Forum

  • By (Deleted User)

    Hi,
    We have an AP (Dlink 2700) attached to a Sectoral (90 degree) on a 200 ft tall building sending signal to another AP via a wireless client attached to an 8db. dish antenna. The two POP's are @ 1 1/2 miles apart. It is clear line of sight over mostly land with a bit of beach/shoreline.
    At times we'll do 1.9 meg other times .7 meg. with no load. Pings will stay consistent but disconnects will occur for @ 2-3 minutes sometimes.
    There are a few P-P signals in the area and a few other AP's.
    My question is whether these dropouts and disconnects might be from interference, Antenna's, water in air, Alignment or?
    Help greatly appreciated,
    Tim

  • By (Deleted User)

    Did you test these devices during a site survey? I think this is critical before actually placing these new APs that have not been readily used in enterprise solution scenarios.

    Many environmental factors to consider. Depending on the clients that are connecting (802.11/b/g) and the number of other 2.4 Ghz devices (telephones) operating in the area you could suffer from interference on the channel you are using.


    Are your cable connections tight? Is the average rain fall high, and causing a reflection and VSWR?

    But then again, it could be the device itself and the firmware that needs an upgrade or return to the factory. (Warranty?)

    Try using an RF/Protocol analzyer to test your spectrum, if you can get one ($$$$). NETSTUMBLER might show you something that helps.

    See what is in the area and what channel they are operating on.


    Also:
    See this forum posting about dlink drops. Not bad mouthing DLink. I think they have some pretty good products.


    I have one of their devices a DI-774 that works well inside the house.

    http://209.123.109.175/forum/remark,11245432~mode=flat?hilite=drops


    got my CWAP book and studying (boy it is exactly what Devin and Kevin said....TOUGH STUFF!)

    compughter

  • By (Deleted User)

    Did you test these devices during a site survey? I think this is critical before actually placing these new APs that have not been readily used in enterprise solution scenarios.

    Frankly, no, we didn't site test. When you first get the visual you think it has to work (perfect Line o sight). Tallest buildings in the area...
    I didn't even have a clue about the suspect RF signals lurking there in the ether. I'm still unsure.

    Many environmental factors to consider. Depending on the clients that are connecting (802.11/b/g) and the number of other 2.4 Ghz devices (telephones) operating in the area you could suffer from interference on the channel you are using.

    On the first hop is a 2700 serving @20 wireless Dlink (b) cameras. They are sketchy sometimes, othertime work great.
    The second hop is the 2700 @1 1/2 miles further on another tall building. the backend of that is another sectorial.

    Are your cable connections tight? Is the average rain fall high, and causing a reflection and VSWR?

    Yep believe so.

    But then again, it could be the device itself and the firmware that needs an upgrade or return to the factory. (Warranty?)

    Yes, I agree with you here. The Dlink 2700 AP (g) device, is relatively new. They like to do the first to market thing and have you prove their product in the field. We've been waiting for firmware for the longest. The repeater mode doesn't work with certain end user chipsets! Imagine that.

    Try using an RF/Protocol analzyer to test your spectrum, if you can get one ($$$$). NETSTUMBLER might show you something that helps.

    Don't have the budget for that at the moment.

    See what is in the area and what channel they are operating on.

    Tried every channel combination possible. Incidentially we have amps on our AP's. And I don't believe any other AP's nearby do, so...
    One would think we wouldn't have issues, maybe?

    Thanks for the reply...
    Tim

  • By (Deleted User)

    Forgot to post name with previous

  • By (Deleted User)

    if there was some way you could rent or borrow an analyzer with a capture that would be great. maybe you could correlate it and ensure something is not going on with the visual loss.

    i would definately check everything from the dish horn all the way to the cable entering the building and ensuring all is secured properly.

    what type of antenna is being used? how were they aligned?

    are all the bits part of a system? do the components match as far as settings, resistance (ohmage), load, power, etc?

    here is something for free...disconnect all the bits until you have one ap, the cable going to the antenna, the antenna at the far side, and its ap. attach one client at each end only. if all is good, then something on the backside of your network, a connection or load, might be affecting the overal stability. that might help segment things for free.

  • I would suspect you have a low S/N ratio. I would check it at both ends, and using an amplifier will generally amplifiy the noise too unless it is of good quality LNAmp.
    At 200 feet there is a lot of rf noise from antennas miles away. Do you really need a 90 degree antenna (it will pick up more noise). Can you do for something less 60 or 30 degree.
    Plus I looked up the 2700 and it doesnt look like it is necessarily your best pick for that distance. I also didnt see an amp for the 2700, if you are using an amp not tested with the 2700 and certified by the FCC then you may be in violation of the FCC rules. You cannot use an amp unless it is sold as a system and tested as such. Now, maybe I just didnt see the amp on the dlink site, but you should know the rules anyway. A good site to spend the day at is the YDI site at www.ydi.com .

    Good Luck

  • By (Deleted User)

    Thanks guy for the great suggestions,
    In regards to:
    I would suspect you have a low S/N ratio. I would check it at both ends, and using an amplifier will generally amplifiy the noise too unless it is of good quality LNAmp.
    At 200 feet there is a lot of rf noise from antennas miles away. Do you really need a 90 degree antenna (it will pick up more noise). Can you do for something less 60 or 30 degree.
    Plus I looked up the 2700 and it doesnt look like it is necessarily your best pick for that distance. I also didnt see an amp for the 2700, if you are using an amp not tested with the 2700 and certified by the FCC then you may be in violation of the FCC rules. You cannot use an amp unless it is sold as a system and tested as such. Now, maybe I just didnt see the amp on the dlink site, but you should know the rules anyway.


    I don't know why your insinuating that we're not complying with FCC rules. We happen to be using Amps sold online of which are in compliance with FCC.
    They are not Dlink products but....
    After all this we just found the Telco supplying the backhaul, has been sending us bad E1. The signal has cleaned up and we're doing 1.9 at the hops and end-users throughout!
    We've been pulling hairs thinking it was us. At the router/switch we've always done 1.9 meg. Wirelessly from there on out the problem manifested from the bad Telco feed.
    What a relief!
    Thanks again,
    By the way we have one customer 18 miles away sourced from a 2700AP, (no affiliation with Dlink) Tiltec Sectorial to 16 db grid antenna.

  • By (Deleted User)

    Good post and lesson learned. I am glad it worked out for your customers. How did they (TELCO) discover the bad T-1, might I ask?

    Thanks again,
    compughter

  • By (Deleted User)

    t-1 and e-1 serial lines are coded in a certain way, with bits in each frame providing line condition information, to include error rates and up/down status. telcos should see a red condition in their logs/systems/indicators. along with that, one can physically place a bert, bit error rate tester, one each end of the line, pass a pattern, and determine integrity, clocking, etc.

  • By (Deleted User)

    Thanks again Gents,
    To start, we're in Baja Calif. The Telco is Telnor.
    They can be dependable but for the most part their terrible in servicing customers.I think we simply complained long enough to get their attention, in order to fix the backhaul feed.
    Very interesting bits of hardware you mentioned, eeyore. We will most definitely look into acquiring something similar to complete the whole WISP picture.
    Regards,
    Tim

Page 1 of 1
  • 1