Home AMXForums Archive Threads Residential Forum

Recomended WAPs for MVP8400

gerrad.ferriergerrad.ferrier Junior MemberPosts: 7
Hi,

Does any one have any suggestions of what type of WAP is best to work with MVP8400 touch panels?
I am currently using the Netgear WG102 WAP (there enterprise version) and it is very good and really versatile but I have problems with the tp's randomly falling off line. the install is in the country where there is pretty much no interference from outside. The system has been extensively surveyed and set up correctly so as not to conflict with it's self. so i dont think there is a problem with the WAPs or the network.

The tp's all have G cards and are running the latest hotfix firmware that addresses roaming issues, this stopped the problem of them randomly locking up when moving from one access point to another, but they still fall offline.

the main problem is poor reception on 1 of the tp's, it has been sent back to AMX for service and they just said there was nothing wrong with it. If I sit 2 of the panels next to each other the 1 I sent back consistently has a worse connection rate or is not connected.

Basically I have run out of options as to what the problem might be. Has anyone had any dud wireless cards?
I have done all the usual tests without encryption etc with no affect.

this problem of the panels falling offline happens to us on a few jobs and is very frustration for us and the client. The only common factor is the tp's as we have used other WAPs in the past.

are there any aerial modifications that we can make? it is really frustrating when my laptop can connect perfectly any where on the whole property and the TP just falls offline!

Any help or suggestions is welcomed

Thanks

Gerrad

Comments

  • viningvining X Member Posts: 4,368
    1st thing you should do is download NetStumbler. Free!

    Once you have it running do a site survey and verify yourself that all WAPs are on different channels and using the same SSID.

    Follow the AMX guideline for channel shifting using channels 1,6,11. (although I don't)

    There's an Active Roaming on Channels Option on the TP's (protected setup>wireless setting) which should be enabled.

    Then verify you dbmv levels on the WAPs the ensure they are operating properly by bringing your laptop to the vincinity of each WAP to ensure it's power output is where it should be as compared to the other WAPs.

    You can also disconnect all WAPs and then one at a time connect one and verify encryption is correct by connecting through it w/ your laptop.

    You can also remove all encryption until you resolve this issue.
  • DHawthorneDHawthorne Junior Member Posts: 4,584
    I doubt it has anything to do with the brand or model of the access point. I've seen this kind of thing with just about all of them; it's an inherent problem with multiple access points. All wireless devices have a tendency to drop one and pick up another that's in range.

    This is what I have found that works the best; of course, there may be factors in your job that make it different, I don't suggest it's the only answer, or even the right one for you, it's just a suggestion:

    1) Make all your IP's static. Give all your masters and all your panels static IP's.
    2) Give all your access point unique SSID's, and set each panel to connect only to the strongest SSID. Even if it does pick up another, it will stay connected to the one defined, but if all your access points have the same SSID, it won't be able to do that.

    As often as not, it's the WAP that lost connection with the rest of the server rather than the panels losing connection with the WAP. This can be easily checked by looking at the Wireless setup page ... it will show the WAP signal strength even when it's not connected to a master. I've found that this is often a DHCP server issue, and the static IP will usually get around it.
  • gerrad.ferriergerrad.ferrier Junior Member Posts: 7
    Thanks mate,

    I have done all that, been using net stumbler for about 3 years, very handy.

    only thing on your list that I dont know about is the roaming, what version of firmware is it on? I am using v2.73.2 it isnt an official release as yet but it was done as a hotfix to address roaming issues.
    The panels i have are MVP 8400 not 8400i. Can i use the 8400i firmware?

    I have 7 WAPs onsite so 1, 6, 11 isnt an option unfortunately as i dont have enough physical separation
  • gerrad.ferriergerrad.ferrier Junior Member Posts: 7
    Thanks Dave,

    All IPs are static.

    Trouble is the panels have to be able to roam around a garden so I cant fix them to 1 access point, which means the SSID must be the same. I did try doing it with separate SSIDs and it was really good until you walked too far away!

    Thanks to you and Vinning for the suggestions

    Cheers

    Gerrad
  • gerrad.ferriergerrad.ferrier Junior Member Posts: 7
    By the way do you guys use WiSpy?

    http://www.metageek.net/

    it is a really good tool for analyzing the 2.4GHz spectrum. you can pick up interference like microwave ovens etc.

    cheers

    Gerrad
  • travtrav Get off my lawn Posts: 188
    By the way do you guys use WiSpy?

    http://www.metageek.net/

    it is a really good tool for analyzing the 2.4GHz spectrum. you can pick up interference like microwave ovens etc.

    cheers

    Gerrad


    Always. The Static IP thing is a good suggestion, I've seen the linux stack fail on other devices not just AMX ones with problems like that.

    Also, do you have a site map and positions of the WAP's I could have a look at ? If you have 7 waps, the site must be very large, and I'm having trouble understanding why you can't have WAP's on the same channel seperated by ones on a different channel.

    But then, you've seen the site and I'm just making it up as I go along ^_^ There have been lots of discussions about using non AMX waps, mostly the ones I've seen work well with roaming have been the ones based around cisco (linksys and , level one) If you have the netstumbler and wispy recordings and a site map I'll have a look at them. I always find that handy to have a seperate set of eyes to look a problem over.
  • viningvining X Member Posts: 4,368
    DHawthorne wrote:
    2) Give all your access point unique SSID's, and set each panel to connect only to the strongest SSID. Even if it does pick up another, it will stay connected to the one defined, but if all your access points have the same SSID, it won't be able to do that.

    Now I've always found it to be the opposite when roaming with laptops or TPs and I believe that if you are associated with an SSID you have actually drop connection before it can re-associate with another SSID where as (in theory anyway) if all are on the same SSID they are supposed to associate with the channel providing the strongest signal depending on defined firmware protocols for doing so. So that you're only changing association if there is a big enough difference between levels to warrant a switch.


    No, but I've been meaning to get a Wi-Spy!
  • DHawthorneDHawthorne Junior Member Posts: 4,584
    vining wrote:
    DHawthorne wrote:


    Now I've always found it to be the opposite when roaming with laptops or TPs and I believe that if you are associated with an SSID you have actually drop connection before it can re-associate with another SSID where as (in theory anyway) if all are on the same SSID they are supposed to associate with the channel providing the strongest signal depending on defined firmware protocols for doing so. So that you're only changing association if there is a big enough difference between levels to warrant a switch.


    No, but I've been meaning to get a Wi-Spy!

    You see, that's why I am always careful not to put forth my experiences as "the definitive way" to deal with something. I have a job with 5 WAP's in it, and probably at least 2-3 more to come. It's a very busy network, and wasn't wired wtih the use it's getting in mind, so I have to share my control network with the house network ... which, by the way, mirrors the clients office server, and all the traffic resulting from that.

    When I set up this system with all the WAP's on the same SSID, the panels were continuously dropping off line, connecting to another WAP, dropping again, going back to the first. Even without physically moving the panels. When I gave each WAP it's own SSID, this problem went away. Clearly, there are other factors involved, and in this case, I know for certain I have a background noise problem - I had a heck of a time even getting cordless phones that would work in the next room over from their base. I suspect it's the huge high-tension power lines cutting across his south 40. But separate SSID's was the only way I could make this system work. Ironically, he does continue to have trouble with his PC's disconnecting,but the panels are stable.
  • viningvining X Member Posts: 4,368
    DHawthorne wrote:
    he does continue to have trouble with his PC's disconnecting,but the panels are stable.

    With that set up you would have to completely drop a connection before the laptop would go to the next SSID in the connections list. So depending where they are they can have a very tenuous connection with extremely low through put. I take it the panels don't move around to much in which case seperate SSIDs would be fine.

    Maybe you do have sort of weird EMI, RFI thing going on or maybe the WAPs have fluctuating and inconsistant transceivers. Who knows? But switching back and forth from WAP to WAP when the TP is stationary is odd. Do you remember what the levels were? I suspect you may have been evenly spaced between the two and wouldn't rule out that the MVPs hand off protocol isn't at fault.
  • ericmedleyericmedley Senior Member - 4000+ posts Posts: 4,177
    Coming from a large university campus with campus-wide wireless, I believe the standard is to keep SSID's the same. I have found that panels do seem to search around and will even go offline breifly while they reconnect. These do not result in much problem as apposed to the panel trying in vain to hold on to an old WAP that has just gone too low.

    All our systems use consistent SSIDs throughout. Is there some kind of standard for this? I've heard both sides of the argument. What I can say is that since we've switched our networks over to 'all SSIDs the same', overall stability has improved greatly.
  • jweatherjweather Junior Member Posts: 320
    I'll just mention that any campus-wide wireless system worth its salt is a little more sophisticated than just "all SSIDs the same". If you take two consumer APs and give them the same SSID, then yes, your panel should roam between them, but it involves disconnecting from one and reconnecting to the other at the crossover point. Cisco et al. provide wide-area "mesh" access points that allow seamless roaming between APs. I'm not familiar with the exact details, but I know it involves making all the APs appear to have the same MAC address so the client never needs to know that it is roaming. Of course, if the client wants to pay as much for their WLAN as they did for the AMX system, they can have the same network...

    Jeremy
  • CT-DallasCT-Dallas Junior Member Posts: 157
    Gerrad,
    Do you have any cordless phones in the install? We recently performed a test while a client was out of town. The RMS log indicated that their panels went offline close to 200 times in a 5 day period. Most of the time, it was for 2-3 seconds, and sometimes the panels could not come back online and would need to be manually rebooted. Environmentally, there are 3 WAP's in the system, and they are all on different channels. The WAP's reside less than 15 feet from the virtually stationary 8400's.

    We turned off the phones for one week. We then turned off the phones and unplugged the panasonic cell stations for another week and compared the logs against the connectivity problems pre-changes. To our surprise, the drop off's in the first week of testing went from roughly 200 to 5. The second week almost had no drop offs at all. That was quite convincing that the phones played a major part in our problem. Unfortunatly, panasonic only sells 2.4 ghz system cordless phones - so the two must be able to coexist in this installation.

    Even though the phones use frequency hoping, they are still on the exact same frequency as the 8400's g signal once every 70 cycles or so.

    We are still working with engineering to devise a work around if any are possible.

    In my opinion, a N wireless card operating in the 5.4 ghz range would solve the problem but we will see what they come up with.

    Based on my conversations with support, I think they will be releasing a tech note on this study. The data that we were able to capture is quite telling.

    Good Luck,
    Chris
  • viningvining X Member Posts: 4,368
    Nice study and virtually the same conlcusion that was being bantered about on this thread.

    http://amxforums.com/showthread.php?t=3007&highlight=multi-cell

    Me thinks a tech note is definitely long over due especially when the above thread eludes to the fact that AMX Engineering has known about this for some time now. Just think of the collective thousands of hours of waisted time that may have been eliminated by a timely tech note. Even as there probalbly isn't an answer to the problem other than eliminating one or the other, the knowing of the problem would have been extremely beneficial to all that use these devices.
  • flcusatflcusat Junior Member Posts: 309
    DHawthorne wrote:
    set each panel to connect only to the strongest SSID. Even if it does pick up another, it will stay connected to the one defined, but if all your access points have the same SSID, it won't be able to do that.
    .

    Dave, how can you do this? I don't see any options for this in the wireless settings?
  • DHawthorneDHawthorne Junior Member Posts: 4,584
    flcusat wrote:
    Dave, how can you do this? I don't see any options for this in the wireless settings?

    Sorry for not being clear - it's strictly a manual setting. If your WAP's have unique SSID's, you can fire up something like Netstumbler on a laptop, and see which one has the strongest signal in your panel area. Then you explicitly tell the panel to connect to that WAP.

    Theoretically, connecting to the strongest WAP should be automatic if they have the same SSID. I may try going back to this at my problem job, in case firmware upgrades have fixed whatever was causing my strange experience. I only offered the above suggestion in case you had a similar situation.
Sign In or Register to comment.