Tomato Firmware: Increased Transmit Power Does More Harm Than Good?

Currently, I run four routers, all part of the WRT54G/GL/GS series, that allow a house to receive wireless internet.  These routers provide sufficient, although not yet ideal, coverage for the house.   Since the WRT54G/GL/GS routers are built on a Linux framework, they are easily upgraded to more powerful firmware, such as the Tomato Firmware available from http://www.polarcloud.com/tomato.  Third-party firmware such as Tomato adds increased functionality to what would otherwise be considered a standard router.  One of the more interesting features that Tomato provides is the ability to increase the transmission power of the wireless signal.  The picture below shows this setting.

Tomato Transmit Power Setting
Tomato Transmit Power Setting

Recently, users had been mentioning that the Internet was unstable, with frequent reconnecting and low signal strength.  These specific users’ rooms are at the extreme ends of the house, which means that their Internet signal is the weakest.  I increased the Transmit Power of all the routers thinking this would be solution to the problem.  It was not.  As the users kept saying the Internet was unstable, I would increase the transmission power of the router.  I found that the recommended maximum transmission power level is 82 mW.

From Tomato Wikibooks : Transmit Power: Sets the transmit power in milliwatts. High settings may cause nonlinearity in the transmitter causing loss of data, interference to other users and channels, and a high “noise floor”. It may also overheat and shorten the life of the transmitter. Tomato default is 42mW, using 84mW is usually safe.

I had turned the routers up to 80 mW and the problem was still occurring.  Upon some research I found that as the Transmit power levels of the routers increased, so did the radio noise produced by the routers. I decided to turn the routers back down, to a level lower than when I first began to administer the network.  After this change, users noticed an immediate improvement in the quality of their wireless connections.  Everything was (and still is) running stable.  The users are no longer losing Internet connection and are maintaining proper network speeds.

In my personal experience, although the Transmit Power feature of the Tomato firmware may seem to be the answer to wireless coverage or wireless signal strength problems, it is not.  Not only does increasing the wireless transmission power of these routers cause more radio noise, it also reduces the life of the router.  On these routers, the settings were only at 80 mW and I was having problems.  The Tomato firmware allows a maximum setting of 250 mW.  Users with this setting will experience an unstable wireless signal, and will shorten the life of their routers.

Have you experienced any similar problems with your router?  What transmit power level are you using for your router?  What other steps have you taken to increase the signal strength of your router?  Let us know by commenting below.

  • Ian

    I think the reason why this is is because the clients radios are not increasing their power, and because of that, the effective range is limited by the transmit power of the clients radios. However, I have found increasing this setting to 84 to be useful when setting up WDS to repeat the network over longer distances. Obviously, since you have two devices that are communicating with each other and BOTH are transmitting at a higher output, it would make sense that the range is increased, which it seems to be.

  • Ian

    I think the reason why this is is because the clients radios are not increasing their power, and because of that, the effective range is limited by the transmit power of the clients radios. However, I have found increasing this setting to 84 to be useful when setting up WDS to repeat the network over longer distances. Obviously, since you have two devices that are communicating with each other and BOTH are transmitting at a higher output, it would make sense that the range is increased, which it seems to be.

  • Tristan

    The stock linksys setting is 43mW which, in my opinion, seems rather low. I noticed a significant boost when upgrading from this stock value to 70. I would never recommend jumping above 100 unless you want a dead router in a short amount of time.

  • Tristan

    The stock linksys setting is 43mW which, in my opinion, seems rather low. I noticed a significant boost when upgrading from this stock value to 70. I would never recommend jumping above 100 unless you want a dead router in a short amount of time.

  • Great writeup. I love the Tomato firmware, and this setting was very well explained in your post. Thanks!

  • Great writeup. I love the Tomato firmware, and this setting was very well explained in your post. Thanks!

  • Mark

    I am not sure if I have just an older router or it is in fact because I run Tomato at 84. Users do still have a weak signal (1 bar) from only about ~50 away. I will try another router, as I have access to many (including brand new wrt54gl's)

    Thx

  • Mark

    I am not sure if I have just an older router or it is in fact because I run Tomato at 84. Users do still have a weak signal (1 bar) from only about ~50 away. I will try another router, as I have access to many (including brand new wrt54gl's)

    Thx

  • Mike

    Nice write up, as for my two cents…"when in doubt, directional antenna." I'd say replace the client antenna with a slightly higher db one. Also many wireless adapters allow for an overdrive on transmission power, or in the least deactivation of wlan power saving options in portables. As for the weak bar at 50 feet I've had this exact issue, and it turned out one of the anteana's on the router was not making good contact. So i did some experimenting and it seems (this is from observation only, not fact) That if you drop the left, you get bars, but horrible connections, and the right, you get stability but no range. I do wonder if the writer has these bridged or simply each on their own hardline. My next curiosity is what channel's he's using for each one. I've noticed some sites do better with certain channels, simply going from 2 to 9 can make a huge difference in specific locations.

    • Each one of the four routers were on their own line. I also had different channels set for each one.

  • Mike

    Nice write up, as for my two cents…"when in doubt, directional antenna." I'd say replace the client antenna with a slightly higher db one. Also many wireless adapters allow for an overdrive on transmission power, or in the least deactivation of wlan power saving options in portables. As for the weak bar at 50 feet I've had this exact issue, and it turned out one of the anteana's on the router was not making good contact. So i did some experimenting and it seems (this is from observation only, not fact) That if you drop the left, you get bars, but horrible connections, and the right, you get stability but no range. I do wonder if the writer has these bridged or simply each on their own hardline. My next curiosity is what channel's he's using for each one. I've noticed some sites do better with certain channels, simply going from 2 to 9 can make a huge difference in specific locations.

    • Each one of the four routers were on their own line. I also had different channels set for each one.

  • I have my router running at 80 mW. As an experiment I did boost it to over 150 mW and some of my wireless devices (a wireless webcam for example) lost connection from time to time. At 80 mW the problem ceased.

    To get better coverage, an antenna with more gain is recommended, as it amplifies the signal in both directions (transmitted and recieved)

  • I have my router running at 80 mW. As an experiment I did boost it to over 150 mW and some of my wireless devices (a wireless webcam for example) lost connection from time to time. At 80 mW the problem ceased.

    To get better coverage, an antenna with more gain is recommended, as it amplifies the signal in both directions (transmitted and recieved)

  • I wonder if the interference is caused by the hardware itself and not Tomato. When setting the transceiver power, my guess is it changes voltage to only one area on the motherboard, and the fact that the WRT54G, as good as it is at accepting third-party firmware, is a really cheap router that muddies up the signal. Linksys isn't going to use top-end, noise-reducing components in a low-end router, especially one not rated for the power output Tomato is pumping into it.

    It'd be interesting to try it with another firmware (DD-WRT) and see if the same happens, or in a router with different components. Unfortunately, it seems like most routers capable of supporting third-party firmware that are not WRT54* routers are in the $100s range. Plus more expensive doesn't always mean better components, it usually just means better firmware =).

  • I wonder if the interference is caused by the hardware itself and not Tomato. When setting the transceiver power, my guess is it changes voltage to only one area on the motherboard, and the fact that the WRT54G, as good as it is at accepting third-party firmware, is a really cheap router that muddies up the signal. Linksys isn't going to use top-end, noise-reducing components in a low-end router, especially one not rated for the power output Tomato is pumping into it.

    It'd be interesting to try it with another firmware (DD-WRT) and see if the same happens, or in a router with different components. Unfortunately, it seems like most routers capable of supporting third-party firmware that are not WRT54* routers are in the $100s range. Plus more expensive doesn't always mean better components, it usually just means better firmware =).

  • scrook

    It is worthy to note that an increase from default 42 mW to 84 mW (twice the power out, also twice the heat generated in the final RF amp transistor) will only increase your signal 3 dB, which may be enough, but it’s not a profound difference. Likewise, a decrease to 21 mW will only reduce your signal by 3 dB but half’s the power dissipated by the final amp. All things in moderation…

    A pair of longer, higher gain omni antennas (7 – 9 dBi) vs. the factory antennas (5 dBi??) would get you another 2 -4 dB (if I recall the stock antenna’s gain correctly), though closer to perpendicular tothe antenna axis, as the gain is achieved by compressing the radiation pattern vertically.

    Increasing to the max of 250 mW is only a 7.7 dB increase in signal, but nearly 6X the heat dissipated in the final amp (not good for life/reliability of the transmitter).

    • Wbarfels

      Actually an Increase of 3DB is twice the signal power…It’s not a Linear Scale

  • Thanks for the great input!

  • me,not,u

    Once you learn about ghosting and what causes it you will then realize RF energy is one wild child and lives by it’s own rules. By increasing the level of output power your signal bounces off the structures around the house and causes what you refer to as “noise” when in fact it is not necessarily “noise” at all.. It is the RF signal interfering with itself.. turning down the power reduces the ghosting effect thus clearing up the issue..

  • guest

    I have had a WRT54 router online for over 3 years with the tomato software and set at 105 milliwatts. I have not reset the router in 2 years. I have no problem with disconnects or unstable transmissions. I had a separate access point in my house before using the tomato software to turn up the signal. I have not had the same experience as Dustin had. Turning it up worked well for me.

    • Nunya

      Same here. Turning up the transmit power solved all of my weak signal woes. I really do not care if it shortens the life as these things are cheap and easy to find.

    • SuperFly

      Same here, turning up the transmission power solved all of my weak signal issues. Now everything connects at 54mbs and stays that way. 

      • Toniballini

        What a load of nonsense. You guys probably believe in witches too.   Go learn some physics!

  • The channels I was using for this setup were 1, 6, and 11, with two of the four using the same channel further away from each other.  In addition, this was in a more-populated neighborhood with several other wireless access points from neighbors in the area, causing even more interference. 

  • Anonymous

    You may have very well answered the problem of many routers having issues these days even with manufacture firmware. Maybe router makers are simply setting the routers up for a problem by increasing output to compete and make customers happy with better advertised coverage. Knowing full well of the pitfalls of shortened lifespan and unstable connection. It may very well answer why some initial review are positive but then turn negative after a few months. I find that very true as my Linksys E1200 drops WiFi every couple weeks but continues to have it Wired connections work just fine.
    Its like anything that you max out in power for a long period. Its just going to fail early.

    • Tommy Neill

      Here’s something interesting, I just took tomato off and put sputnik’s ddwrt brand on and my e3000 temp has gone way down. I’m running wl0 at 120 and wl1 and 30 ( I don’t use the 5ghz anyway )