Intermittent Comm Problem (also how do I read level 3 log)

[ELA]

Thanks IndyMike,

I see what you are saying now on item #2. That makes sense assuming the intended receiver was dual band and was within RF range of the sender. Otherwise I see it as you are eliminating that hop #0 completely and thus the device must rely on subsequent simulcasted hops. In my experience hop#1 being the preferred.

 

Yes it does appear that there is some bit shifting going on. I do wish We knew the checksum calculation used cause I have always wondered why the PLM passes these up as valid messages?

[Vyrolan]

Do you guys think too many dual-band devices (or too many devices in general) is (or can be) a problem? I was wondering about that sometimes... With everything as a repeater, it seems like with lots of devices (particularly on the same circuit) you run the risk of just flooding the powerline (or the RF). As an example, I have one circuit with 10 devices (and 3 other RF-only that communicate with those 10). Of the 10, 9 are dual-band.

[IndyMike]

Do you guys think too many dual-band devices (or too many devices in general) is (or can be) a problem? I was wondering about that sometimes... With everything as a repeater, it seems like with lots of devices (particularly on the same circuit) you run the risk of just flooding the powerline (or the RF). As an example, I have one circuit with 10 devices (and 3 other RF-only that communicate with those 10). Of the 10, 9 are dual-band.

 

Vryolan,

 

I'm sorry, but this is an area where I don't have much experience. For the past 7 years I've been running my installation on a hardwired coupler and 1 Accesspoint (for motion sensors). I have recently installed 2 dualband LL's in my family room (same phase as the accesspoint) with no ill effects.

 

The best I can say is the SmartHome doesn't appear to believe there will be problems - they're pretty much forcing us to move toward dualband modules.

[ELA]

Vyrolan,

I do firmly believe you can have too many devices in close proximity to each other. I do not know about what a good mix of SB and DB is however.

 

The reason I believe you can have too many is from my experience in my living room with 11 devices.

Only one of them is DB.

I posted a lot of testing information on that room here:

 

http://forum.universal-devices.com/viewtopic.php?p=54325#p54325

 

Signal strength is decreased with each additional Insteon load added. Depending upon where the signal originated from ( its initial strength entering a room), the initial hop signal strength can become very low with multiple devices loading it. One would expect that would not matter since subsequent simulcasted hops signal strength would be very strong. In my experience in this room the hops #1-3 signal strength is very strong. Yet communications would occasionally still fail.

 

This is what we have been discussing. Why is this? I have my theories and can speculate but no hard proof.

 

On this problem room I worked hard to get a stronger hop #0 ( initial send) signal to all devices. This worked and that room is very reliable now.

As has been discussed another possible option is to force the PLM hop #0 to be blocked completely. Rely on RF as the only initial send and subsequent hops 1-3 as the main source to remote devices.

 

That is why I say that I believe that devices can sometimes try too hard and reach into the noise floor for that initial send (hop #0). Then either they get confused and ignore subsequent hops, or simulcasting just does not work very well. I am unsure as to which is occurring. And also a little burnt out on any extensive testing efforts

 

If a person had such a marginal situation and cared to spend hours testing:

blocking the initial hop #0 (PL filtering the PLM), to force RF as the only hop #0, could be used to help prove which is happening. Receiver needs to be SB or out of range of the PLMS RF.

 

In my opinion a person has to do extensive reliability testing to make the case though. It is not enough to just say things appear to have gotten better. Often times they appear to get better only to get worse again later.

Having reliability statistics is much more convincing.

 

I would love to hear any results you might gather from your testing.

[Vyrolan]

The reason I believe you can have too many is from my experience in my living room with 11 devices.

Only one of them is BD.

Yea that's how this one is...the whole master suite (including the ceiling lights in the two closets and the master bath) are all on one circuit...that is the one with 10 insteon devices. Since I've only recently gotten into Insteon, basically all of my stuff is dual-band (except a few old switchlincs I recently got from Best Buy on the cheap and the KPL Dimmers).

 

 

Signal strength is decreased with each additional Insteon load added. Depending upon where the signal originated from ( its initial strength entering a room), the initial hop signal strength can become very low with multiple devices loading it. One would expect that would not matter since subsequent simulcasted hops signal strength would be very strong. In my experience in this room the hops #1-3 signal strength is very strong. Yet communications would occasionally still fail.

I kind of was wondering if that was my problem when I moved my PLM. Before, it was on an outlet in the middle of that 10-device circuit...so it was short trips to all the relevant devices. In the new location, it was on a different circuit on the same phase so it had to make the trip through the breaker box and then out that 10-device circuit. It seemed like all the devices and all the repeating was maybe gumming up the works, but it seemed more likely that noise/attenuation at the PLM's new location was the problem (since there are the power strips there). We will find out soon enough as my Filterlincs should be here on Thursday. If I filter the two power strips and the subwoofer, I expect the PLM performance to be back to normal. If not, it will be interesting to see what the messaging looks like and I will be open for more testing/ideas.

 

This is what we have been discussing. Why is this? I have my theories and can speculate but no hard proof.

Yes I have been following along with much curiosity. =) Sadly, we can probably never really know for sure without additional information from SmartLabs that they are not likely to provide. To reverse engineer or determine by observation everything we want to know would require a pretty extensive testing setup. =p

 

In my opinion a person has to do extensive reliability testing to make the case though. It is not enough to just say things appear to have gotten better. Often times they appear to get better only to get worse again later.

Yea our ability to measure these things isn't very scientific. =p

[ELA]

Vyrolan,

From your description it does sound like moving your PLM made the for the decrease in reliability. It did also sound like you had some signal suckers to filter.

Hopefully filtering signal suckers will revive your reliability. Best of luck to you.