Question About Insteon over RF

[shannong]

I have an all Insteon setup with the ISY994i Pro and a 2413S. I have about 130 devices spread throughout every room in my home (~approx 3300 sqft). Almost every device is dual band except for a handful of devices like the On/Off plugin modules. At least 110 of the devices are dual band if not more. There isn't a room in my house with fewer than two dual-band devices on two different walls. On the branch circuit where the ISY and PLM plugin, all devices are on filters except for two table side lamps with low wattage LED bulbs. I haven't noticed a difference in performance with those lamps On/Off. The PLM is not connected to a power strip, UPS, etc.

 

I'm about to undertake some efforts to identify and filter noise, signal suckers, etc as I have a several devices that don't always respond reliably. I've been researching this in preparation to develop a strategy and acquire any necessary tools.

 

After reading a lot about noise sources, noise types, filters, etc I kept running across the often recommendation for adding more APs. I mostly ignored that recommendation as most of my devices are dual-band and already repeat RF messages. It seems in the earlier days dual-band devices were not the norm.

 

Then I got to thinking... Hey! Wait a minute. Since most of my devices are RF-enabled, they all repeat messages, and RF is retransmitted before powerline then why would noise on the power line matter? My RF-network should be robust enough for full coverage as long as there is an RF-enabled powerline device reliably communicating to my PLM.

 

Sure, there can also be interference in the RF 915Mhz range affecting wireless but I don't have any known sources in my own house except for maybe my microwave which is rarely used. Within each room, I don't seem to have RF issues as I can use the RemoteLinc2 in every room without issue.

 

I'm not asking for assistance troubleshooting the wired issues. I'll have a different thread for that. What are your thoughts on that being a non-issue with a good wireless mesh coverage with the dual-band devices?

[Michel Kohanim]

Hi shannong,

 

First of all, I am by no means an expert so my suggestions are based mostly based on my experiences and subjective.

 

1. I do not think you need any Access Points simply because the RemoteLincs work in every location of the house. This is especially true IF you take the RL and go as close to the PLM as possible and see if it still covers all your devices spread out through your house (of course controlled by RL). The exception is #2

2. I have seen best performance when phases are bridged (either by Access Points or dual band devices) very close to the breaker panel and close to each other. This way, phases are bridged with on RF hop to one another

3. Unless you have only a handful of devices in a very large house, most communication issues are noise related and not signal strength related

4. In a lot of cases, the location of the PLM is instrumental. If you have it plugged into the same outlet as all other electronic equipment, rest assured that you are going to have some type of noise and communication problems

5. Transformers AND "electronic" low voltage lights (not the magnetic ones) are pretty much the biggest source of noise

6. Baby monitors, older INSTEON thermostats, SwitchLinc v35s can definitely cause problems

 

With kind regards,

Michel

[shannong]

Thanks for the tips.

[larryllix]

Then I got to thinking... Hey! Wait a minute. Since most of my devices are RF-enabled, they all repeat messages, and RF is retransmitted before powerline then why would noise on the power line matter? My RF-network should be robust enough for full coverage as long as there is an RF-enabled powerline device reliably communicating to my PLM.

I was just reading (and posted in another thread) Insteon's repeating algorithm where they state the RF data received is repeated on the powerline and vice versa, but I haven't seen anything stating how the arbitration between bands is done. IOW who gets first crack if they differ? Who gets repeated?It also states that RF is synchronised with the Powerline data so is only as fast as the powerline frequency data speed allows. (yay! for N.America 60Hz. )

 

I make these assumptions:

-If a garbage packet is received on either band it will not be repeated on either band.

 

-Simultaneous repeated transmissions, from different modules, are stated to reinforce signal strength. Anybody dealing with audio or any signal equipment probably knows this is not always true. Even if all modules are exactly in phase the phase lag between location distance, wire lengths and impedances (line proximities or concrete walls) can cause complete blind spots at a receiver's location where signals from multiple transmitters, out of phase, cancel each other out completely or weaken the signal strength to the point of uselessness. Simply moving a receiver over a few inches or injecting a loop in one wire may make all the difference. In the powerline frequency multiple unit synchronisation may be possible with the low frequency carrier used, but in the high frequency RF (915MHz) even the CPU being out of sync by one CPU clock cycle, at these lower clock rates, may result in the bits being out of sync by more than multiple RF wavelengths. I haven't done the actual signal wavelength delay math, yet, so don't get too excited if I'm too far off.

 

IOW: Mess network units that don't use collision detection avoidance at the carrier level of the protocol may not play nicely with each other and the more you have may make the occasional failure MORE common. Designing units to talk simultaneously may not always work well. Six or eight, well placed, repeating units and the rest receive only may just function more reliably.

 

In short: you may have too many dual band units at play and I would be willing to do my part to help you out in that endeavour. I can PM you my shipping address for some possible relief from your situation.