WNDR3700 Wireless TX Power

Discussion in 'General Wireless Discussion' started by tipstir, Jul 7, 2010.

  1. tipstir

    tipstir Very Senior Member

    Joined:
    Aug 16, 2008
    Messages:
    1,632
    Location:
    South Florida
    I tell these vendors they need to let the average customer know what they're getting for their money. This router the data for TX is hidden. Not anymore.. Again every major vendor as the TX: set @ 2.4GHz very low except for a few other vendors who set to the max: 20 dBm (100 mW)

    2.4GHz 802.11b/g/n
    Tx-Power:16 dBm (40 mW) default

    5.0GHz 802.11a/n
    Tx-Power:18 dBm (63 mW) default

    root@WNDR3700:/# iwlist ath0 txpower
    ath0 8 available transmit-powers :
    0 dBm (1 mW)
    7 dBm (5 mW)
    11 dBm (12 mW)
    15 dBm (31 mW)
    19 dBm (79 mW)
    23 dBm (199 mW)
    27 dBm (501 mW)
    31 dBm (1258 mW)
    Current Tx-Power:16 dBm (40 mW)

    root@WNDR3700:/# iwlist ath1 txpower
    ath1 8 available transmit-powers :
    0 dBm (1 mW)
    7 dBm (5 mW)
    11 dBm (12 mW)
    15 dBm (31 mW)
    19 dBm (79 mW)
    23 dBm (199 mW)
    27 dBm (501 mW)
    31 dBm (1258 mW)
    Current Tx-Power:18 dBm (63 mW)
     
    Last edited: Jul 7, 2010
  2. Log in / Register to remove this ad

  3. antsh

    antsh New Around Here

    Joined:
    May 13, 2010
    Messages:
    6
    Yup, I can confirm those are the stock Netgear values for the WNDR3700. Signal strength and quality, however, cannot be measured just by TX value. I can tell you that with DD-WRT firmware both values are indeed set to 20 dBm, or 100mW, but the stock firmware has both greater range and throughput. I tested this over and over using a high powered AR9220 card in the client.

    Here is an example of the great throughput of this router on the 5GHz band with the laptop about 10 feet (3M) from the router. Notice not only the speed, but consistency. Stock firmware, btw. DD-WRT averages about 20megabits less.


    Done.

    bin/iperf.exe -s -P 0 -i 1 -p 5001 -w 1.0M -f m
    ------------------------------------------------------------
    Server listening on TCP port 5001
    TCP window size: 1.00 MByte
    ------------------------------------------------------------
    [256] local 192.168.1.146 port 5001 connected with 192.168.1.105 port 1763
    [ ID] Interval Transfer Bandwidth
    [256] 0.0- 1.0 sec 10.6 MBytes 88.9 Mbits/sec
    [256] 1.0- 2.0 sec 17.4 MBytes 146 Mbits/sec
    [256] 2.0- 3.0 sec 19.6 MBytes 164 Mbits/sec
    [256] 3.0- 4.0 sec 21.7 MBytes 182 Mbits/sec
    [256] 4.0- 5.0 sec 20.2 MBytes 169 Mbits/sec
    [256] 5.0- 6.0 sec 18.0 MBytes 151 Mbits/sec
    [256] 6.0- 7.0 sec 19.5 MBytes 164 Mbits/sec
    [256] 7.0- 8.0 sec 22.5 MBytes 189 Mbits/sec
    [256] 8.0- 9.0 sec 22.3 MBytes 187 Mbits/sec
    [256] 9.0-10.0 sec 23.6 MBytes 198 Mbits/sec
    [256] 10.0-11.0 sec 23.7 MBytes 199 Mbits/sec
    [256] 11.0-12.0 sec 23.4 MBytes 196 Mbits/sec
    [256] 12.0-13.0 sec 23.0 MBytes 193 Mbits/sec
    [256] 13.0-14.0 sec 22.8 MBytes 191 Mbits/sec
    [256] 14.0-15.0 sec 23.8 MBytes 200 Mbits/sec
    [256] 15.0-16.0 sec 23.0 MBytes 193 Mbits/sec
    [256] 16.0-17.0 sec 21.5 MBytes 180 Mbits/sec
    [256] 17.0-18.0 sec 23.6 MBytes 198 Mbits/sec
    [256] 18.0-19.0 sec 22.8 MBytes 191 Mbits/sec
    [256] 19.0-20.0 sec 22.7 MBytes 191 Mbits/sec
    [ ID] Interval Transfer Bandwidth
    [256] 0.0-20.1 sec 428 MBytes 179 Mbits/sec
    ------------------------------------------------------------
    Client connecting to 192.168.1.105, TCP port 5001
    TCP window size: 1.00 MByte
    ------------------------------------------------------------
    [280] local 192.168.1.146 port 59930 connected with 192.168.1.105 port 5001
    [ ID] Interval Transfer Bandwidth
    [280] 0.0- 1.0 sec 7.98 MBytes 66.9 Mbits/sec
    [280] 1.0- 2.0 sec 17.7 MBytes 148 Mbits/sec
    [280] 2.0- 3.0 sec 18.7 MBytes 157 Mbits/sec
    [280] 3.0- 4.0 sec 19.4 MBytes 163 Mbits/sec
    [280] 4.0- 5.0 sec 19.8 MBytes 166 Mbits/sec
    [280] 5.0- 6.0 sec 20.1 MBytes 169 Mbits/sec
    [280] 6.0- 7.0 sec 20.1 MBytes 169 Mbits/sec
    [280] 7.0- 8.0 sec 20.2 MBytes 170 Mbits/sec
    [280] 8.0- 9.0 sec 19.6 MBytes 164 Mbits/sec
    [280] 9.0-10.0 sec 20.8 MBytes 175 Mbits/sec
    [280] 10.0-11.0 sec 19.6 MBytes 165 Mbits/sec
    [280] 11.0-12.0 sec 20.8 MBytes 174 Mbits/sec
    [280] 12.0-13.0 sec 19.9 MBytes 167 Mbits/sec
    [280] 13.0-14.0 sec 20.3 MBytes 170 Mbits/sec
    [280] 14.0-15.0 sec 20.5 MBytes 172 Mbits/sec
    [280] 15.0-16.0 sec 20.1 MBytes 169 Mbits/sec
    [280] 16.0-17.0 sec 19.4 MBytes 163 Mbits/sec
    [280] 17.0-18.0 sec 19.6 MBytes 164 Mbits/sec
    [280] 18.0-19.0 sec 19.1 MBytes 160 Mbits/sec
    [280] 19.0-20.0 sec 18.7 MBytes 157 Mbits/sec
    [ ID] Interval Transfer Bandwidth
    [280] 0.0-20.0 sec 382 MBytes 160 Mbits/sec
    Done.

    One is uplink, one is downlink.
     
    Last edited: Jul 8, 2010
  4. tipstir

    tipstir Very Senior Member

    Joined:
    Aug 16, 2008
    Messages:
    1,632
    Location:
    South Florida
    Unit has 8 internal ANT how's it going to have greater range with weaker TX with the stock firmware limit 16dBm @ 2.4GHz as just as the ESR-9850 in 802.11n mode 40mW. Your chips AR there I have two of those internal on 2x laptops. 2x netbook have AR also. The others have Airlink N300 USB 1T/2R 2x 2dBi in the USB. I see most test are done with Intel 5300. They need to use AR/Realtek/Broadcom as most users have those. Still Comparing this router with TEW-673GRU which uses the same chips but doesn't have 8 internal ANT but two high power 2.4/5GHz but it's pushing about the same as the WNDR3700 @2.4GHz. The only strong player here is WZR-HP-G300NH with 20dBm @ 2.4GHz on the wireless side or the DLINK DAP with 18dBm @ 2.4GHz and TP-LINK TL-WR1043ND 20dBm @ 2.4GHz.

    Strong the mW better the RSSi to the client adapters. TX stronger transmitter will help those are have weaker signal. Again now everyone going to reflash to DD-WRT, Open-WRT or Tomato firmwares. The stock firmware should be set to the max which is stable to run without drops or disconnection or have lousy throughput in wireless side.
     
  5. antsh

    antsh New Around Here

    Joined:
    May 13, 2010
    Messages:
    6
    The stock firmware is based on OpenWRT, but there a lot of propriety things going on behind the scenes that Netgear refuses to release source code for. For all we know the values returned with the hacked telnet interface are only half the story, or the values are being returned incorrectly. A clue to to this would be that the firmware is saying that the max power is 31dBm. No way, there might be another parameter that Netgear threw in that isn't being reported and is not public code. Trust me when I tell you, that at this point, the stock Netgear firmware performs better under wireless, and routing for that matter, than OpenWRT and DD-WRT. If this was not the case there would be no reason for me to go back to the Netgear firmware. In my case anyway, the stock firmware is performing amazingly, and better than the opensource derivatives.
     
  6. stevech

    stevech Part of the Furniture

    Joined:
    Mar 13, 2010
    Messages:
    5,845
    Location:
    San Diego
    DD-WRT et al has a user interface web page that has many power settings.
    Most hardware ignores attempts to go beyond 40-60mW in 11g/n mode. This is because 11g/n, unlike 11b, uses an RF signal modulation method called OFDM. For detailed technical reasons (called peak-to-average power ratio), the average power of OFDM must be reduced (by about 5dB) as compared to the signal produced by 11b which is not OFDM. This is called OFDM back-off.

    So while DD-WRT et al may offer you a choice of higher power settings, the radio module ignores settings which would violate the OFDM back-off requirements. Violating this will cause the chipset to produce a distorted transmit signal which leads to higher error rates and thus lower throughput (speeds). For this reason, the WiFi alliance has pass/fail criteria for products' transmitted signal distortion. The FCC cares not if the signal is distorted with high error rate - their job is interference management.

    A few non-consumer products can get to 100mW or more and not violate this waveform distortion limit (called the rho), but these tend to cost more. The power amplifier in a WiFi product is the most expensive part, and with OFDM in 11g/n, most vendors chose to keep the same amp they had for 11b rather than adding cost. Curiously, the reduced average power for OFDM (about 5dB) is offset by the signal processing benefit of OFDM over DSSS used in 11b. So it's a wash in terms of range, even with the lower power in OFDM modes.

    ALSO
    The difference between 50mW and 100mW is 3dB. A common transmission path, indoors, has about 50-70dB of loss, due to RF principles. So that 3dB increase is really negligible. And worrying about the router's power ignores the fact that the client device's transmitter needs to be of equal strength, unless there is a high gain antenna at the router. Image a pair of walkie-talkies, where one has a lot more transmit power than the other. The range is not balanced.
     
    Last edited: Jul 9, 2010
  7. tipstir

    tipstir Very Senior Member

    Joined:
    Aug 16, 2008
    Messages:
    1,632
    Location:
    South Florida
    DD-WRT on the Smart Router (Buffalo WHR-HP-G54) can go beyond those limits for 802.11g using DD-WRT. It can do 100mW, I've tested it since 2006 the two I have can do it. I don't know about the current crop after the court settlement so those might not be able to exceed those limits you had mentioned. Still well all want the strongest signal the unit can provide. What's on paper might be crack with DD-WRT, if they test it right on the sample routers that are sent to them from users all around the world. Still 802.11n and DD-WRT doesn't play nice on some of the routers DD-WRT supports. 16dBm, 17dBm, 18dBm, 19dBm and 20dBm are mostly what you see in home wireless routers. The vendor who produces these routers have to rely on the chip maker to produce the speed and signal the unit can handle. Firmware on stock units can't be change if the chipset maker has put in limiters. That's why a lot of these units don't put out as you think they do. I can't see spending so much on a wireless router then find out the unit can't cover the square feet you want it do. I've also noticed that some wire routers, wireless access points and CB don't play nice with each other and that's another issue to be continued!

    What gets me is the fact the thinking of the vendor of these wireless routers, are removing a lot of advanced features thus you get a limited feature router where you can't change the WAN port speed and some other neat features like LAN port speed. These routers should have these features if you going to spend over 150 bucks.

    I've ask the makers of ESR-9850 if some features could be implemented in to the firmware. They said yes but, here it comes, that most home users won't need to use such features it would just confuse them. Too bad. So everything is auto except the wireless security and wireless mode. Minor tweaks.
     
  8. thiggins

    thiggins Mr. Easy Staff Member

    Joined:
    May 18, 2008
    Messages:
    9,809
    Tipstir: If you want to continue to harp on your "more power is better" theme, please do it on your own Forum.

    I've shown here that simply cranking up power can actually degrade performance.

    Stevech obviously knows what he is talking about. You'd do well to pay attention.
     
  9. DaveMcLain

    DaveMcLain Regular Contributor

    Joined:
    Jun 24, 2009
    Messages:
    167
    Isn't one of the reasons distorted output is to be avoided in a transmitter is because if the waveform is distorted by a high percentage the distortion artifacts are actually harmonics of the fundamental frequency? These harmonics can produce interference that's "out of band" for the device producing the distortion and thus cause a lot of problems with interference with other devices which are seemingly unrelated...
     
  10. tipstir

    tipstir Very Senior Member

    Joined:
    Aug 16, 2008
    Messages:
    1,632
    Location:
    South Florida
    I do agree with you cranking up the power can cause issues. I've tested that myself on Netgear wireless N and reported the issue with DD-WRT. But they seem not to care about that. I think 50mW to 100mW range is good enough for the average home user. If you need to go higher from 600mW to 1 watt range then you should look else ware for hardware that more on the professional level.

    I've been this too long now over 32 years for me as software developer in this industry. We all can learn from each other. My own business in the real world is now going 10 years in Network Services here which I run daily. I know what I am talking about (might not be in the right terminology) as you do here with knowledge. Well enough said on this matter.

    As for my forum it's there for helping those who need more help on certain nodes, OS, software and FB. I am glad to provide such a service for free. I've been at that since 1991 on BBS & SVP well not the forum but helping those with all sorts of in this industry.
     
    Last edited: Jul 9, 2010

Share This Page