Some thoughts on phase coupling in relation to Access Points

[IndyMike]

Hi Rand,

 

I agree with half of the below - an accesspoint takes a powerline command and converts to RF during at the end of the current hop (time alloted for RF communication). This does not count as a hop, but the AP hasn't put any information on the powerline as yet.

 

The receiving AP must wait until the next message window (Hop) to begin transmitting on the powerline. This is the 1 hop that I'm referring to. If a message were to be transmitted with a 0 hop count, the AP's would not be able to couple the phases.

 

IM

 

 

Thanks so much for all that information. From what I recall of the developer's guide I was led to believe that an AccessPoint repeats a command from the power line to the RF without subtracting a hop and vice-versa. Then it subtracts a hop and repeats back on the method it received on. I am going to have to search for that again.

 

Rand

[Sub-Routine]

Hello IndyMike,

 

The power line needs 5 cycles and the RF uses one cycle. The power line signals always pause for that sixth RF cycle. As soon as the AP receives an RF signal it can begin to transmit power line on the next cycle, right?

 

AFAIK, no hops should be deducted at this point. The transmitted message contains the hop count, not how many cycles it takes to reach the destination.

 

Then the AP modifies the message to delete one hop and sends it back out as RF on the sixth power cycle so further APs can hear it. When they receive it they transmit on the PL without subtracting, then again on RF with subtraction. Until the count is zero.

 

This is what I have assumed according to the Insteon documentation.

 

I have spoken with quite a few people who have seen big improvements installing the http://www.smarthome.com/2406H/SignaLinc-INSTEON-Phase-Coupler-Hardwired/p.aspx especially those with multiple breaker boxes.

 

Thank you for all the time you have spent helping us, I am learning a lot.

Rand

[Illusion]

Wow, heck of a post there IM!! Impressive amount of work. One note just for archival reasons:

 

3) The three branch lines are all 400 feet long with units at 400 ft intervals.

 

100 ft intervals?

 

It all looks great to me, and is pretty close to how I though it should go. Of course, that is not how it went in real life. I put non, responders on the circuit, and if they heard the command, they should have repeated it increasing success. I did not see that. And the APs did not seem to perform per your wonderful drawings in what I would categorize as extensive testing

 

You had also asked about the Leviton filter function being affected by the powerstrip (e.g. your test where adding the filter appeared to improve communications). I'm again at a loss to explain this. Any noise reduction benefit the filter could have provided should have been more than offset by the Insteon signal loss through the filter. Is it possible this was another anomaly?

 

I do not know how a signal gets past a filter as an anomaly. I have no answers here. I am at a loss. I am going with the components of the surge suppressing strips as de-tuning the filter.

 

Rand,

 

These are the hop counts at the PLM from the message sent by the device.

The ISY cannot tell you how many hops are left when it's signal reaches the responder.

 

Okay, so those were replies from the module in question. Got it. I still do not understand the variation in the Hop count as each line in my event viewer was taken from a seperate query. The variation of the hop count was in the same place in the comm each time.

 

I have spoken with quite a few people who have seen big improvements installing the http://www.smarthome.com/2406H/SignaLin ... red/p.aspx especially those with multiple breaker boxes.

 

My tests show the terrific power of passive phase coupling. I would have felt better if the APs did as much in my tests as the coupler.

[IndyMike]

Hi Rand,

 

We may be saying the same thing. I agree with your statement that the AP's synchronize during the 6th cycle of the initial transmission. The AP's first opportunity to transceive to the powerline is during the first Hop window. When they do this, they transmit with 2 Hops remaining.

 

The following is a capture of an extended transmission (11 data bytes) with only my AP's installed. The AP's are synchronizing during the 12th cycle here. The signal during the 1st HOP window is my AP on the opposite phase. It should be transmitting with 2 hops remaining.

 

 

Hello IndyMike,

 

The power line needs 5 cycles and the RF uses one cycle. The power line signals always pause for that sixth RF cycle. As soon as the AP receives an RF signal it can begin to transmit power line on the next cycle, right?

 

AFAIK, no hops should be deducted at this point. The transmitted message contains the hop count, not how many cycles it takes to reach the destination.

[Sub-Routine]

Hi Rand,

 

We may be saying the same thing. I agree with your statement that the AP's synchronize during the 6th cycle of the initial transmission. The AP's first opportunity to transceive to the powerline is during the first Hop window. When they do this, they transmit with 2 Hops remaining.

 

Okay, that is what I was wondering. Thank you for that info.

 

Rand

[Illusion]

IM,

 

Is it possible that an AP that hears a command on the power line will not repeat that command on the power line?

 

I am trying to come up with an explanation for my AP test results. This actually would explain all of them I think.

 

If you go back to my fist post, you will see that I mentioned that my hop counts were actually better with no APs at all. Maybe the AP just sits there and sucks up some signal, but will not repeat it. If the AP sees an Insteon command on the powerline, then gets that same command synced on the 6th cycle via RF, maybe it does not push it back out on the powerline as it knows that it was already there where it is located since it saw it. This senerio also would explain why it seemed to work when it was behind the 6288 filter.

 

In all my test, we know there was some signal getting to the test devices because some of them would always respond. If an AP were plugged into the same circuit, but with higher sensitivity, it could suck up some of that signal, and then not repeat it. This would also explain others complaints that adding APs can actually degrade system performance in certain conditions, which I did in fact see in my tests

[Brian H]

I found with my newer ApplianceLinc Modules. One Access Point for RF devices and a passive coupler worked better than two Access Points. One on each phase.

[IndyMike]

IM,

 

Is it possible that an AP that hears a command on the power line will not repeat that command on the power line?

 

I am trying to come up with an explanation for my AP test results. This actually would explain all of them I think.

 

Hi Illusion,

 

Based on my observations, AP will repeat when there is a valid signal on the powerline. The plot below shows my opposite phase AP repeating a signal that it heard on the powerline (passive coupler installed).

 

In this test, the AP on the opposite phase saw both the primary powerline transmission and the RF transmission from the primary phase AP. It had to make a decision on which signal to repeat. I'd presume that the AP would check the message CRC's and, assuming the powerline CRC was good, repeat the powerline message. I assume it would use the powerline message because it has no way of knowing whether the RF came from a Remotelinc or other transmitter. If there is valid powerline data with Hops remaining, I would assume it would repeat this, rather than the RF, to prevent collisions with other repeating devices.

 

That's how it appears to function in my setup. Unfortunately, both you and Brian have different experiences. I will say that I periodically receive NAK's while my AP is communicating back to my motion sensors. I do have 900Mhz phones which may be contributing to this.

 

I don't think there is any way for a AP to "inform" the plm if it's RF transmission isn't heard, or if it hears an invalid transmission (AP to AP coupling).

 

[Illusion]

Based on my observations, AP will repeat when there is a valid signal on the powerline. The plot below shows my opposite phase AP repeating a signal that it heard on the powerline (passive coupler installed).

 

Of course. As expected. I was just looking for a plausible explanation for my test results that seemed to show the AP failing to work for me. Thanks for dashing that to pieces.

 

I think I have come up with a great product. A test module you link to the system. You plug it in, and the ISY can interface with it. The ISY sends it a command, then through extended messaging the ISY gets a report from that module. Signal level, Hops left, Line Noise, and what Command Sent all get returned to the ISY through the powerline. The module is not outrageously expensive because it has no display, no external user interface or the like. We can move this test module around and see what is going on in our system. I am going to be RICH!. I will sell hundreds of these for virtually tens of dollars profit!

[TJF1960]

 

I think I have come up with a great product. A test module you link to the system. You plug it in, and the ISY can interface with it. The ISY sends it a command, then through extended messaging the ISY gets a report from that module. Signal level, Hops left, Line Noise, and what Command Sent all get returned to the ISY through the powerline. The module is not outrageously expensive because it has no display, no external user interface or the like. We can move this test module around and see what is going on in our system. I am going to be RICH!. I will sell hundreds of these for virtually tens of dollars profit!

 

That sounds like a great idea. Sign me Up!!!

 

Its funny, I was trying to think of a way the ISY could be used to indicate and report things like noise, signal level and such from each device etc., and with your idea of a plug in device reporting to the ISY that info..... Thats perfect!!!! With a device like that and ISY support you could plug the device in anywhere within your house or property and get a report from each location. Seems like that would sure help during troubleshooting.......

 

Tim

[IndyMike]

Of course. As expected. I was just looking for a plausible explanation for my test results that seemed to show the AP failing to work for me. Thanks for dashing that to pieces.

 

I think I have come up with a great product. A test module you link to the system. You plug it in, and the ISY can interface with it. The ISY sends it a command, then through extended messaging the ISY gets a report from that module. Signal level, Hops left, Line Noise, and what Command Sent all get returned to the ISY through the powerline. The module is not outrageously expensive because it has no display, no external user interface or the like. We can move this test module around and see what is going on in our system. I am going to be RICH!. I will sell hundreds of these for virtually tens of dollars profit!

 

Sorry Illusion,

 

I didn't intend to be a "dasher". I'm simply relating how things function in MY system. I became very interested in the reported I2 and V.35 problems months ago. I've tried to reproduce these problems in my system using noise sources, signal absorbers and the like. This has been largely unsuccessful, these things just seem to work in the environment I have. The opposite is obviously true for others. Based on some posts, V.35 devices are only suitable for use a paperweights. I don't seem to have an issue with them.

 

In regard to the diagnostic tool, You can accomplish pieces of this using a second PLM/PLC. At one time I used a PLC and a laptop to go around and monitor HOP counts at various locations. Due to the "addressed" nature of the Insteon protocol, there are definite problems in doing this:

 

1) The PLC has to be linked to all of the scenes in order to hear the transmissions.

2) The PLC will not hear direct device transmissions or queries.

3) Since the PLC is itself a repeater/signal absorber, moving it around in a system changes the configuration. Just installing the device has the capability of fixing/breaking a marginal system.

 

In the end, I wound up causing problems with this mobile setup.

 

I had also approached a "learned friend" regarding a Insteon troubleshooting tool based on the AccessPoint. Said friend currently designs X10 components and tools and is a respected authority in the field. He quickly pointed out that "hacking" an Accesspoint (or developing a device) to listen in to Insteon communications without linking, would likely be a violation of Insteon IP. Such a device could be used to listen in on a homes communication and determine installed devices (addresses). This would be a serious breach of the claimed "Insteon network security".

 

There I go again, dashing hopes. I'll try to do better in the future.

 

IM

[IndyMike]

Hello again Illusion,

 

I've been going over your data trying to develop a rational explanation for the behavior you've documented. While I haven't been able to reproduce the Accesspoint communication problem, I do have some data on how they "should" perform.

 

I've been focused on a number of your apparently conflicting test results:

 

Non-Hostile Insteon Environment. [/i]

Test #3 Put an AP in the circuit and in the PLM (100% failure rate)

 

As expected from my earlier post.

 

Test #5 Added Passive X-10 coupler to the system (11% failure rate)

Pulled the APs

Clearly showing the power of this little jewel

 

Test #7 v.35 IconLinc Relays (4% failure rate)

Added 2 v.35 devices on each phase (these are my outside flood lights)

Removed the passive coupler

Removed Non Responder from circuit[/u]

 

Test #9 Add X-10 Passive coupler back in (9% failure rate)

Statistically insignificant difference

 

The above implies the following:

1) Test #7 - due to their placement,your V.35 units can hear the PLM coupled through the transformer. They repeat the signal and your test units respond well.

2) Test #9 - your passive coupler provides sufficient signal that your test units respond well.

3) Test #3 - something in your test circuit is interfering with the AP and either preventing it from communicating, or corrupting it's output.

 

I base #3 on your results below -

 

So I started thinking, maybe my APs are doing something for me in this setup. I had done another test in the Hostile Insteon Environment that was not listed. Actually I have dozens of test that are not mentioned here. I expected absolute failure in the Hostile Insteon Environment Test #1 but some of the devices always responded. Not always the same ones either, seemingly random, so I tried to isolate the circuit by putting a Leviton 6288 filter in before the power-strips but devices kept responding so I assumed that the filter was junk and pulled it and moved on. Well, now that my system is back together but with lower failure rates I decided to do another test. I isolated 2 of the modules after the 6288 filter. It worked. Total isolation and I could no longer control those devices. I added an AP after the filter and once again had control of the devices. Clearly APs do something, I just do not know why they do what I would expect in some situations and not in others.

 

IndyMike, could some components of the surge suppressing strips be de-tuning the 6288 filter? The filter did not seem to work very well when the strips were plugged into it, but when it was straight AC the filter seemed to be 100%

 

Filters simply reduce noise. Your 6288 is a low pass filter the will attenuate noise in the 130Khz range by ~ 40 db. The noise (or signal) will still be present, but at a reduced level. My interpretation of the above is that you have both significant Insteon signal and noise on the line. Adding the filter reduced the noise and allowed the devices to discriminate the Insteon signal.

 

Your second test, conducted with your entire system in place, used different input conditions. The implication is the the Insteon signal had actually dropped to the point where the filter was effective (filter reduced the Insteon level below your receivers min sensitivity).

 

Most Insteon devices have a quoted input sensitivity of 1mv. This sensitivity is obviously dependent on the noise environment present in the installation. As noise increases, the device AGC circuit (firmware I believe) will reduce the input gain (sensitivity) to reduce false triggers.

 

Accesspoints are a bit of a question mark.

 

1) If the device can reliably receive data on the powerline, it seems it would repeat this information and ignore (or at least delay) rf information from an opposite phase AP.

 

2) If the device is unable to retrieve powerline information, it would again seem reasonable that the device would use the RF couple to repeat on the powerline during the next hop window.

 

3) If the device receives corrupt powerline data (bad CRC) operation is a bit fuzzy. Since the powerline data is corrupt, the device can't verify a RF receive as being the same transmission (i.e. this might be a remotelinc or motion sensor sending RF rather than an AP).

 

I ran some tests to trying to resolve the above, and was partially successful. The following summarizes the tests and what I believe I confirmed. All testing was conducted using the following:

 

1) Passive coupler at panel

2) PLM at panel

3) AP on same phase as PLM - mounted ~ 250 feet away.

4) Filter - Filterlinc model 1626 (bandstop) V1.1

5) Responder LampLinc 2856 (Rev 1.2) connected to a 60 Watt incandescent.

6) Powerstrip - 250' unfiltered.

 

Most of the following was conducted using the ISY direct device mode (I was monitoring cleanup responses). The "Main circuit repeat" column indicates whether a device was installed in the unfiltered output of the filterlinc (close coupled repeater upstream from the filter).

 

 

The schematic below details the "Test 5" setup. The opposite phase AP was installed in the filterlinc "unfiltered" output. The AP was able to hear the LL through the filter and repeated the information on the powerline.

 

 

I'm running out of time but wanted to note one thing that I did learn:

 

The Insteon devices themselves appear to have extremely good sensitivity. Monitored levels with the filter installed were on the order of 10mv P-P and lower. The lamplinc was able to reliably discriminate this low level signal from the ambient noise (impressive).

 

Back for more later,

IM

[Illusion]

IM,

 

Extremely interesting data points. So as I understand your test results, you were able to closely simulate some of my test with similar results.

 

You created a hostile environment with the filter, and yet the LL was still able to respond 98% of the time. Great! Now we have a marginal system that closely resembles the dynamic that caused me to begin all this testing in the first place.

 

In test #9 you have marginalized the system so much that you have complete failure. This closely simulated my Hostile Environment testing. My total confusion comes from the fact that in this situation I added an AP to the circuit and still had high failure rates. Of course in my case I was adding it after the equivalent of your filter on the powerstrip with no improvement.

 

These test results also help explain some results I got that I did not post from other tests. I put a non-responder upstream on the same circuit and got statistically significant higher failure rates. (signal absorption)

 

This helps explain what is going on with my failing office scene, with the exception of why when I add an AP to the weak circuit the AP does not "fix" my issues. It would be like if you added an AP to test #9 on the powerstrip after the filter.

 

This did not help in my testing except in the one test where 2 modules were isolated after the filter. Why the APs did not help when put on the weak circuit in all the other tests I do not understand.

 

Thanks for your time on this IM, you have been very helpful. Good to know you also had similar results from testing the v.35 switches.

[Illusion]

Also,

 

We see that adding devices can actually decrease Insteon reliability. Correct me if I am wrong, but in these controlled experiments this is largely and likely due to signal absorption.

 

If enough devices are added to a circuit, each one absorbs some signal, and eventually there is not enough signal for any of them to respond to. Hence, adding a device on the circuit to act as a repeater has the very real possibility of hurting performance.

 

Am I thinking right here, IM?

 

I still do not get how this could apply to adding an AP to the suspect circuit though.

[IndyMike]

Illusion,

 

Per your suggestion, I ran two additional tests (10 and 11) with the AP on the filtered circuit. Without the AP (LL + 2 units) the response was 0 out of 20 attempts. With the AP, the LL responded 20 out of 20 times.

 

IM,

This helps explain what is going on with my failing office scene, with the exception of why when I add an AP to the weak circuit the AP does not "fix" my issues. It would be like if you added an AP to test #9 on the powerstrip after the filter.

 

This did not help in my testing except in the one test where 2 modules were isolated after the filter. Why the APs did not help when put on the weak circuit in all the other tests I do not understand.

 

Our results are a bit different in that your AP only functioned when on the filtered circuit. I believe you have a noise source in the vicinity that is preventing the AP from functioning properly when it's on the unfiltered side. I've been trying to construct a noise source that would prevent my AP's from repeating, but have been unsuccessful so far.

 

As I said above, I believe your filter was functioning properly. There's not much you can do to upset a low pass filter. The fact that your units functioned with the filter installed in-line indicates that your Insteon level on the circuit is actually quite good (you were getting some signal through to the units). With the filter in place, the AP functioned properly because the noise had been reduced.

 

Anyway, that's the best theory I have at the moment.

 

IM

[Illusion]

IM,

 

More tests and a new theory I am really liking!!

 

I recreated my tests that started this thread, except this time I used 7 modules. When I saw high failure rates I put an AP on the circuit and the PLM and once again it did not help. Then I put a filter on the circuit and put the AP on this side of the filter and low and behold 100% success.

 

It is not noise! I used my cute little Elk meter and the line was quiet. It is some kind of data corruption that is at issue here. That is my current theory.

 

By creating a hostile environment where the Insteon has to travel long (maybe through the transformer or a neighbors AC motor) the AP on the circuit does not fix the high failure rate. But filter the circuit and let the AP create a new signal and perfection ensues.

 

When I had high failure rates in my test I still was showing nice strong 2V signals (on the Elk meter's scale), but only one or two modules would respond. The AP did not help this when applied to the circuit. When I filtered the circuit I got nice strong 5V signals as the AP was placing the signal on the line right where I was metering. I think that inductance from high turn coils is offsetting the signal, corrupting its timing, and preventing the AP from trying to reproduce the signal. I could be wrong on all this, but I think the problem is along this thinking at least. I wonder if this could also explain why putting a non-responder on the circuit did not help in so many of my tests. If the signal were mucked up, I am guessing a module would not try to repeat it. But some of the modules did respond to the command, so maybe only part of the signal was damaged. Maybe responding to a command requires less signal perfection than repeating a command does.

 

The interesting thing here is that I may have found a way to fix that troublesome circuit of mine. If I isolate the circuit completely with a filter and add an AP to the circuit then I should be at 100%, whereas adding the AP as I have been without isolating the circuit does not help.

 

I am also seeing the potential to solve Insteon comm issues by isolating the PLM behind a filter and using an AP to get the signal out there. It seems counter intuitive, but may actually improve some situations.

 

IM, what do you think? I like this line of thinking as it also has the ability to explain all of my seemingly conflicting test results. Your clearly have some theory that is at least in this vain as per this quote of yours from an earlier post:

 

I do have a theory on why AP's (and devices in general) may have a problem operating as repeaters, but I've gone on long enough for one post. As a teaser, it has to do with localized noise and devices improperly detecting the 60 Hz zero crossing.

 

Also I should take this opportunity to say this:

 

I recreated IM's test 5,6,7 and was shocked to see the modules respond with a rate of 100% behind the filter by putting a non-responder repeater on the up-stream side of the filter. Really incredible when you think about it! We are all very quick to condemn Insteon and Smarthome but that result is really remarkable. They have single-handedly created a standard and products that really do work most of the time and they do it at a price lots of people can afford. All the while, the oft talked about alternatives still do not have a deep product selection for me to buy and use. If we want perfect reliability we can always pony up the dough for a professional hardwired networked system that will be really near 100% but cost as much as a house. The Insteon system, while brilliant in its Hop count and all modules repeating a command, does suffer from a complete lack of control of the local power line environment.

[IndyMike]

Ok Illusion,

 

You've been holding out on me - you have an ELK ESM1 and only now fessed up (kidding).

 

The Elk is very useful in detecting noise in the X10 frequency range. Unfortunately, noise comes in many different types and frequencies which the Elk will not detect.

 

The bottom line here is that there is something on that circuit that prevents repeaters from boosting (Hopping) the Insteon signal. This is supported by the following:

 

• 1) A repeater (AP or normal module) placed in line does not improve reception.
  2) When when a filter is placed in line, a downstream AP functions normally.
  3) Repeaters placed on other circuits (your V.35 units) are able to properly boost/hop the signal.
  

 

It is not noise! I used my cute little Elk meter and the line was quiet. It is some kind of data corruption that is at issue here. That is my current theory.

 

Sorry, but I don't like the idea of isolating the PLM and relying on RF. It will burn to many hop counts.

 

My Zero crossing idea doesn't work well here. High current switching devices can affect the 60hz waveform. Since the Insteon (and X10) devices use the waveform zero crossings for timing, this can have disastrous effects on communication. I'm not sure that this makes sense since the 6288 filter would have no effect at 60Hz.

 

I am also seeing the potential to solve Insteon comm issues by isolating the PLM behind a filter and using an AP to get the signal out there. It seems counter intuitive, but may actually improve some situations.

 

IM, what do you think? I like this line of thinking as it also has the ability to explain all of my seemingly conflicting test results. Your clearly have some theory that is at least in this vain as per this quote of yours from an earlier post:

 

I do have a theory on why AP's (and devices in general) may have a problem operating as repeaters, but I've gone on long enough for one post. As a teaser, it has to do with localized noise and devices improperly detecting the 60 Hz zero crossing.

 

It's almost as if you have a noise make that is synchronously locked to your Insteon transmissions. I've seen Boosterlincs do exactly that. They can be fooled into trying to boost the Insteon signal and wind up polluting it with X10 noise.

 

Don't have any boosterlincs or boosterlinc enabled modules installed do you?

 

IM

[Brian H]

Insteon signals start 800 us before the Zero Crossing.

Maybe noise in that time frame could effect things.

[Illusion]

IM,

 

No bossterlincs or bossterlinc enabled modules. I do have an active X-10 repeater, but that is not in play here. Nearly all my test have always had that device out of the system. Even if it was not out of the system I have not been able to find any negative effects of putting in back in. This was proven through EXTENSIVE testing.

 

But in any case, it was not a player in the system for almost all the enumerated tests in this thread. That said, I agree that something is preventing repeating. I think it is the Insteon devices themselves. I am not sure exactly what dynamic causes it. I do not think I have the filters and supplies to create an environment that can prove this theory. I am still thinking about it though.

 

The office circuit that caused me to start all this testing is a different circuit than the one used for the majority of the testing, yet it also is not helped by adding an AP.

 

Further, when testing on the test circuit, virtually every electronic item on that circuit is behind a 6288 filter.

 

I believe I have a nice quiet circuit. I have looked at it with a Graphical Multi-Meter and it looks as nice as can be. I know that is no O-Scope, but that combined with the ELK and my ability to get 100% with a passive coupler I think helps prove that. It is a nice quiet circuit except when I start adding Insteon signals to it.

 

I am sticking with my theory that the Insteon devices themselves have the ability to conflict with each other on the power line. I am not educated enough to understand this, but we have lots of people on this forum cautioning against adding too many APs. Well in the Insteon system and protocol, we should be able to add 100 of those and the system should just keep getting better and better. But it does not! I am no where near the only person who has not had issues resolved by adding an AP when logic says it should help. I think my tests show that. I do not know why my test confirm this, but there it is.

 

I am still working on a test that you could perform to get the same results I saw. I know it is repeatable, and I am sticking to my guns that it is not noise, unless you consider transmissions by Insteon noise. The key thing is that when the APs do not work for me, the signal is traveling all the way to the xforrmer and back. The test circuit shows high failure rates as expected, but adding an AP or other non-responder (repeater) at this point does not help. In fact is seems to hurt. But now filter the very circiut and the AP will reproduce the Insteon from its RF, but it will not do this if the circuit is unfiltered. If noise was the culprit just adding in the passive coupler should not massively improve things, but it does. It does not all add up.

[IndyMike]

Illusion,

 

OK, I'll back off on the noise theory. I think our basic problem here is that we do not understand how the AP's react when they see a garbled Insteon powerline transmission and a valid RF transmission.

 

From your description, the Insteon transmission is being garbled as it passes though the transformer and your devices on the opposite phase refuse to repeat it. Furthermore, the AP refuses to repeat the valid RF signal.

 

I also do not understand how you can be getting a garbled 2V level through the transformer. I view a transformer as a passive device. It should attenuate and phase shift the signal. At 2 volts you have plenty of signal coming through. I don't know what the transformer could be doing to the signal to cause your units to reject it.

 

Assuming the transformer is the problem. could you try reversing your configuration?

1) Filter your PLM with an accesspoint installed.

2) Plug your test units and another accesspoint directly into the opposite phase.

 

The filter should prevent signal from looping through the transformer and you'll have an RF connection to phase B.

 

If this works, you've made a believer of me.

 

IM

[Illusion]

Cool idea IM,

 

It will be a while before I can try that test sequence. It is hot here in Dallas, and I cannot ask my neighbors to turn off their ACs any more for a while. I am turning into that annoying neighbor.

 

I am not saying that the transformer is causing this IM. I am saying that currently I have a strong correlation with that test condition. As we know causation and correlation can be from very different simultaneously occurring conditions. I will keep thinking about this, but it is tough to test with such limited tools and such limited control of the power line. What I really need is an isolation transformer to create my own power line that I have complete control over.

[Illusion]

IM,

 

Your experiments gave me an idea. I wanted to come up with something to prove that it was not noise causing my problems (outside of other Insteon devices). As I am unwilling to shut off all phase coupling (read neighbors AC and Water Heaters) for additional testing, I came up with a better test based on your tests.

 

I put a 6288 filter in.

 

Put the PLM on this side of the Filter

 

Created a simulated Insteon network on this side of the Filter with Insteon Devices, X-10 devices (mini controllers and maxi controllers for signal suck) 6288 Filters, FilterLincs, Cable loops, and surge suppressing strips (for signal suck) etc.

 

Changed the relationship of those devices until I had created a dynamic that cause high failure rates talking to an APL on the simulated network

 

Then added an AP to the back of the PLM and right next to the problem APL

 

And low and behold, I still saw roughly the same high failure rates. I watched my ESM1 and there was still lots of signal there, even on the failed attempts to control the APL. I had lower failure rates during some tests without the APs with no visible signal on the meter at all. I am not educated enough to figure out why this is though.

 

I do not believe I had any noise generators on my simulated and isolated power line.

[Illusion]

Note on last post:

 

I have not been able to recreate the previous isolated Insteon network test sequence with similar results. I am unsure what variable changed. I now begin to wonder if some devices (PLM, Access Points) do not always provide consistent behavior.

[IndyMike]

Hi Illusion,

 

Sorry for the long delay in responding. It has taken me a bit of time to assimilate your test data...

 

As you indicated, some of your test data conflicts with earlier results. Your latest test setup appears to have eliminated phase coupling as a variable, yet it produces worse results.

 

The following is what I have envisioned for your test circuit. The modules on the right side of the schematic are simply intended to represent loads. Please confirm whether this is correct.

[Illusion]

Yep IM, that is the test network. You understood perfectly. I did have some additional devices in between the filters on some tests though. I did dozens of variations trying to learn from the results. Mostly what I learned is that I am out of my league, and the ESM1 is not going to help me understand what is going on.

 

Thanks Guys!

 

I have been adding access points (and filters) in an attempt to increase reliability, it is good but not as good as I think it should be, and with a few troubled spots.

 

I am up to 8 access points but they seem to do little good. They don't seem to help even when putting them right on top of a two devices that have trouble with each other.

 

I just got a hardwired phase coupler, I just need to figure where to put it (Full breaker box).

 

I am glad I am not imagining things,

 

Nick

 

If the APs worked like I thought they should, it would seem to me that 8 APs are better than two. If all the APs get the RF at the same time, and lay a newly created signal on the power line in unison, things should be great. These things are not the problem solver I think they should be. I personally have lower reliability with my real world system if I add APs to try and solve problems.