MVP's Wireless internet issues
ericmedley
Posts: 4,177
I have been experiencing some problems that have began sprouting up on all our systems that use MVP-8400s. Size of system, number of panels, busy communications between panel and master or not, different programmers, etc... doesn't seem to be a factor.
The one factor that does seem to be present is the presence of 2.4GHZ phone systems with multiple cells.
The most prominet has been with the Panasonic TDA-50 system. They state in their documentation that they used 2.4Ghz channels 1-8. also, they recommend that you only use channels 9,10,11 for 'other' things. Further discussions with Panasonic's engineering department have yielded little more than a ,'too bad, that's how we do things. sucks to be you' attitude. There is no way to make any changes in the way the phones operate. We cannot change the channels it uses or whatnot.
the same is also the case with several other manufacturers of this technology. We've pretty much seen the same thing in Samsung, Bang & Olufson, and a couple others that I can't remember at the moment.
For those unfamiliar with this freq band, it's the same one used for 802.11b/g wireless networks, cell phones, microwave ovens, in other words all kinds of stuff. In the US we have 11 channels to choose from. In addition, it is not a good idea to use multiple channels that are adjacent do to the spurious frequenies that are transmitted by these devices that can cause interference. In addition we've found that it is very bad to try to put WAPs on the same channels even if they are resonably far apart. We've found that WAPs on the same channel that are even over 1000 feet apart seem to still interfere with each other.
Most of our resi systems are homes in the over 9,000 sq ft range. We find that, in order to get strong enough coverage for the MVP-8400s we need to blanket the home with wireless internet. As many of you know, the MVPs are not as forgiving as an average laptop wireless card when it comes to low signal from/to the WAP. In most cases, due to the size of the property, we need to use more than 3 WAPs to get coverage. We've experimented with using lots of WAPs on very low power settings, fewer WAPs but much more powerful settings. We've used a whole variety of WAP manufacturers. We are currently switching all the systems over to the high-end Ciscos. They seem to offer the best reliability and logging. but even with these, the problem still persists at some level.
We've also tried letting all the WAPs choose the channel themselves. This doesn't seem to work at all since the WAPs appear to only concern themselves with IP WAPs. They will regularly set themselves up on channels that are already screaming hot with phone cell waps or whatnot. We've also seen a Linksys WAP set itself up on a channel occupied by a Cisco WAP that was within 30 feet. (this during a test at our shop)
So, to my points...
1) I know this frequency band is getting very crowded. We are seeing a trend in our clients to put more and more wireless phones in homes. our average 2 years ago was something like 1-3 hard wired stations and 1-2 wireless. Now, we are putting in many more wired and upwards of 6-8 wireless. If this trend is happening everywhere (and world-wide sales trends would seem to support this) where does that leave us poor little integrators who have to now squeeze our wireless internet transponders into the three remaining channels? (9,10 and 11) Also, note that these channels are not the best of the bunch. 9 is notoriously bad since that's where microwave ovens sit. I can specifically pin the phones to the problem in some cases as we've found that when we turn off the phones, the problem goes away. But not always...
2) Is the problem more an AMX issue? We've already seen that the MVPs are not as robust as an average laptop wirless. Sitting side by side, the laptop has a way better connection. Has anyone seen occurances of the MVPs going offline for no apparent reason? I've noticed that some of the reports I get back show that they fall off line and reappear in exact 3 minute intervals. Others seem to be that the panels fall away for seemingly random amounts of time. (several 10s of minutes.)
Our clients report that the panels are locked up. They are not locked up per se. It's just that they're not communicating with the Netlinx master, so they see no response. At times they say the signal strength bar (i put that on their interface) is very low even if the WAP is within feet of the panel. Other times, it is perfectly normal. The fix seems to be to reboot the panel. Once they do that the devices comes back online.
The manufacturers are not very helpful in that they ultimately must toe the company line. It is not in their interest to determin why their box is causing problems with another manufacturers box. (see part 15 of the FCC regs)
For us, it's becoming a problem in that I cannot offer a reliable AMX solution that uses wireless touch panels. This may or may not be AMX's fault. However, it is also not reasonable to ask the client to forego using wireless phones. And since no one seems interested in making any moves we're just stuck.
I would appreciate any feedback. We have an entire team of people who are skilled in all these areas working on the problem. We're all reaching the sad conclusion that there is not much that can be done at the moment. It's the same problem as the freeways. The problem's not the road, it's the number of cars wanting to drive on the road.
I remember all the discussion in engineering/networking circles several years ago saying that is was foolish to put 802.11 on the public bands. However, manufactureres were screaming about the cost of fcc licensings and whatnot. In the next few years we may find that the roosters have come home to roost...
The one factor that does seem to be present is the presence of 2.4GHZ phone systems with multiple cells.
The most prominet has been with the Panasonic TDA-50 system. They state in their documentation that they used 2.4Ghz channels 1-8. also, they recommend that you only use channels 9,10,11 for 'other' things. Further discussions with Panasonic's engineering department have yielded little more than a ,'too bad, that's how we do things. sucks to be you' attitude. There is no way to make any changes in the way the phones operate. We cannot change the channels it uses or whatnot.
the same is also the case with several other manufacturers of this technology. We've pretty much seen the same thing in Samsung, Bang & Olufson, and a couple others that I can't remember at the moment.
For those unfamiliar with this freq band, it's the same one used for 802.11b/g wireless networks, cell phones, microwave ovens, in other words all kinds of stuff. In the US we have 11 channels to choose from. In addition, it is not a good idea to use multiple channels that are adjacent do to the spurious frequenies that are transmitted by these devices that can cause interference. In addition we've found that it is very bad to try to put WAPs on the same channels even if they are resonably far apart. We've found that WAPs on the same channel that are even over 1000 feet apart seem to still interfere with each other.
Most of our resi systems are homes in the over 9,000 sq ft range. We find that, in order to get strong enough coverage for the MVP-8400s we need to blanket the home with wireless internet. As many of you know, the MVPs are not as forgiving as an average laptop wireless card when it comes to low signal from/to the WAP. In most cases, due to the size of the property, we need to use more than 3 WAPs to get coverage. We've experimented with using lots of WAPs on very low power settings, fewer WAPs but much more powerful settings. We've used a whole variety of WAP manufacturers. We are currently switching all the systems over to the high-end Ciscos. They seem to offer the best reliability and logging. but even with these, the problem still persists at some level.
We've also tried letting all the WAPs choose the channel themselves. This doesn't seem to work at all since the WAPs appear to only concern themselves with IP WAPs. They will regularly set themselves up on channels that are already screaming hot with phone cell waps or whatnot. We've also seen a Linksys WAP set itself up on a channel occupied by a Cisco WAP that was within 30 feet. (this during a test at our shop)
So, to my points...
1) I know this frequency band is getting very crowded. We are seeing a trend in our clients to put more and more wireless phones in homes. our average 2 years ago was something like 1-3 hard wired stations and 1-2 wireless. Now, we are putting in many more wired and upwards of 6-8 wireless. If this trend is happening everywhere (and world-wide sales trends would seem to support this) where does that leave us poor little integrators who have to now squeeze our wireless internet transponders into the three remaining channels? (9,10 and 11) Also, note that these channels are not the best of the bunch. 9 is notoriously bad since that's where microwave ovens sit. I can specifically pin the phones to the problem in some cases as we've found that when we turn off the phones, the problem goes away. But not always...
2) Is the problem more an AMX issue? We've already seen that the MVPs are not as robust as an average laptop wirless. Sitting side by side, the laptop has a way better connection. Has anyone seen occurances of the MVPs going offline for no apparent reason? I've noticed that some of the reports I get back show that they fall off line and reappear in exact 3 minute intervals. Others seem to be that the panels fall away for seemingly random amounts of time. (several 10s of minutes.)
Our clients report that the panels are locked up. They are not locked up per se. It's just that they're not communicating with the Netlinx master, so they see no response. At times they say the signal strength bar (i put that on their interface) is very low even if the WAP is within feet of the panel. Other times, it is perfectly normal. The fix seems to be to reboot the panel. Once they do that the devices comes back online.
The manufacturers are not very helpful in that they ultimately must toe the company line. It is not in their interest to determin why their box is causing problems with another manufacturers box. (see part 15 of the FCC regs)
For us, it's becoming a problem in that I cannot offer a reliable AMX solution that uses wireless touch panels. This may or may not be AMX's fault. However, it is also not reasonable to ask the client to forego using wireless phones. And since no one seems interested in making any moves we're just stuck.
I would appreciate any feedback. We have an entire team of people who are skilled in all these areas working on the problem. We're all reaching the sad conclusion that there is not much that can be done at the moment. It's the same problem as the freeways. The problem's not the road, it's the number of cars wanting to drive on the road.
I remember all the discussion in engineering/networking circles several years ago saying that is was foolish to put 802.11 on the public bands. However, manufactureres were screaming about the cost of fcc licensings and whatnot. In the next few years we may find that the roosters have come home to roost...
0
Comments
Now throw in ZigBee, which so far I'm liking, but when you start trying to use multiple gateways and there fore need use channels other than 25, 26 and delve into the realm of wi-fi and multi cell phones, that won't be any fun what so ever.
The TPs now support an active roaming feature that can be enable but only works if WAPs are on channels 1,6,11. I use a variety of channels, usually 4. 1, 5, 9, 11. Where near channels are located as far apart as possible and then if I need to repeat a channel for say a 5th or 6th WAP I repeat but usaully off 1 channel.
If I need to tweak WAP levels there's a Must Die frimware availble for Linksys WAPs than will allow me to change output power levels. I don't recall if I can tweak recevier sensitivity.
The new wire antennae included on the new G nics are a big improvement over the older B nics but having a tuned antennae that runs acroos the top the length of the panel would likely significantly improve range.
I've also noticed that when transferring files to the TPs that it often fails when strictly on battery power. I know it's recommended to transfer when on AC power but some times thats just a pain.
http://amxforums.com/showthread.php?t=2151
Actually, that thread had more to do with Encryption/security. However, thanks anyway for the help.
I have had a chance to talk to AMX at length today (as well as other network engineers) All signs seem to point to known problems with FFHS (Frequency-hopping Spread Spectrum) devices. The 3 second brown outs I was experiencing with the touch panels match up with the hop time of many devices out there. All of the phone systems I mentioned use this method.
The overall conversation has not gone well in general. The consensus seems to be that there's not much we can do about it really. the only control we have is the placement of all of these devices within the space. All of us in Residential know that even there, we're at the mercy of the interior desecrators (I mean decorators)
to add to the fray, we have Zigby and Bluetooth all occupying the same band.
We can get in and tweak the WAPs to some degree and also use the MAC address authentication and minimum bandwidth settings to keep out the riff-raff. However, this still doesn't address the interference issue.
I'll keep posting any news I hear on this issue.
That doesn't sound encouraging! I believe the Pana cordless phone are basically passive when not in use or the transmitter side of the transceiver just goes to sleep. When the handset is nested in its cradle or on the coffee table the TPs appear to function properly but as soon as some one makes a call on one is like a crap shoot. The TP may or may drop communications and the person making the call usually has the TP on his/her lap for maximum interferance.
I asked an AMX trainer about this about 4-5 years ago and then asked why they didn't consider using 802.11A with all the 2.4ghz congestion and I got an evil look and then he said why doesn't Panasonic just make their multi-cell systems at 5.8ghz. It appears no one what's to take a strategically defensive posture and try for the cleaner air space.
I know I've research the freuquency for the multi cells before and if not mistaken the values are in the medical reference section but I can't seem to find that right now. Perhaps the upper freq limit is below ZigBee channel 26 freq which would make them a prefferable choice in certain installations. For some reason though I'm thinking the phones go up to 2485mhz which will cover ZigBee channel 26.
Typically, at least for me, when were on site testing the owners aren't around and if they are they are else where in the house. Now, when they are there and nearby it's usually cuz we're instructing them or showing them something and typically don't make calls at this time. Every one else on site typically uses their cell phones so there aren't that many times when the cordless phones are in use in close proximity to the MVPs while we're there to observe these subtle communication drops.
I did however just do a job with an open kitchen family room layout and often the owner or their assistants would be there and I would be sitting at a table with the MVP less than ten feet from where they would periodically make calls on their wireless handsets. Often during this time I would notice button pushes wouldn't execute and a general sluggish behavior. I have panel online feedback next to my battery indicator on the main page that always show me connected. I also had a situation where I told a customer to call me back in 5 minutes to give me time to connect to their system (VNC through VPN). I would connect w/o any problems and have complete control ot the TP and when they called back my control was gone. There's lots of reason this could happens but I'm thinking multi cell!
I had every intention of taking a handset and spending some time testing this but of course I spaced out, fell behind what I was during and left late w/o running any tests.
If any one has a situation in the near future where they can test a MVP w/ a multi cell phone in their hand to see how responsive button pushes are during a live call I be interested in knowing your results.
I'd like to believe they can co-exist in one big happy RF family but from an EMI, RFI and IMI perspective they shouldn't.
Jeff
Typically not more than one or two. However, I have a job that has 6 WAP's and a dozen or so cell stations, and there is no interference. I have had lots of RF issues on that particular job - like a WAP that won't work in a room only 10' from the WAP, but I had that problem before putting in the Panasonic. His old phone system wouldn't work cordless at all ... one room away from the base, and the phone died. We changed it out to the TDA just to solve that problem. I am convinced it's either the extensive Lutron control wiring or an environmental thing (some huge power feeds run past his south 40). But the Panasonic has not been a problem with the wifi.
Hmmm...
This is yet another case I can make for getting an RFSpec analyzer (they cost well over $5000 for a good one)
In our situation, if we unplug the Panasonic phone system and power down all the cell phones, the problem goes away with the exception of when the client puts his blackberry next to his master bedroom touch panel.
There is a lot of Lutron in this house. It's a full blown Homeworks system with a ton of dimmers and whatnot.
We've run into mysterious wireless issues at almost every home in the past two years. I do think it's time to geta RF Spec Analyzer and get an idea of what's going on.
Thanks everyone for the input.
ejm