Can you have too much signal strength with Insteon?

[ELA]

While doing some testing on a small "isolated Insteon network" I have seen some strange behavior with regard to the Insteon signals.

 

Here is a post regarding this:

http://forum.universal-devices.com/viewtopic.php?t=5923&start=30 (near the bottom)

 

I was wondering if Insteon can explain whether there is such a thing as having too many devices located close together?

[ELA]

Does anybody really care?

[apostolakisl]

I care, I swear, I really do!!! I just don't have a clue.

 

But I look forward to you finishing your Insteon signal diagnostic device.

[Brian H]

When you get more data. I do care also.

 

I joined the Developers Group in the Insteon infancy. It was being pushed for both software developers and those that are interested in how Insteon Hardware works. I am the second choice. Hardware interested.

 

We even got to talk to Smartlabs Insteon Engineers. That is all changed now.

 

I do find your data very interesting. Just usually don't have much to add.

[TJF1960]

While doing some testing on a small "isolated Insteon network" I have seen some strange behavior with regard to the Insteon signals.

 

Here is a post regarding this:

http://forum.universal-devices.com/viewtopic.php?t=5923&start=30 (near the bottom)

 

I was wondering if Insteon can explain whether there is such a thing as having too many devices located close together?

 

Smarthome may not monitor this forum very much. Perhaps an email or phone call to SH would get better results. It would be nice to know who you could contact though!

 

I am, by the way, also very interested in your tests and findings as well. Please, keep up the good work!

 

Thanks,

Tim

[LeeG]

If you are expecting Smarthome/Smartlabs to jump in and provide an answer or technical data on the subject I’m afraid you will be disappointed. The contribution and analysis you and others have made on this subject are the most information I have seen published on the subject.

 

Anecdotally I have 18 devices plugged into metal outlet strips which serve as part of a test bed. This is separate from the 50+ devices installed as operating units in the house and garage. No surge suppression, the outlets are simply hard wired together. The devices are anywhere from 6 inches to a few feet apart for the wire length to connect the various outlet strips. The mix includes various PLMs (standard, HL enabled, and Dual Band), 2 EZSrves, a half dozen other SHN device types, KeypadLincs (dimmers and relays), ICON switches (dimmers and relays), LampLincs, ApplianceLincs, a ToggleLinc, a FilterLinc (functional to isolated PC equipment and UPS) and an Access Point. All devices that are driving loads use incandescent bulbs.

 

The combination varies depending on what is being tested, exercised or analyzed at the time. The basic setup has been used for years, increasing the number of outlet strips as the number of test bed devices has increased. It has the capacity to hold 21 devices with another outlet strip to be added in a few weeks. I have never experienced a powerline quality problem resulting from that many devices essentially plugged next to each other. Of course I have not examined the actual powerline signals as you and others have done. My dual trace is 25 years old with no memory.

 

The lack of response is more the lack of useful contribution than lack of interest. I hate it when folks jump in with opinion and speculation not supported by objective data. Keep up the great work!

[ELA]

If you are expecting Smarthome/Smartlabs to jump in and provide an answer or technical data on the subject I’m afraid you will be disappointed.

 

Thanks to you all for your input.

In my testing I have not confirmed whether or not the actual communications is getting errors. I was only looking at signals on the scope. For all I know I may be having an "instrumentation error". In addition I am working on a "scabbed prototype" and this could be at fault.

 

I was just looking for more information from anyone who could offer some input.

I put the request in this thread since it is Smarthome related and I saw that Steve Lee contributed here.

 

Anecdotally I have 18 devices plugged into metal outlet strips which serve as part of a test bed. This is separate from the 50+ devices installed as operating units in the house and garage. No surge suppression, the outlets are simply hard wired together. The devices are anywhere from 6 inches to a few feet apart for the wire length to connect the various outlet strips.

 

 

Thank you LeeG for your description of your test network.

if I understood you the test network is not isolated from the home network?

If this is the case, and you are willing, it might be interesting to attempt to isolate the small test network. I used a Filterlinc followed by a 4mH common mode choke. The filterlinc alone may be enough.

 

I was trying not to get too distracted from my goal but this symptom did distract me. I am not intending to pursue it any further for now. I just was hoping that someone might have read something in the past on the subject.

I was also curious if anyone was capable of looking at the schematic and interpreting whether or not it looked like a large 131Khz signal could adversely affect zero cross detection.

When I build my first integrated prototype I will revisit this.

[LeeG]

You are correct. As it is today the test bed is not isolated from the rest of the house. Just lots of devices very close together powerline wise. I think I do have spare FilterLincs. I will isolate the test bed with a FilterLinc (maybe two) and see if I get errors device to device or with automation through the PLMs. Will post back the results.

[IndyMike]

Hello ELA,

 

Sorry for the long hiatus - work is becoming complicated.

 

Let me start by saying that I have performed "isolated network' testing and have not observed the signal distortion that you have described. To be fair, I was testing the "hopping" on I2 devices and was not looking specifically at signal addition with multiple repeaters.

 

My test setup utilized filterlincs for isolation. As you noted in a different post, these devices do present a load to Insteon signals. This loading may be the difference in our test setups. If your 4mH choke is between your Insteon devices and the Filterlinc, the transmitter outputs may be looking into a high impedance.

 

Aside from not being a real world environment (typical home powerline impedance ranges from 5 to 32Ohms at 100 KHz), this may actually cause problems for newer V5.x Insteon units. SH is using a redesigned switching supply in their newer units. If your 4 mH choke impedance is close to the impedance of the unit power supply (at the control frequency) you could send the supply into oscillation.

 

Please re-try with the filterlinc by itself. I will attempt to set up a parallel test on my end. Can't commit to a date for this at the moment.

 

Yet another observation - you are referring to the schematic for the 2412S and the zero cross detector. I have not seen a schematic that details the zero cross circuit for other Insteon modules (SWL, APL, etc.) but in general, the power supplies of these units are very different. I opened my Appliancelinc and did not see a zener that I could associate with the Zero cross input. SH may be using an entirely different approach on these units.

 

I agree that you should not get too distracted by this anomaly. I don't believe the transmitter loading is real world and I am very selfishly interested in seeing you continue to progress your test monitor.

 

IM

[ELA]

Hi IM,

Thanks for your thoughts.

 

1) I have already tried a Filterlinc only (no 4Mh choke).

It still happens but less frequently because the signal level is reduced by the FilterLinc.

 

2) The schematic I refer to is at the end of the 2412 developers guide. (shows zero cross circuitry)

 

3) I am going to retest a little on this subject today. I need some better data. I posed the question in hopes that someone had a direct answer and then I would ignore what I saw. Since that has not been the case I now want to review my test setup.

 

What I observed was an increased 60hz level observed at the scope when the 131Khz level exceeded a certain level(approx 5V p-p).

I will take another brief look and then move ahead.

 

 

Moving ON:

I was unable to duplicate the odd behavior I saw last week when I ran the same test today.

I rerouted some breadboard wiring in between tests and now believe that I may have had some unintended cross coupling going on that resulted in the odd signals I saw.

At least that is what I am going to blame it on for now so that I can move on.

 

Thanks for your input.