RF Comms Issues

[Scyto]

I only just learnt of the 4-button test. So I tried it.

 

Most of my dual band dimmers and keypad dimmers around the house blink green or red.

 

However there are 5 that don't (out of about 30) and I am not sure what to do next.

There is one downstairs in the center of the house it is on a pillar and on the other side of the pillar are 3 dimmers that do blink. The one that doesn't blink is actually closer and has better line of sight to the ISY on the floor below.

 

Then I have several upstairs towards the back of the house. What is odd is that in one switch box I have 2 dimmers and one blinks and the other doesn't.

 

Generally everything is reliable over power line.

 

What should I do to troubleshoot. How can I prove it is a general RF issue rather than the devices being defective?

[stusviews]

Once you have at least one pair of devices communicating on opposite legs of the split, single-phase electric supply, then you have adequate coupling. But, not every device can communicate with every other device by RF alone. That's OK as long as the devices can communicate with any other device.

 

Start the 4-button test from a device that didn't respond. Does it communicate with at least one other device that did respond?

[Scyto]

Oh, how do I start the test from any device?

[Brian H]

Try tapping the set button four times. Most of the modules start the test.

Some modules like the latest 2457D2 LampLincs. Now use a local set button sequence flow chart to do it. In that case it is in the full users manuals you can download.

[stusviews]

The 4 taps must be really, really fast.

[Scyto]

Ok I did this on one of the KPDs that is in the bedroom that previously didn't flash.  Now it and the neighbors around all flash.  So that tells me that, yes, the back of the house has some dead spots.

 

Should I juts get one of the insteon range extenders, would that create a complete RF mesh and plug the dead zone?

[LeeG]

Any Dual Band device if positioned between the PLM and that KPD will improve RF coverage.  I think it better to resolve the powerline issues so that Insteon mesh network functions as designed.   

[Teken]

 

Should I juts get one of the insteon range extenders, would that create a complete RF mesh and plug the dead zone?

 

I like to consider dual use when installing any plug in wall wart. The Lamp Linc has a specified 150 RF range where as the Range Extender has a 200 RF range.

 

That extra 50 feet might make a difference in your environment but like LeeG. I would suggest you hunt down what are the noise makers / signal suckers in the home.

 

If the extra 50 feet isn't important than going with a Lamp Linc will offer you light control for dimming. A Range Extender only does one thing and that's to couple and re-transmit RF to power line. 

[oberkc]

I would also add that I am not convinced that it is even a problem, nor should it be an expectation, that all dual-band insteon devices be within direct RF range of all others. The point of insteon (and zwave, I expect) is that the network offers multiple communication paths, over powerline and radio waves, for one device to talk to another. The more paths, the better, but it is normal that there exists some devices that cannot directly communicate with others.

 

The exception to this is when devices are on opposite legs of the electrical system. As I understand it, the beacon test is based on direct communication between the initiator (four taps in most cases) and receiver. Its purpose is, solely, to determine if one has communication across the legs of your electrical system....not to verify that every device can hear a radio signal from any given transmitter. Unlike a normal insteon command, I am not sure that insteon devices repeat the beacon test signal, so the addition of range extenders may not even guarantee different test results.

 

Fortunately, you passed your test with flying colors. Whatever problems you have are, I expect, unrelated to RF coverage. To me, the only value of new range extenders in your case is if you happen to have rf-only devices that cannot communicate with the nearest dual-band device. You would be, in my estimation, wasting your money otherwise.

[Teken]

I would also add that I am not convinced that it is even a problem, nor should it be an expectation, that all dual-band insteon devices be within direct RF range of all others. The point of insteon (and zwave, I expect) is that the network offers multiple communication paths, over powerline and radio waves, for one device to talk to another. The more paths, the better, but it is normal that there exists some devices that cannot directly communicate with others.

 

The exception to this is when devices are on opposite legs of the electrical system. As I understand it, the beacon test is based on direct communication between the initiator (four taps in most cases) and receiver. Its purpose is, solely, to determine if one has communication across the legs of your electrical system....not to verify that every device can hear a radio signal from any given transmitter. Unlike a normal insteon command, I am not sure that insteon devices repeat the beacon test signal, so the addition of range extenders may not even guarantee different test results.

 

Fortunately, you passed your test with flying colors. Whatever problems you have are, I expect, unrelated to RF coverage. To me, the only value of new range extenders in your case is if you happen to have rf-only devices that cannot communicate with the nearest dual-band device. You would be, in my estimation, wasting your money otherwise.

 

 

I believe the problem for some folks is trying to understand how can two DB units in the same wall. One will Ack the RF beacon test while another beside, not.

 

The other issue is where an initiator makes a receiver respond and confirms bridging / coupling. But, when you do the reverse the (new) receiver which was the initiator does not Ack coupling / bridging.

 

These two examples are the pain points which I am hard pressed to explain to people. 

[Xathros]

I have to agree with oberkc on this.  Your testing seems to indicate reasonably decent RF comms and successful phase bridging.  If there are any devices in common responding to the PLM 4 tap and the far end 4 tap tests then you are only 2 hops from the PLM to the far end by RF only and this is assuming you are not already making it to the far end on 1 hop via the power line.  With the event viewer running at level 3, control one of the far end devices via the admin console and inspect the event log results for hops remaining count.  report your results.

 

-Xathros

[Scyto]

LeeG, on 09 Feb 2015 - 10:36 AM, said:

Any Dual Band device if positioned between the PLM and that KPD will improve RF coverage.  I think it better to resolve the powerline issues so that Insteon mesh network functions as designed.   

 

I have lots of dual band devices that meet this description both horizontally and vertically.  For example in the master bathroom there are two dual band dimmers in the same plastic wall box.  One sees the RF signals from the ISY the other does not.  It does not seem that the RF devices form as repeaters to other RF devices.

[Scyto]

Xathros, on 09 Feb 2015 - 12:08 PM, said:Xathros, on 09 Feb 2015 - 12:08 PM, said:

I have to agree with oberkc on this.  Your testing seems to indicate reasonably decent RF comms and successful phase bridging.  If there are any devices in common responding to the PLM 4 tap and the far end 4 tap tests then you are only 2 hops from the PLM to the far end by RF only and this is assuming you are not already making it to the far end on 1 hop via the power line.  With the event viewer running at level 3, control one of the far end devices via the admin console and inspect the event log results for hops remaining count.  report your results.

 

-Xathros

 

This was my conclusion - but given the issues I have had programming the KPDs (the ones that didn't blink when I did the ISY test) I am unconvinced that the power line is reliable OR these devices are just plain crappy...

 

It seems I have some sort of intermittent issue (see last one)

 

Mon 02/09/2015 12:52:33 : [iNST-TX-I1  ] 02 62 29 4A C3 0F 11 FF

Mon 02/09/2015 12:52:33 : [iNST-ACK    ] 02 62 29.4A.C3 0F 11 FF 06          LTONRR (FF)

Mon 02/09/2015 12:52:33 : [iNST-SRX    ] 02 50 29.4A.C3 2A.1C.FF 2B 11 FF    LTONRR (FF)

Mon 02/09/2015 12:52:34 : [std-Direct Ack] 29.4A.C3-->ISY/PLM Group=0, Max Hops=3, Hops Left=2

Mon 02/09/2015 12:52:34 : [D2D EVENT   ] Event [29 4A C3 1] [sT] [255] uom=0 prec=-1

Mon 02/09/2015 12:52:34 : [  29 4A C3 1]       ST 255

Mon 02/09/2015 12:52:36 : [iNST-TX-I1  ] 02 62 29 4A C3 0F 13 00

Mon 02/09/2015 12:52:36 : [iNST-ACK    ] 02 62 29.4A.C3 0F 13 00 06          LTOFFRR(00)

Mon 02/09/2015 12:52:37 : [iNST-SRX    ] 02 50 29.4A.C3 2A.1C.FF 2B 13 00    LTOFFRR(00)

Mon 02/09/2015 12:52:37 : [std-Direct Ack] 29.4A.C3-->ISY/PLM Group=0, Max Hops=3, Hops Left=2

Mon 02/09/2015 12:52:37 : [D2D EVENT   ] Event [29 4A C3 1] [sT] [0] uom=0 prec=-1

Mon 02/09/2015 12:52:37 : [  29 4A C3 1]       ST   0

Mon 02/09/2015 12:52:39 : [iNST-TX-I1  ] 02 62 29 4A C3 0F 11 FF

Mon 02/09/2015 12:52:39 : [iNST-ACK    ] 02 62 29.4A.C3 0F 11 FF 06          LTONRR (FF)

Mon 02/09/2015 12:52:39 : [iNST-SRX    ] 02 50 29.4A.C3 2A.1C.FF 2B 11 FF    LTONRR (FF)

Mon 02/09/2015 12:52:39 : [std-Direct Ack] 29.4A.C3-->ISY/PLM Group=0, Max Hops=3, Hops Left=2

Mon 02/09/2015 12:52:39 : [D2D EVENT   ] Event [29 4A C3 1] [sT] [255] uom=0 prec=-1

Mon 02/09/2015 12:52:39 : [  29 4A C3 1]       ST 255

Mon 02/09/2015 12:52:39 : [iNST-SRX    ] 02 50 29.4A.C3 2A.1C.FF 23 11 FF    LTONRR (FF)

Mon 02/09/2015 12:52:39 : [std-Direct Ack] 29.4A.C3-->ISY/PLM Group=0, Max Hops=3, Hops Left=0

Mon 02/09/2015 12:52:39 : [iNST-DUP    ] Previous message ignored.

 

[Xathros]

This was my conclusion - but given the issues I have had programming the KPDs (the ones that didn't blink when I did the ISY test) I am unconvinced that the power line is reliable OR these devices are just plain crappy...

 

for example I just turned a scene on and off several times - it worked perfectly, however if I do a scene test intermittently get 'scene failed' - even though my eyes show it worked just fine.

Do the test I indicated above and see how the hops remaining count looks.  If it shows consistent remaining hops, then communications may not be your issue.  I have seen a scene test fail with excellent comms.

 

-Xathros

[Teken]

Based on your trace you have noise issues that need to be addressed.

[Xathros]

It does look like something is interfering with comms to that device.  What else shares that circuit?

 

-Xathros

[Scyto]

My breaker map says just lights. I will double check when I get home Friday.

 

 

Sent from my iPhone using Tapatalk

[Scyto]

It does look like something is interfering with comms to that device. What else shares that circuit?

 

-Xathros

The only thing on that circuit are lights and two velux windows motors (controlled by IR) the motors only run when the windows open and close. I see no obvious way to disconnect them (I could go buy a ladder and get into the roof space if really needed..)

 

What do you think next steps should be?

[stusviews]

Anything controlled by IR has power on 24/7 so that it can respond to the IR signal.

[Scyto]

Anything controlled by IR has power on 24/7 so that it can respond to the IR signal.

Sure, I understand that, my point is that motor noise would only be generated at time of use. As for wether it has some sort of transformer that could be causing an issue I don't know. I am hoping someone has some experience with velux and can weigh in.

 

I am looking to understand if I need to go out, buy a ladder and start crawling around the attic

[stusviews]

IMO, if you're a homeowner, then you should own a ladder

[Xathros]

It's not necessarily the motor. Often power supplies for electronic device can either be noisy or have line filters that absorb signals.

 

I would try plugging in a dual band lamplinc or range extender into that circuit and see if things improve.

 

 

-Xathros

 

Sent from my iPhone using Tapatalk

[Scyto]

stusviews, on 16 Feb 2015 - 1:42 PM, said:stusviews, on 16 Feb 2015 - 1:42 PM, said:

IMO, if you're a homeowner, then you should own a ladder

yeah, I have only been a homeowner for a year in the US had plenty in my house when I was in the UK over a decade ago, I keep looking at the Costco megamax / little giant ladder and trying to decide on biting the bullet.

[Scyto]

Xathros, on 16 Feb 2015 - 1:58 PM, said:

It's not necessarily the motor. Often power supplies for electronic device can either be noisy or have line filters that absorb signals.

 

I would try plugging in a dual band lamplinc or range extender into that circuit and see if things improve.

 

 

-Xathros

 

Sent from my iPhone using Tapatalk

 

The dimmers in question are dual band, can you help me understand how plugging more dual band devices in will help? 

 

Also given this is a lighting circuit there are no receptacles to plug anything in to... I guess I could get an insteon light bulb but I don't want that permanently on the circuit...

 

--edit---

 

Also remember that when I initiate the 4-tap test from these dimmers, all the other insteon devices upstairs flash fine, they are distributed across about 4 or 5 other circuits on both sides of my panel....

[Xathros]

Ok. Did these dimmers respond to the beacon test started at the PLM?

 

 

-Xathros

 

Sent from my iPhone using Tapatalk