Insteon Sucks just a little less than X-10 did :(

[ELA]

I am upset that I am having to spend so much time diagnosing Insteon communications issues.

I recently completed upgrading all my devices from X-10 to Insteon just because I was tired of playing around with X-10 communications issues.

 

I am hoping for some insights to renew my faith. I am an electrical engineer so I am capable of diagnosing signal levels if required (using Oscope), but do not feel that should have been required. After all this is INSTEON!

 

I have about 30 Insteon devices - No X-10 devices. Using the ISY99i which is a dual band unit. I have three other Dual Band Lamplics (DBLL).

This makes for (2) RF devices on each phase. Small 1300sq ft house.

 

I was having issues getting commands from a KeypadLinc to the ISY99i. I moved one of my DBLLs closer to the service cabinet after monitoring signal levels and seeing they were weak at the ISY99. This increased the signal levels and that particular key worked better.

 

This morning another function from the same keypad failed. The command made it from the keypad to the ISY but did not propagate fully back to all devices. A second keypress a moment later and it worked fine.

 

I am concerned about this. Why should a command fail and then a moment later work fine (when no device states were changed in between)?

 

In addition after moving the DBLL now a 2420M is failing to get acknowledgement (fast LED blinks). There are two DBLL within 25 ft of it. The DDLL I moved was within 10 ft of this 2420M previously.

 

I feel that I now need to add another access point closer to the 2420M to get it happy again.

 

I have been reading up on the Event viewer and will increase it to level 3 and do some testing. What a pain this is turning out to be.

 

Yes I retained several filterlincs in the system on the known noise generators and signal sinks.

 

One question I have is:

I am using Dualband lamplincs. Are the normal access points any stronger RF wise?

[Walrus]

I'm new to this, but so far I have found the dual band PLM has a crappy range. I have it plugged into a ceiling mounted plug in my basement. I couldn't link a 2420M motion sensor on the main floor unless it was directly above where the PLM is plugged in, I'd say it worked no more than 12 feet away from the PLM.

 

Possibly the DBLL are the same, and you need 2 access points.

[LeeG]

ELA

 

Sorry you are having comm problems. They can be difficult to identify. From posts on various forums related to Insteon that I follow some folks have indicated a preference for Access Points as they seem to have better range. I happen to be one of those but have NO objective proof. I keep an Access Point at the ISY PLM plug point and have two Access Points placed at outlets about midpoint of the house that provide 120V leg coupling and RF coverage. I think all the Dual Band devices add to the reliability of the mesh network, I just prefer Access Points as the primary phase coupling and RF support devices.

 

Unfortunately powerline communication is rarely an absolute, it works or it does not. There are cases where certain devices have completely stopped Insteon communication on a particular circuit but more often than not communication failures are intermittent because of marginal signal strength or intermittent noise.

 

There are some things to look at that are often involved. CFLs are common sources of interference. Seems to depend on the manufacturer. Many CFLs are fine, I have them all over my basement. Folks often find they cause problems when they are turned On. Some cell phone charges will wipe out powerline communication due to the noise they place on the line. There was a case just a few days ago where there were two switches in a J box, one worked and one did not. Cause was a TV that when unplugged completely resolved the problem. When you mentioned service cabinet that often means a collection of devices that can absorb powerline signals. PC and UPS power supplies are often sources of signal attenuation which can easily be solved by putting all those devices on a Smarthome FilterLinc. An easy test is to put the PLM on an extension cord to move the plug point away from a collection of PC/UPS/AV equipment. Make sure Insteon devices are not on power strips that have surge suppression. One technique for identifying problem devices is to unplug them and see if things improve. You are actually in a better position than most as you have a scope to observe the results of unplugging devices.

 

Some X10 devices with BoosterLinc technology interfere with Insteon signals. Two weeks ago there was a user new to Insteon. One SwitchLinc on a 3-way circuit worked, the other SwitchLinc on the opposite end of the traveler wires (rewired to supply unswitched 120V) did not. Turned out to be an old X10 KeypadLinc with BoosterLinc.

 

The one thing to keep in mind, once the sources of interference have been identified and resolved, Insteon is a very reliable powerline protocol. I spent years trying to get this house working reliably on X10. The best X10 coupler/repeaters on both power panels. Never did achieve the reliability I wanted. When Insteon came out I purchased a KeypadLinc installed in the living room and an ICON switch installed on a dock some 200’ away. This was always unreliable with X10. Insteon worked all the time right out of the box. That one test installation convinced me to move to Insteon. One I have never regretted from a technology point of view. I have all my computer equipment and UPS on a FilterLinc. Also have other UPS devices scattered around the house on FilterLincs. Found an LCD flat screen TV was also causing a problem that a FilterLinc resolved.

 

There are other powerline protocols such as UPB (I think) that operate on a higher signal strength. These have less issues from powerline interference but are much more expensive devices. Insteon will be reliable when the sources of interference are resolved.

 

Lee

[mmknox]

I am in a similar situation to ELA. One of the reasons I hung on to X10 for so long is that I really liked my old Enerlogic controller. It had a very powerful programming language, but was DOS-based and wasn't Y2K compliant. I had to keep an old PC around with DOS 6.2 and set the year back into the last century whenever I had to make a program change! When the old Enerlogic finally died, the X10-branded interface just didn't cut it, so I moved to Insteon.

 

I have pulled out all my X10 devices except for a PowerFlash/Universal combo in the garage for door status/control (will replace this weekend with Insteon IOLinc) and a couple of Leviton branded in-wall controllers (to be replaced this weekend with Insteon SwitchLincs). I bought a ISY/99iPRO controller and am starting to get the hang of it.

 

The most frustrating thing for me is that intermittently, some of my devices fail to turn on or off on schedule. I'll come home to find one or two of my four outside lights didn't come on (never the same ones!) or one of my two bedroom wall sconces (with switches right next to each other) won't turn off. Yet when I manually send on/off commands from the ISY through the same PLM that it uses to run schedules, everything works perfectly!

 

So far I have:

 

1. Systematically checked EVERY electrical connection in the house, including the entire breaker panel. All tight and good.

2. Re-positioned my 2443 Access Point pair (for phase coupling) several times. (I also have a hard-wired coupler at my panel).

3. Installed almost a dozen 5 amp, 10 amp and 15 amp noise filters on all the "usual suspects" even though I could never firmly establish that the devices were causing interference.

4. Tried running on/off commands with various combinations of other electrical things on or off.

5. Tried plugging my 2430 Tabletop controller in various outlets and running scenes.

 

Whenever I am there and watching, everything works fine. As soon as I turn my back and leave it to run automatically things start to not work!

 

With some help on these forums, I've added some programs to periodically run Queries and take action if devices are found to be in the wrong state (i.e. a porch light that didn't go out at sunrise).

 

Insteon is better than my old X10 system had gotten, but then some of my X10 devices were over 20 years old (some still with the original owner, BSR, branding!).

 

Right now, I'm hanging my hat on the premise that more Insteon devices will improve the situation due to their message forwarding capabilities. I currently have a dozen or so responders in addition to several controllers and, of course, the dual band PLM. We'll see as I start building out further.

[ELA]

Thanks for the responses,

I do so want this install to work reliably but it is definitely not as simple as the sales make it out to be.

The other issue is that my wife gets very irritated at me working and reworking the system when it does not work right

 

As I mentioned I am somewhat experienced from working a lot of X-10 issues in the past.

 

I hate CFLs. Have none in my house. I do have some Fluorescents which can have the same sink and noise issues as CFLs. These are not on at the time of my Insteon failures.

 

All other noise and sink devices have Filterlinc filters on them.

 

No X-10 in house.

 

I would like to take this opportunity to clarify that to simply say, "DO NOT put Insteon devices on power strips with surge suppression" is FLAT OUT INCORRECT!

ALL Insteon devices contain surge suppresion devices inside the device itself!

 

What they mean to say is: do not put devices on a plug stip that contains an EMI filter. This will suck signal strength. What is important is the difference between an MOV = OK, EMI filter = Not OK.

 

I will work through this. I am going to start by adding two actual Access points in hopes that they send/receive a stronger RF signal.

I am just very upset at the moment that Insteon is not as robust as claimed. The fact that they are having issues is highlighted by the introduction of all these new Dual Band devices.

 

 

I do appreciate your input.

 

Regards,

Ela

[Goose66]

I have a 4600 sqft house with two separate breaker panels. I have 3 Access Points and all UPSs on older X10 filters (3 5amp and 1 20amp). I would say my system is 80-90% reliable. It does appear that when it is not being reliable, it is easily recreatable. So something being on is likely my problem. Could be any of our appliances, the CFLs throughout the house, who knows.

 

For my old X10 I have diagnostic equip. to diagnose such reliability issues. I can't find such equipment for Insteon. I would like a plug in device that logged commands, hops, and signal strength at that plug. Anybody know where to get one of these?

[ELA]

I am in a similar situation to ELA.

 

2. Re-positioned my 2443 Access Point pair (for phase coupling) several times. (I also have a hard-wired coupler at my panel).

 

Right now, I'm hanging my hat on the premise that more Insteon devices will improve the situation due to their message forwarding capabilities. I currently have a dozen or so responders in addition to several controllers and, of course, the dual band PLM. We'll see as I start building out further.

 

My sympathies mmknox,

I am a little discouraged by your note about your access points. I will still try them though.

 

I have detailed my system on a drawing to attempt to understand why it is not working.

Looking at the diagram - I would have expected additional device "message retransmission" to have helped but I am not seeing that in my system. Not large enough ??

[unispeed]

Boy, I am with you guyz....

I am a Mechanical engineer by Have been in the process and building automation for 25 years.... This Powerline is killing me... I know one thing is that if there is a signal that cand be interpreted than it can be measured..

I am ready to invent create or what ever it takes to help these guys out (including me).....You EEs think we could get something to measure strength....?

Mark

[Goose66]

You EEs think we could get something to measure strength....?

Mark

 

Or even a utility software program that I could put on a laptop and move around the house with a PLC or PLM attached.

[mmknox]

On a somewhat related note, does anyone know how many "attempts" the ISY makes before giving up? I have seen in the logs reference to 3 "hops" and am wondering if the answer is 3.

 

I get that you don't want to make that number infinite as it'll hang on a missing device etc., but I think a lot of my problems would be resolved if that number could be increased a bit.

 

Is this configurable in the ISY or is it something that's hard-coded into the Insteon protocol?

[Brian H]

I agree. The 2413S PLM is not as good as my pair of older revision 1 Access Points. When it comes to phase bridging or signal repeating.

Before installing the 2413S on my ISY99i.

I tried it in both of the older Access Point locations.

In both locations. It fails the communications test with the other Access Point.

 

Insteon signals can be effected by power line noise and signal suckers just like X10 is.

 

I use a JV Engineering XTBM X10 signal meter and it is great. We have asked Jeff Volp if he would ever consider an Insteon version. Not in the plans right now.

Most times if I find an X10 noise maker or signal sucker. It also would be a problem with Insteon.

 

I may have stuck with X10 as I did get the JV Engineering XTB-IIR that blasts over ten volts of signal back to the modules, but I had committed to Insteon. The ISY99i sealed the deal as it was much better than my older X10 controllers.

[LeeG]

Max Hops and retries are related but separate. When an Insteon Direct command is issued the application decides what Max Hop value to use. The Event Viewer trace record of an On issued by the Admin Console shows the Max Hops count is set to the maximum value of three. The 0x0F in the following command specifies (among other things) the Max Hop count in the last two bits which are 11 which is 3 in binary.

 

Wed 01/12/2011 03:08:18 PM : [iNST-ACK ] 02 62 04.56.50 0F 11 FF 06 LTONRR (FF)

 

The Max Hop count controls how many devices are allowed to repeat (not retry) a given message. The greater the Max Hop count the better chance a message has for reaching a given device. The higher the Max Hop count the slower the response from the responder device. Has to do with how long the responder has to wait for all the possible repeated messages to arrive before the responder can send an ACK. The insteondetails.pdf document on the Insteon.net web site has an extensive description of Hops, Max Hops, time slots, and the effect on powerline data transfer rates.

 

Message retry is something else. The Controller device, where a command is issued, where a paddle/button is pressed, automatically retries most Insteon messages. The retry of a message is independent of the Max Hop count but there is a relationship. If the message was originally issued with a Max Hop count of 1 (best performance) but an ACK is not received from the responder, the Controller will bump the Max Hop count by 1 for each subsequent automatic retry. Again, the insteondetails.pdf document describes all of this in detail if you want to know how in internals of Insteon message protocol work.

 

Max Hop counts and retries are not something you have to be concerned about. The Insteon Controller takes care of retries automatically. Does make for some interesting reading if you have an interest in how the Insteon technology works.

[mmknox]

Thanks, LeeG.

 

If I get some spare time, I would like to read up on the details in the insteondetails.pdf document.

 

I guess I was just hoping for a "try a little harder" or "try a little longer" option that I could configure in the ISY

[Michel Kohanim]

Hi Guys,

 

I just wanted to jump in and provide our experience with INSTEON in 1000s of installations:

1. I have never seen an installation that can NOT be fixed

2. When it starts to work - and as long as something does not go bad - it works quite reliably

3. The major sources of unreliability have been SwitchLinc v35s. If you have any of them, get them replaced

4. 2413S PLM is much more reliable than the previous models

 

And, I know how HARD it is to find where the problem is originating from. Actually, it's quite frustrating especially due to the lack of troubleshooting tools, one has to basically figure things out based on trial/error.

 

And, if you do not have 100% reliability, there's something wrong. Please do not ignore it as usually they may get progressively worse.

 

And, finally, if you need help please do not hesitate to contact our tech support team.

 

With kind regards,

Michel

[ELA]

Hi Guys,

 

I just wanted to jump in and provide our experience with INSTEON in 1000s of installations:

2. When it starts to work - and as long as something does not go bad - it works quite reliably

3. The major sources of unreliability have been SwitchLinc v35s. If you have any of them, get them replaced

 

And, I know how HARD it is to find where the problem is originating from. Actually, it's quite frustrating especially due to the lack of troubleshooting tools, one has to basically figure things out based on trial/error.

With kind regards,

Michel

 

Michel,

My system was working well. As it grew in size it it got worse, not better.

I will check for any Switchlin v35s. Can you tell us why they are a problem?

 

I have used an Oscope to measure signal strength with a 0.2uf cap and 100 ohms resistor. I am in the process of using a modified Powerlinc to better filter out the 60Hz.

Taking adavantage of the signal transformer inside the unit.

 

I had read the "insteondetails.pdf document" previously and that helped some. If I remember correctly it talkes (5) zero crossings to send the standard message. How long after that will you see the ACK message?

 

 

Q1) Do you think that adding a set of standard access points would be better than my (3) existing dual band lamplincs + DB PLM?

 

Q2) Do all powerline based devices repeat messages? Did I read right that they repeat (or boost) right on top of the original message?

[mmknox]

Thanks Michel.

 

I would like to say that I am very impressed with my ISY/99i controller and the support it gets here on the forums. And in fairness, my Insteon setup is working much better than the X10 system it replaced.

 

I am curious, however, about how to track down a problem you can't replicate. I have tested everything I can think of, turning things on and off through the console in all possible combinations and it works great. I then go to work one day and come home to find my porch light off. The ISY reports it On until I run a query and then it shows it as off. I initiate an On from the console and the porch light turns on! Unfortunately, I'm not home when the signal problem occurs (in this scenario), but I can't imagine the electrical conditions in my house are any different than the scenarios I tested in person. The next day, or week it may work fine, then it's another device that doesn't go on or off.

 

Do you think a running refrigerator or chest freezer could be the culprit? Unfortunately, I can't unplug these for days on end to see if my intermittent problem goes away (but I have tested locally with them running and it seems to work fine then).

 

Just wondering what to try or do next in this scenario.

[LeeG]

ELA

 

How long after that will you see the ACK message?

It depends on the max hop count of the priginal message the device is ACKing. In the scenario where the device sees the message on one hop but the max hop count is 3 the device has to wait for the possibility that additional messages will be received because of more hops before placing the ACK on the powerline. Otherwise one message could step on the other. I don’t know the timing of the various possibilities but I think the insteondetails.pdf covers all the variations.

 

Do all powerline based devices repeat messages? Did I read right that they repeat (or boost) right on top of the original message?

They do all repeat. The message is repeated in the next time slot. If it was simply amplifying the same time slot message there would be no need for the responder to adjust the ACK time slot based on hop count versus max hop count. The insteondetails.pdf document has several graphics showing various time slot usage for repeated and acknowledged messages.

 

Lee

[Michel Kohanim]

Hi mmknox,

 

Thank you.

 

From my experience - and it seems you have scenes that are not activated reliably - in the majority of cases, the problems are this:

1. SwitchLinc v35 ... this is not your case

2. The load on devices participating in the scene. i.e. when they are turned on, signal reliability goes down

3. Too MANY or Too little number of Access Points especially in the case of having to bridge phases

 

From what I understand, I think you have a problem with #2. Have you tried Scene Test on your scenes while all the members of the scene are ON?

 

With kind regards,

Michel

[IndyMike]

ELA,

 

I glad to see that you're rolling your sleeves up. However, as Michael mentioned, this is not a trivial task given the insteon "Hop" and "repeat" protocol. That said, I've been down this road and found it extremely enlightening. Being and engineer, I can't think of any other method of rationalizing "Insteon theory" other than using an oscilloscope. But then that's me - I'm so anal, Websters uses my photo for a definition...

 

I doubt that your 0.2u / 100 ohm High pass filter will give you the resolution you need to pick up lower level Insteon signals. You simply won't have enough 60 Hz rejection.

 

You may well be able to modify the Powerlinc to suit your needs, but there is a far easier path. The ACT CP-000 is a tuned coupler with transformer isolation (safe). Slap a plug on one end (powerline interface) and connect the opposite to your scope (loading of your choice) and you have a very good 120 - 140 KHz interface with excellent 60 Hz rejection. These can be had for $7.14 at Smarthome today.

 

Assuming your scope is dual trace, you should also consider a 60 Hz interface for zero crossing correlation with the Insteon signal. A simple voltage divider will work, but we again dealing with safety issues. A simple 24V transformer works nicely. BLH would likely have some good suggestions here (he's been "around the block" almost as many times as I have - no disrespect).

 

I cheat and use the ACT Scope-test2. It's not isolated, not cheap, not widely available.

 

The following is an example of what you'll be observing. It's an Insteon Extended message (used during programming) and shows the "hop" nature of Insteon. The blue traces are the Insteon communication packets and the red traces are the 60Hz content.

 

 

This was conducted on a power strip close to the PLM. I could trigger on the initial insteon transmission from the PLM.

 

If you are trying to determine signal levels at "other" points in your home, the signal levels will vary wildly and you will not know which "HOP" you are triggering on. This isn't X10 - you don't have a simple command signal. You have commands, hops, and retries to deal with.

 

The following is the same setup with the "destination device" removed. The low level signals are being passed back through a passive coupler from an accesspoint (Hop 1).

 

 

So where am I going with all this - I am most definitely not trying to talk you out of measuring your communication signals. I am attempting to show you the magnitude of the task and implore you to thoroughly review the Insteon white paper (it's really quite good and concise, albeit with some PR mixed in).

 

I'm very much looking forward to any results/conclusions you may come to,

 

IM

[ELA]

ELA

 

Do all powerline based devices repeat messages? Did I read right that they repeat (or boost) right on top of the original message?

They do all repeat. The message is repeated in the next time slot. If it was simply amplifying the same time slot message there would be no need for the responder to adjust the ACK time slot based on hop count versus max hop count. The insteondetails.pdf document has several graphics showing various time slot usage for repeated and acknowledged messages.

 

Lee

 

Thanks Lee,

I read and thought I understood timeslots.

Then I reread the section on pg 40 of the "details" on Simulcasting. That is where it talks about devices sending at the same time??

Gets confusing fast.

[IndyMike]

Sorry for butting in,

 

yes the device transmit synchronously (on top of each other). The 60 Hz zero crossing is used for synchronization. There is some phase cancellation from distant devices, but in general the "Mesh" signal is increased substantially.

 

Yes - it does get confusing.

[ELA]

ELA,

 

 

I doubt that your 0.2u / 100 ohm High pass filter will give you the resolution you need to pick up lower level Insteon signals. You simply won't have enough 60 Hz rejection.

 

You may well be able to modify the Powerlinc to suit your needs, but there is a far easier path. The ACT CP-000 is a tuned coupler with transformer isolation (safe). Slap a plug on one end (powerline interface) and connect the opposite to your scope (loading of your choice) and you have a very good 120 - 140 KHz interface with excellent 60 Hz rejection. These can be had for $7.14 at Smarthome today.

 

 

IM

 

Hi IndyMike,

Thanks for all your input.

The cap and resistor work acceptably with the 60hz content at 1V as long as the signal strength is strong. I was worried I might be missing some lower level signals though. So I do want to improve it.

Your suggestion of the ACT CP-000 sounds just what I need and I will look into that. Thanks for mentioning it.

I am also an engineer am very proficient at using a scope. My issue is learning the communications protocol, even though I really had hoped not to have to

 

In your other post you confirmed that devices simulcast or repeat over the top of each other. Is this only in special cases or all the time? How does that relate to the hops and timeslots? Is the signal level observed in each time slot the result of all devices combined then?

 

Thanks for your time,

Ela

[IndyMike]

In your other post you confirmed that devices simulcast or repeat over the top of each other. Is this only in special cases or all the time? How does that relate to the hops and timeslots? Is the signal level observed in each time slot the result of all devices combined then?

Ela

 

Hello again ELA,

 

After the initial transmission, all devices within reception range will synchronously repeat the message in the next "timeslot" (can't remember if this is timeslot 1 or 2 - nomenclature). The white paper defines a "progression" of how listening devices repeat (HOP) transmissions.

 

The repeated or "hop" signal is the summation of all the transmitting devices.

[LeeG]

For example 1 the messages are repeated in unique time slots. DeviceA sends a message to DeviceD. The only device to receive the message from DeviceA is DeviceB. The message is repeated by DeviceB to DeviceC which repeats it to DeviceD. Each repeat is in a different time slot.

 

DeviceA to DeviceB to DeviceC to DeviceD

 

For example 2 the messages can be simulcast. DeviceA sends a message to DeviceD. DeviceB and DeviceC receive the message sent by DeviceA. DeviceB and DeviceC each repeat the message from DeviceA in the next time slot. This puts the DeviceB and DeviceC message in the same time slot as DeviceB and DeviceC do not know that the other has also received and repeated the same message.

 

DeviceA to DeviceB to DeviceD

DeviceA to DeviceC to DeviceD

 

These are simple examples. When a house has50/60 devices scattered over various circuits the potential repeat/simulcast situations get complex. Assuming the same Max Hop count is used in example 1 and example 2, DeviceD sends the ACK in the time slot for both examples. DeviceD cannot know whether there will be another repeat message after receiving the simulcast message from DeviceB/DeviceC.

 

The combination of unique versus shared time slots makes no difference. There is no control over how this works out. In a perfect world a message from Device A to DeviceD would result in a repeatable time slot usage. In reality noise can disrupt the message to any given device so the end result can change even though the same DeviceA to DeviceD message is issued.

[Goose66]

As far as SwitchLinc v.35, what is the current selling version? All mine seem to be v.27. Would those have a similar effect, or is it specifically v.35 that is the problem and not just anything old.