bit like Russian Roulette :) Intel cards especially are dreadful. Often rebooting the router will force the offending client to wake up, I don't think it is the router at fault because other clients function normally while the Intel card snoozes .....

    So - at the clients - turn off any power reduction/negotiation settings you can see in the wireless card's advanced setup, and set it to maximum power. Turn off any sleep mode. Look suspiciously at any other settings if you have connection problems. Be aware that Intel wireless devices (even the latest ones) are often very unreliable. Intel's power saving "features" are to a wireless connection what a steamroller is to your cat :eek:
     
  6. Catalin

    Catalin Addicted to LI Member

    Thank you very much for the tests.
     
  7. Toastman

    Toastman Super Moderator Staff Member Member

    Today I was testing the new wireless drivers used by Fedor in his latest updates posted in the tomato repository. One thing led to another .... and hence this post.

    There are still quite a few people in the forums saying that the RT-N16 and other MIPSR2 routers have very weak radios and poor performance. Some people have tried to back this up with various screenshots and measurements, most of which have been wildly inaccurate and contradictory, making impossible statements that contradict the laws of physics.

    I've always been curious about this. You see, I bought one of the very first RT-N16's and I've used it ever since, blissfully unaware that it's supposed to be crap. I now have a lot of them. You'll see in the posts above quite a few router measurements, carried out over long periods of time and checked and rechecked, which all show that there isn't anything wrong with the RT-N16 transmitter.

    Now, as a final sign that all might be well with it, here's a new screenshot:

    [​IMG]

    I put an RT-N16 set to tx level 60 and a WRT54GL with tx level set to 150 on a table at the exact same distance (10m) from, and line-of-sight to the receiver (which is a DELL Latitude D420 Laptop) running InSSIDer. I repeated the tests many times with several different Tomato firmwares and wireless drivers, using slightly different positions of the routers relative to each other and the receiver, to rule out transmit path changes and reflections. The results are repeatable every time and with any firmware and any ND driver. I also swapped routers from the one I bought right at the beginning to one of the latest purchesed only a month ago. No difference.


    EDIT- The new 5.100.138.9 driver for the more recent routers such as the E4200 also gives the same results (Oct 5 2011).

    I repeated the tests with the more common defaults, the WRT54GL set to 42 and the RT-N16 set to 17. Apart from being a few dB weaker, both were once again almost identical.

    [​IMG]

    AT MY LOCATION THE WRT54GL and the RT-N16 ARE PRETTY MUCH IDENTICAL IN SIGNAL STRENGTH.

    As you can see, there really is no significant difference between these router's wireless signals. It is absolutely certain that I never see the huge difference that some posters say they experience. Both routers were accessible in the street outside about 100 meters away. There was no significant difference in download speeds between the two routers either.

    I have no explanation for the poor results experienced by some posters. However, it is probably a minority. Statistics show that my builds alone have been downloaded (2010) more than 57,000 times, and most of the builds have been for the RT-N16 and the E3000. TomatoUSB and Tomato-RT, Victek, and Shibby's builds added to this must be at least a few hundred thousand downloads. But actually there aren't so many complaints of poor wireless when this fact is considered, I believe that most people don't have any problem. (early 2013 now more than 250,000 downloads).

    *

    Later ...

    I am trying to perform the same test using DD-WRT as a comparison. Now that's proving to be a real PITA. After a few minutes DD-WRT seems to shut down it's radio and then come back up again at a random, unstable, weaker level, before doing the same all over again. It's totally unstable, even worse than I remember it.

    [​IMG]

    Much later ....

    Well for what it's worth, I waited until it stayed up long enough to get a few minutes of graph. As you can see, it's broadly similar at first. You can also see a first dropout just after 8.30, followed shortly afterwards by two more at 8.31, followed by a period of real instability and poor signal, as the signal strength dropped off for no apparent reason. Once this instability has started, it doesn't stop, and throughput drops badly. Reading the DD-WRT forums gave no explanation for this behavior.

    Anyway, you can see that DD-WRT is NOT any stronger than Tomato, but in my tests it was less stable. Many posters on the DD-WRT forum found the same.

    Now I need a drink before getting the router fumigated. Every time I use DD-WRT I feel like screaming.

    EDIT:

    If you wish to repeat these tests, bear in mind that moving a router a few inches in any direction or repositioning the antennas can raise or lower the signal strength at the receiver by a large amount. You need to repeat the tests many times to get consistent results, changing position to get maximum strength befoore you commence the measurements. Likewise, the site you choose for your router is very important - keep it clear of obstacles and as high as you can get it. If it has internal fixed antennas you may have to try several locations to find a good one to cover your room or house well. Just move a few inches and you'll see big changes in strength. ]
     
  8. Gaius

    Gaius Networkin' Nut Member

    Interesting tests, Toastman. I've found that the Asus has about the same range as my WRT54GL, which is not a bad thing for me but could be a limiting factor for others.
     
  9. Toastman

    Toastman Super Moderator Staff Member Member

    Agreed. Cover is broadly the same. Since the WRT54GL with it's external antennas has been kind of the "standard" for many years to compare everything against, this is what I was trying to point out.

    I know of no normal router with a stronger signal except possibly the Buffalo units with built in amplifiers. More recently, the E3200 looks as if it has high power amplifiers, but contrary to several forum posts the E4200 does NOT have them.

    I did a few little tests with the newer N driver for E4200 (but on the RT-N16 and E3000) - same strength.
     
  10. Gaius

    Gaius Networkin' Nut Member

    Wait, are you saying the E4200 doesn't have much more range than an RT-N16?
     
  11. Toastman

    Toastman Super Moderator Staff Member Member

    No reason why it should, no, except that MIMO implementation may improve performance slightly, but the actual transmitter power is similar. On 2.4GHz it runs the same power amplifiers, on 5GHz there are no external power amplifiers at all and it runs about 150mW max. direct from the SOC chip on each of the 3 antennas, which is quite low. It might run even less than that depending on what voltage the chip is running on and what the default power is in the wireless driver. Take a look and see what you think after perusing the available info.

    http://www.linksysinfo.org/index.php?threads/toastman-releases.34549/page-19#post-173315
    http://tomatousb.org/forum/t-303285/linksys-e4200-support#post-998805
     
  12. mito

    mito LI Guru Member

    Hi, agree evreything with Toast about WRT54GL to N16 comparison, but when in N16 set Transmit Power to "0" , signal went really up, plus i found three 9db antennaes and RT-N16 is really a champ in range and speed.
    Rgds.
    mito
     
  13. Gaius

    Gaius Networkin' Nut Member

    Yep, I can confirm as well that I got a measurable improvement in signal when I set it to "0".

    However, a value of "60" gives you even more improvement. The maximum as stated by Toastman.
     
  14. Toastman

    Toastman Super Moderator Staff Member Member

    0 seems to allow power to be varied by (something) - the signal strength does vary. Sometimes, depending on channel, country, interference settings, and the direction of Mecca, setting 0 can give lower strengths. Setting 60 gives me a consistently strong signal with minimum variation and best signal on RT-N16, E3000 (and most probably all routers using the same wireless driver). It's hard to know what what is going on. However, I think it's probably safer to set 0 as default, and that's what I am going to do.
     
  15. occamsrazor

    occamsrazor Network Guru Member

    Very interesting thread, though starting to get a bit above my head :) Can anyone recommend Transmit power settings for optimum range on an E3000 that's using 20Mhz channel on the 2.4Ghz, 40Mhz channel on the 5Ghz, both set to Singapore, interference mitigation = none, APSD = Off, with Transmit Power = 0.
     
  16. Toastman

    Toastman Super Moderator Staff Member Member

    No idea about 5GHz because I never did any tests on that, but 60 for 2.4GHz.
     
  17. occamsrazor

    occamsrazor Network Guru Member

    Thanks, will give it a try. Out of interest do you know what equivalent level "0" is using on an E3000 on 2.4Ghz?
     
  18. Toastman

    Toastman Super Moderator Staff Member Member

    I "think" it is probably roughly equivalent to 17, but hard to say for sure. Mostly because it's not a clear cut level, selecting 0 allows something, (we don't know what), to vary the tx power according to some algorithm (we don't know what) and that causes the signal strength to vary from model to model. Let us know if changing from 0 to 60 does anything for you, as it did for Gaius (above).
     
  19. occamsrazor

    occamsrazor Network Guru Member

    Hard to say a precise change as around the same time I added an additional access point in a slightly different location which solved the problem more easily. As all working OK now in the location that was problematic I don't really have time to do any serious testing as busy trying to get a new 3G modem to work :)
    I would be interested in some transmit-power figures for the 5Ghz band, as that range really sucks. But I knew that before I bought it and am very happy with the close-range speeds (getting about 13MB/sec transfers).
    Finally, just want to say that all your lengthy and well-written explanations of stuff like this and QoS etc are fantastic and much appreciated.
     
  20. Toastman

    Toastman Super Moderator Staff Member Member

    The 5GHz band uses around half the power of 2.4GHz, which gives it an immediate -3dB disadvantage. Some channels are restrictyed in power anyway. Often to very low powers. Use Google for info. Then there's the shorter range and lower penetrating power of 5GHz - so it will never be as good as lower frequencies, this is bound by the laws of physics. For long-range links however, it does make the use of relatively small parabolic dishes more inviting :)
     
  21. tvlz

    tvlz Addicted to LI Member

    These links may be of interest
    General Info
    rt-n16 specific

    I think '0' would be a better default value for all new builds, stronger than '17' one less thing to forget about changing.:D
     
  22. CardinS2U

    CardinS2U LI Guru Member

    did you mod your router at all? fan etcs?
     
  23. Toastman

    Toastman Super Moderator Staff Member Member

    You can add fans if you like, is your router overheating? Normally, whenever a new router comes out everyone starts adding fans. It's usually completely unnecessary, and a new source of maintenance problems.

    Please, everyone, don't expect 5GHz to be anywhere near as good on range as 2.4GHz, the range reduction is wholly dependent on the laws of physics. Look upon the shorter range as an added benefit - less interference!

    You should also do some research - some 5GHz channels are allowed to run more power than others by the wireless driver - the range will therefore vary with the channel used.

    Google is your friend here ...
     
  24. mito

    mito LI Guru Member

    Hi, agree with Toast, no need of fans.

    If you want to get your N16 run freshier lower the frequency a bit, could be to 300 and be in the safe side.
    Rgds.
    mito
     
  25. Toastman

    Toastman Super Moderator Staff Member Member

    I wouldn't agree with that, mito. Crippling a 500Mhz router by running it at almost half speed is rather a waste. I would never go lower than 453, if you need to go lower, then you really do have a duff processor - but I personally never found even that is necessary.

    I have always found 480 MHz to be 100% stable. I have a LOT of RT-N16's and I am in the tropics where the ambient temperature and humidity is always very high. Actually, I've always found 500MHz to be stable too, although I don't use it myself.
     
  26. mito

    mito LI Guru Member

    Sure you are right, yes i know that it is a waste having such a great touter! I meant it only to be in the safe side and to have peace of mind about the high temperatures, because always somebody is afraid of that. My avatar pic router is running at default frecuency, Tx power 62 and no fans.
    Rgds.
    mito
     
  27. Planiwa

    Planiwa LI Guru Member

    I am trying to relate what has been written in this thread about transmit power to the wl command.

    These appear to be the wl command invocations that are both relevant and not "unsupported":

    wl txpwr -1 # Set the transmit power to default value.
    wl txpwr # display current transmit power.
    wl txpwer N # Set tx power in milliwatts. Range [1, 84].

    wl txpwr -1 # appears to have no effect on WRT54GL or RT-16. (!)
    wl txpwr N # appears not to be limited to 84 (!) (Country Canada)
    Tomato setting of 42mW results in 42mW = 16.25dBm
    Tomato setting of 0 ("hardware default") results in 1496mW (!!!) = 31.75dBm

    This looks like a serious problem.

    * * *
    I should also mention that when simply changing that setting in the Tomato GUI, which should affect nothing but the transmit power of the radio, the following side-effects also happen:

    1. Syslogd restarts, and destroys the contents of the entire message log in /var/log/messages. (!)
    2. All interfaces are restarted, including the WAN connection. (!)
    3. The firewall is restarted, and even the Init script is run. (!)


    [always running in "nocommit" mode]
    WRT: Tomato v1.28.7821 MIPSR1-Toastman-ND K26 Mini
    N16: Tomato v1.28.7494 MIPSR2-Toastman-VLAN-RT K26 USB VPN-NOCAT
     
  28. Inkrypted

    Inkrypted Serious Server Member

    Asus list the Power settings on their website (http://www.asus.com/Networks/Wireless_Routers/RTN16/#specifications) as.
    N mode: 15.8~19.5dBm
    G mode: 15.5~16.5dBm
    B mode: 15.8~19.5dBm
    So I used a dBm to mW calculator I found on the web to convert the numbers (http://www.rapidtables.com/electric/dBm.htm) and found.
    15.5dBm=35mW
    15.8dBm=38mW
    16.5dBm=44mW
    19.5dBm=89mW
    So I would say that anywhere between 35 and 89 mW would be acceptable. If you only have G equipment than a setting off 44 would be appropriate. But if you have N in the mix or N only that 89 would be the best setting although I must say that sounds a little high to me. I have mine set at 44mW and it's working beautifully.
     
  29. Planiwa

    Planiwa LI Guru Member

    dB is just a way of expressing multiples or ratios. a deciBel is 1/10th of a Bel, which simply means 1 power of 10.

    so 10^2 == 100 == 20 dB
    Thus, 19.5dB simply means 10^1.95 == 89.1251
    You can easily find out the value of 10^1.95 by feeding it to Google, which will give it to its calculator.
    In this context, X dBm is the same as saying Y mW, where Y=10^(X/10).

    NB:
    Code:
      wl txpwr ?
    txpwr    Set tx power in milliwatts.  Range [1, 84].
    
    Note that 84mW is less than 89mW. Not that this matters much.

    But 1496mW *does* matter much, being 18 times the maximum limit, but thus far *nobody* has said a word about that.

    Either the software is bogus, in which case it should be fixed,
    or there is a serious problem, in which case it should be fixed.
    In any case, there is a problem that ought to be acknowledged, no?

    But no one has said -- I verified it on my router and it is indeed so.
    Or -- I tried to replicate this on my router but it is not so, instead it is like this: ______.
     
  30. Inkrypted

    Inkrypted Serious Server Member

    84 mW sounds safer than 89 mW by a long shot. I use Shibby's firmware but I would imaging that the base or Teddybear Mod would be the one that is reporting it wrong or overclocking to an extreme.
     
  31. Toastman

    Toastman Super Moderator Staff Member Member

    Setting 0 does hand over control to the wireless driver, but what it displays for tx power is not necessarily correct. The 1496 figure is clearly nonsense and quite probably refers to a maximum that is *possible* on some other hardware. The RT-N16 and most other common routers can only run 251 mW as that is the maximum output power of the PA chip. In most modes the average power is actually half of this.

    In short, the figure of 1496 is clearly meaningless, and we don't know what it is supposed to mean, if anything. Stop worrying about it, just ignore it.

    Like most things in the wireless setup, we can't change much or fix the display. It is probably there for a reason. There are probably hundreds of settings for, or related to, the wireless driver - most of which are completely undocumented as the driver is closed source, and Broadcom will not release the source code and documentation. Each time you try changing one, and get a positive result, you have no way to know if it's just coincidence, or is repeatable over time. Different wireless drivers and different issues of the same drivers by different manufacturers also have different support for many of these commands. Thus some people trying your tests on different hardware will find they don't work or even do something different.


    The maximum power output is obtained on the RT-N16 (2.4GHz) with a setting of somewhere around 60. This appears to hold across all routers I've tested when using the same wireless driver. The figure we enter into the transmit "power" box should be viewed as an arbitrary, relative, number and not an actual power.

    Some people find that setting of 0 (zer0) (which is supposed to allow the wireless driver to use it's own internal default, depending on country) gives much better results. Others say the opposite and that 60 is best for them. Setting zero does hand over the control to the driver, it is a valid setting, and the power level can be and is varied by the wireless driver. However, it is also possible for the setting to be overridden.

    The difference between 84 and 89 is negligible.
     
  32. Planiwa

    Planiwa LI Guru Member

    Let's start with [#2]:

    It is Tomato, not Broadcom nor Linksys that states:
    "Transmit Power [ ] mW (range: 0 - 400, actual max depends on Country selected; use 0 for hardware default)"
    Choosing this Tomato option results in a setting of 1496mW according to the wl command.

    It appears that, the setting of 0 results in the syntactically *maximum* setting, not the default.
    (By syntactically I mean what value can be assigned to the variable, not what can be implemented.)

    You can easily verify this with
    Code:
         wl txpwr 9999;wl txpwr
    31.75 dBm = 1496 mw.

    That brings us to [#1]:

    Yes, wl is a horror. But this particular problem seems to be with Tomato, and easily fixable.

    Since it is Tomato, not Broadcom that makes the false statement about "use 0 for hardware default", it seems a simple (and necessary) move to fix it in the Tomato web page. The fix is as simple as deleting that false text on that web page.

    Once it is fixed, a Tomato user will no longer try to specify 0, at Tomato's suggestion, to set the HW default. And a Tomato user will no longer see "wl0_txpwr=1496" mW in his NVRAM settings.


    FWIW, the wl command claims that the way to restore the default Transmit Power is "-1".
     
  33. fubdap

    fubdap Addicted to LI Member

    @Planiwa - It appears you know more about wireless technology than the average user in this forum. We will like to leverage on your knowledge. Can you provide a general wireless settings, either by hardware (Linksys, Asus, etc) or by software drivers (RT, RT-N). Thanks.
     
  34. HHawk

    HHawk Networkin' Nut Member

    Wow thanks for the research Toastman!

    I always thought anything higher than 24 would damage both my RT-N16 routers. So it's safe to use a setting of 60 with your Tomato builds?

    Another question; for some reason I still have a crappy signal. I already tried several things. Maybe I am using the wrong settings, but I think I tried a lot of different settings, along with changing channels, regarding wireless and I always receive a crappy signal. And yes, I know wireless wasn't meant to stream 720p or 1080p .mkv content, however it still should be possible.

    Now I am considering to buy som better antenna's again (I did this before, but probably some crappy quality) to increase my wireless signal. But now I have read it matters what kind of dBi antenna's you use? Is this correct? I always thought the bigger the better. Currently I am considering quality brand antenna's of approx. 8 or 10 dBi. Those antenna's are long (~40cm), but I don't mind, as long as they will work. What is your take on this?

    It's really annoying that I cannot watch a movie or serie correctly sometimes when I am in bed. And it doesn't happen always. Sometimes movies and series play great at night, while the next day, it stutters like.... Well you get the picture. Not only is it very annoying, but tiresome as well.

    Also a different question; I read in your earlier post when you select 'Singapore' as your country / region, you have more channels available. Wouldn't selecting such channel work better for me? Less interference from my neighbours?

    I already went through a lot of trouble by purchasing a 2nd Asus RT-N16 router along with your Tomato-build firmware and placing it in the living room (directly underneath our bedroom), but still the signal is really bad. I also tried the wireless 'Interference Mitigation'-settings. Though it improved a little, it is still a hit or miss situation.

    I also thought of creating a wired connection, however that is even more problematic (would save me a lot of frustration though), because I cannot get the wired cable up to my bedroom in our house. And not to mention that my girlfriend would be angry if she noticed a cable running through our bedroom... :(

    Can you please provide me with some answers or even a working solution? I understand you are busy, but hopefully you can find the time to answer the above questions.
     
  35. Planiwa

    Planiwa LI Guru Member

    ASUS PL-X32M, HomePlug AV Powerline Adapter Kit (?)
     
  36. Toastman

    Toastman Super Moderator Staff Member Member

    HHawk, as I don't live there, I can't really comment, but if the router is right under your bedroom, then there's something rather wrong.

    If I place an RT-N16 under my apartment, then I receive full bars on all clients, Laptops, USB adapters, and phones - and good strength on the floor above that. Floors are steel rebar concrete.

    What happens if you turn the 2 outer antennas horizontal _|_ ? Usually \ | / is adequate.

    Those longer antennas will probably make things worse on the floor above, not better. Read through this thread:

    http://www.linksysinfo.org/index.php?threads/is-likuda-9dbi-antenna-worth-it.36976/page-2


    Incidentally, what is inside these fancy antennas that people sell, is just a bit of wire, no magic ... the housing and shiny dibdabs are just snake oil.
     
  37. HHawk

    HHawk Networkin' Nut Member

    Thanks for the reply Toastman. Highly appreciated.

    Uhmz... That's what I thought. I wasn't expecting the greatest signal of all time, but at least something solid and stable. But after reading your reply, I should expect something way more stable.

    The Asus RT-N16 is wall mounted and I fooled around with the stock antenna's yesterday and did what you mentioned. Didn't make much difference (signal was in fact become worse). So I moved the antenna's again. And also changed 'Interference Mitigation' to 'Non-WLAN'. Apparently that did improve things, however I need to test it several days more to make sure.

    Also the laptop shouldn't be a problem. The laptop is pretty new, along with a Centrino Ultimate-N 6300 WLAN card. Which is one of the best options out there. And I am guessing the height from the router to the laptop (in the bedroom) is around 3 meters only and with only the concrete floor/ceiling in between.

    Okay I will scrap the new antenna(s) idea. Since it will only worsen things most likely. I was hoping it would improve things a bit.

    Transmission power (for RT-N16 with Tomato Toastman firmware) set at 60 is good and safe, right?

    I did notice a lot of wireless networks in our area. I also tested stuff with InSSIDer and my router's RSSI is always around 50 (measured in the bedroom with laptop). That seems to be okay, right?

    That's why I was wondering if changing my country to 'Singapore' to get more different channels, would make a difference?
     
  38. Toastman

    Toastman Super Moderator Staff Member Member

    50 is very good, stop worrying about the signal. Even -70 works OK usually. The issue isn't a crappy signal, it's either interference, or something else causing your transfer speeds to be slow?

    Try 0 (hardware default) and then 60 for txpower, and see which works best for you. Turn interference mitigation off is usually best. All power saving functions disabled (wireless/APSD mode). In the client, do the same. In short - turn OFF all the unwanted stuff and see if that improves things. Don't use AUTO channel to set the routers channel, stick to 1,6,11 as a general rule.

    Changing to Singapore to get extra channels isn't usually going to help.

    There have been many, many posts regarding this on the forum, so dig around a bit ...
     
  39. HHawk

    HHawk Networkin' Nut Member

    I already tried 0 (hardware default) and 60. It seems 60 is better.

    Turning off 'interference mitigation' results in a really bad signal. Setting it to 'Non-WLAN', improved things, as far as I can tell. Or at least yesterday evening, because the movie didn't stutter once. I have to test this more though.

    I also disabled all power saving functions, same goes for the laptop (as it's connected to a power outlet).

    I am not using the AUTO channel function, I know using the auto-setting is bad for reliable performance. I already checked with InSSIDer which channel would be the best for my router and put it on it. Though it seems every channel in the neighbourhood is taken.

    And yeah there are a lot of different topics; one says A, the other says B and someone else says F. Maybe it's an idea to get a single thread with the ideal settings which are supposed to work best?
     
  40. Planiwa

    Planiwa LI Guru Member

    TXPWR -- I have stated repeatedly that "0 == hardware default" appears to be entirely unsupported by any fact. It appears to be a Tomato error. I challenge anyone to produce any evidence that "0 == hardware default".

    Selecting "0" Tomato, in the expectation of getting "the hardware default" will instead cause wl to report 1496mW.

    No one has actually stated what they think the hardware default actually *is*, for the RT-N16, and actually stated that setting "0" in Tomato will set that default value.

    The whole thing appears to be based on no evidence or reproducible method whatsoever.

    In contrast, the wl command itself clearly states that the hardware default setting in the wl command is "-1".

    I would be very, very happy if someone could "prove me wrong", that is -- produce evidence that "0" means "hardware default" and that setting "0" in Tomato sets that hardware default, and evidence about what that default actually is.


    Interference Mitigation -- I have presented some evidence on this.

    It seems that when wl is allowed to "fix" problems, two things happen:

    1. It "sees" problems that don't exist,
    2. It "fixes" those imaginary problems in a way that creates actual, serious problems.

    In particular, letting wl "mitigate" Non-Wlan Interference causes it to imagine such interference, and then to reduce the transmission speed.

    I have found these settings to be very stable:
    Code:
    chanim_mode            # CHANIM mode: disabled.
    interference           # All interference mitigation is disabled. (mode 0)
    interference_override  # Interference override disabled.
    
     
  41. Toastman

    Toastman Super Moderator Staff Member Member

    There do not seem to be any settings which work "best" for everybody. I think that everyone's situation is different.

    Planiwa - we don't need to respond to your challenges. Perhaps one day instead of trying to provoke people you may come up with something a little more constructive.
     
  42. davidh44

    davidh44 Addicted to LI Member

    I've put Tomato Toastman build on Asus RT-N12 (v1), Netgear WNR3500L, and Belkin F7D7301 (same as F7D3301). Is transmit power of 60 a good generic setting to maximize power on all of these routers without going overboard? Or does the maximum level vary wildly between router brands and models?​
     
  43. Planiwa

    Planiwa LI Guru Member

    If you are concerned, run this command:

    Code:
    wl txpwr1
    
    Whatever you do, do not use the Tomato "default" setting of "0". It is bugged, and no one is fixing this Tomato bug.

    (No one is even acknowledging that this is a Tomato bug.)

    wl documentation clearly states that the default is "-1", not "0".
     
  44. bluenote

    bluenote Networkin' Nut Member

    I think you might be entertained by taking a couple pieces of wireless equipment and running inssider on them while you adjust those values.
    From my experiments, there is very little to be gained by messing with that setting whether you are at 0, -1, or 1496.
     

Share This Page