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

Buffalo WHR-HP-G54 and Tomato 1.07 performance. Info Thread.

Discussion in 'Tomato Firmware' started by BHTeam, Sep 4, 2007.


Was this Thread usefull to you?

Poll closed Nov 3, 2007.
  1. Yes.

  2. No.

  1. BHTeam

    BHTeam LI Guru Member

    I want to share with all of you my findings about the "power" potential of
    Buffalo AirStation WHR-HP-G54.
    The default transmit power setting on Tomato is 10mW.
    After some research on Buffalo AirStation WHR-HP-G54, I established
    that it's end step of amplifying circuit is an Anadigics AWL6153.
    Anadigics AWL6153 have an output power levels of
    25 dBm for 802.11g witch is approximately 316 mW, and
    28 dBm for 802.11b witch is approximately 630 mW.
    This output can be achieved by suppling Anadigics AWL6153 with 5.0V.
    Anadigics AWL6153 can run on 3.3V, witch will provide
    21 dBm for 802.11g witch is approximately 130 mW, and
    25 dBm for 802.11b witch is approximately 316 mW.
    I really don't know what is the default voltage of Anadigics AWL6153,
    but there are some mods on the internet for increasing chips voltage
    from 3.3V to 5.0V.

    I've tested changing the power settings my Buffalo AirStation WHR-HP-G54.
    Used Hardware:
    1. Buffalo AirStation WHR-HP-G54
    2. Dell Latitude 120L (Intel® PRO/Wireless 2200BG miniPCI card)
    3. Custom made Asus P5WDG2 WS Pro PC.
    Used Software:
    1. Real-Time Bandwidth monitor of Tomato v1.07
    2. NetStumbler® v0.4.0 (Build 554)
    3. DC++ v0.674 on my ISP's local hub.
    4. A 538 MB avi file.

    Terrain definition:
    Wall - 20 cm (7.9 inches) of concrete with structure steel inside.
    Point 1 - 40 cm (1.31 ft) from the router. Clear sight of vision.
    Point 2 - 4 m (13.12 ft) from the route. Signal going trough one Wall.
    Point 3 - 12 m (39.37 ft) from the router. Signal going trough two Walls.

    Power settings:
    1. 10 mW Tomato's default setting.
    2. 130 mW Anadigics AWL6153's maximum power on 3.3V.
    3. 251 mW Tomato's maximum setting.

    And here are the results summarized in exel.

    First - Noise Floor chart. Closer to -100 the better.

    Second - Signal Strength chart. Closer to 0 the better.

    Third - Speed comparison carts. The higher, the better.
    PEAK value is the highest value achieved during the test. The average value is more important.

    10 mW Chart.

    130 mW Chart.

    251 mW Chart.

    Question number one:
    If Buffalo's default voltage of the Anadigics AWL6153 chip is 3.3V,
    what will happen when I set Tomato to a higher value then 130mW?

    Nothing. The router will work with no problems. Even when going to 251 mW
    the router performs as intended from manufacture. No jamming, needs of
    reset or any other unusual stuff.

    Question number two:
    Is there any change in performance after going to 251 mW?

    It depends on your terrain and the frequency load around your location.
    You can see that after increasing the power output to 251 mW, there
    is a significant increase in signal strength. I really can't do the test
    in a frequency quiet environment, so you make your own conclusions from
    the charts.
    There is no way to set Tomato to value higher then 251 mW.

    Question number three:
    Is there a temperature change after setting 251 mW power output?

    No. The router's casing has exactly the same temperature on 10 mW and 251 mW.
    I've not opened my router to measure the exact temperature.
    I've set my router to 251 mW since June 2007. Exactly the first day i bought it.
    And i have not experienced any problems with it.

    Here is a link to all screen shots a made of NetStumbler and Bandwidth monitor of Tomato.
  2. GeeTek

    GeeTek Guest

    That is excellent work ! Thanks for posting this info. I am going to set up a test radio to run 251 all the time and see how well it holds up.
  3. u3gyxap

    u3gyxap Network Guru Member

    The Anadigicis amplifier on the HP is powered by 4.5V, not 3.3V. You can go as high as 5V, since it is already available on the board, but the power output increase is insignificant.
  4. Toxic

    Toxic Administrator Staff Member

    AFAIK 251mW is the maximum setting of the wl driver and not a limitation of the "Tomato" firmware setting.
  5. jowahl

    jowahl Network Guru Member

    thank you for the very usefull report.
    did you use default boardflags? changing the boardflags may change the results dramatically ...
  6. BHTeam

    BHTeam LI Guru Member

    I used default board flags.
  7. GeeTek

    GeeTek Guest

    The board flags change only the receive pre-amp, not the transmit amp.
  8. jowahl

    jowahl Network Guru Member

    i know, but a better receiving power may cause a better signal strength and maybe higher speed. or am i wrong?
  9. GeeTek

    GeeTek Guest

    You are correct. Depending on environment conditions and other factors, sometimes bandwidth is better with the pre-amp and other times better with out. It is a great suggestion, but the results would tend to be specific to that test environment and not be very applicable to the pre-amp yes/no question as it applies to general deployment.
  10. guillaumy

    guillaumy LI Guru Member

    I'm still wondering how this is all working in Tomato 1.07.

    I mean, the signal is perfectly serviceable at the default setting of 10 mW. How on earth can it be set to 251 mW and still work! What is the REAL effect of this on the "insides", e.g. the pre-amp TX signal, the post-amp TX signal?

    Personally, I am piqued but I am still waiting for confirmations and reports from other users, before I dare to set 251mW in field applications (where I want to be 100% sure that things will continue to work, say, at least a year from date of deployment).

    Thank you BHTeam!
  11. guillaumy

    guillaumy LI Guru Member

Share This Page