IndyMike

Thank you Paul. I am familiar with the switch monitoring technique. I've been using it since my first ISY26 back in 2007. It most certainly does work. The technique allows me to monitor each floor of my house (1st, 2nd, basement) and display the status on keypads so I can turn monitor/turn off sections of the house as I am leaving or going to bed. It is also, unfortunately, very easy to break. Devices moving in/out of scenes, devices being replaced by dissimilar devices, naming issues all contribute to broken status programs over the years. With IF statements of 25 to 75 devices, it's easy to miss things. This maintenance issue could be avoided if we had access to the scene state (on/off). Somehow the ISY is communicating this information to Home Assistant (very nicely). As @Guy Lavoie stated, this is a feature request - not a problem to be solved through a work around.

This discussion has been going on since scenes were implemented by Smartlabs. The bottom line is Insteon does not support either Scene "Status" or "control states" and creating an architecture for this has been LOW PRIORITY for UDI. One workaround that I have found is to use the Home Assistant integration for the ISY. Home Assistant treats scenes as "switches" or "relay devices". They are either ON or OFF. Curiously, I noted that Home Assistant shows scenes as active (ON) if ANY scene member has a Level > Off. Conversely, all scene members must be Off for a scene to be shown as Off. I use this functionality for triggering programs on Home Assistant. I seriously doubt that Home Assistant is divining the scene status on it's own. Unfortunately, we don't have access to the information on the ISY platform. Seems a little ridiculous, but there are far worse things in life.

That's it. The top "On" button and bottom "Off" button on the 5 button KPL are essentially Non-toggle buttons. The ON can only send On commands (or bright) and the Off can only send Off commands (or dim). As we discussed above for the Motion sensor, multiple consecutive ON commands will NOT show up in the Log (same for multiple Off). The device communicates and will show up in the event viewer (and trigger programs), but repeated commands will not be reflected in the logfile. I have never used the "buttons grouping" feature. As recommended by UDI, I use scenes to accomplish similar things. Look for programs that may be triggered by the devices turning ON. You may want to disconnect the loads to see if this eliminates the issue. Good answer. I've used the vacation mode feature on a few controllers. Being from the "dogpatch" section of the country, I don't have a real application for the feature. I can understand it's value for others. Not sure how repeatable this problem is... If you manually run your program can you force the problem to occur? If so, you can copy the program and try removing devices to see if the problem goes away. I think there is a way to accomplish what you want. It involves switching the KPL to 8-button mode and then joining the A+B and G+H buttons in a scene. From the outside the KPL looks like your 5-button. A+B buttons are Non-toggle ON (as they are on a normal 5 button) and the G+H buttons are Non-toggle Off. It allows you to control the lighting on all of the buttons - when all buttons are off, the KPL can be completely dark. If you're interested, we could start another thread on the configuration (may already be something on the forum). Probably better to figure out your current issues 1st.

Hey @CoolToys, sorry to hear that problems are rearing their heads again. I looked over your logs. There's a couple of things that I see, and some that I don't: I do not see a trigger for your program. Neither your "MB Window Keypad.ON" nor your "MB Door Keypad 70 9B C0All On" show as having been switched on. It's possible that you have these set as Non-Toggle On. In this case the Log would only show the 1st on event. Please confirm. Your HallwayOutletCurio is showing a "ON" status and is immediately followed by an "ON". This can sometimes happen when a device like a lamplinc is connected to a non-resistive load. The load can provide and electrical kick-back and fool the lamplinc into thinking it has turned back on. Please confirm the device type and model. Turn off the local load sensing if equipped. You have many devices and scenes in your program. 11 devices and being addressed individually and you are using 17 additional scenes. Why not put all of them in a single "bedtime" scene and issue a single "fast off" command? This would eliminate a ton of communication and program waits. Are you experiencing issues communicating to some of the devices? Beyond the above, I would guess that you have another program that is being activated and turning (some) things back on. Look at the program summary tab for programs that have run in the same timeframe. You can sort by "Last run time".

Answers to your questions above. Use the scene in your programs and from the Admin console.

@twired, I find it extremely interesting that your program is ONLY executing the Else section. If the IOLinc status NEVER becomes true, I'm not sure how you could be activating the Else multiple times (there is no state change). Are you basing your statement that the IF section didn't run on the Lack of and Email? I'm asking because this might be interrupted by a quick change in the IOLinc status. A quick Off - On - Off might exit the If section and run the Else. Do you have another program that calls the Else section of the alarm program? Are you running a periodic query program on the IOLinc? Please do perform the changes as suggested by @paulbates and @Techman. If things still aren't happy, we can take a look further.

@Blackbird, I'm thinking this is coincidence. Home Assistant shouldn't do anything to prevent an ISY scene from activating. If you have Automations (programs) set up on Home Assistant, It could turn off devices using the ISY. The ISY should have knowledge of this - it will show up in your event viewer and Logfile. Please check - Run a device link table compare on your KPL. If there are corrupted/missing links, do a restore. Manually activate the scene from the ISY console. Chances are you are having communication problems and not talking to the KPL If the scene doesn't work, try device direct communication from the Admin console (turn the KPL itself on, not the scene). Let us know how things go. If you absolutely can't talk to the KPL, we'll need to get a bit more invasive. Edit: Are you running Home Assistant on the EISY or another device?

Could you post the program that is running the Else clause (IOLinc activated)?

Sorry - looks like I was the one that was confused. Back to your normally scheduled broadcast...

As @oberkc indicated, the ISY log should show X10 activity received by the PLM. In the view below, activity on house code F is valid. Activity on House Code P is not. Unfortunately, X10 signals are likely very low level. If you don't have X10 repeaters in your system, it's unlikely the PLM will receive ALL the X10. Time to pull out the age old X10 troubleshooting tools: Look for battery devices that may be failing (motions, etc). These can spontaneously change house/unit code as the battery drains. Trace circuits (flip breakers, unplug devices)- if the problem is isolated to a circuit or phase try filtering or improving signal levels. As @paulbates indicated, try changing house codes. Certain house codes are more prone to noise issues. J is one of them because of the bit coding. I have always gotten spurious signals on P. House code A is typically where battery devices communicate when they start failing. It is possible that an Insteon device has a X10 house code programmed. The only way to eliminate the House code is to factory reset/ restore the device. You may need to head over to the old X10 forum for additional tools.

It sounds like we are confusing terms here. Tazman is referring to the Secondary Buttons on a KeyPadLinc. It looks as if you are referring to a 2477 Dimmer Switch that is a SECONDARY (responding) device in a scene. Please respond whether this is correct. If this is the case, the load should react when the 2477D Leds react. If it does not, you have a defective load, problems with your wiring, or a blown Triac (output stage) in you 2477D.

@eric_allman, glad you made it to the end of the long and winding road... As far as the IRLinc is concerned, I believe these are powerline only devices. You could try plugging it in next to the PLM to see if it will link.

@arzoo, glad to hear you have things under control again. As far as why, not a clue. As stated earlier, there was no evidence that your IOLinc was communicating during the events. But then, disconnecting the IOLinc eliminated the problem. Doesn't make sense on the surface, and probably isn't worth chasing unless it reoccurs. One additional observation - It sounds like you are running the IOLinc in Voltage sense mode (On > ~1.4V/ Off < 1 V). That's a bit more challenging than a simple contact closure. It's possible that you had a bad connection that finally reared it's ugly head making your system more sensitive to line level disturbances (communication on the powerline). People sometimes see this type of powerline activity when their LED bulbs flicker. The communication on the powerline is affecting the dimmer drive level to the Triac powerline the LED. It's possible that the IOLinc could be similarly affected on the measurement side (I don't have a schematic to make an assessment). This is why hitting a solid gnd level ( 0V = off) on the IOLinc is important. It's effectively increasing the signal to noise ratio. You may want to measure the "Sense to Gnd" voltage level (both on and off) to see what your alarm system is putting out. From memory, the IOLinc will indicate OFF with a 0.9V input (< 1V). Unfortunately, that provides you very little noise margin in the event of a disturbance.

That's an excellent response. You must have a 5 V activated siren module. Please check to make sure that you are NOT using "trigger reversed" on the IOLinc. It produces erroneous indications when queried.

@arzoo, when you say you're connecting the "alarm siren" to the GND and S terminals, what type of output is this? The IOLinc terminals normally accept a contact closure type of input. The Sense terminal is capable of 5 Vdc max. It can't handle a speaker level AC input. Beyond that, most siren "Modules" that I've seen require a 12V input. There are probably ways that you can connect the IOLinc to your panel, but I don't think a Siren output is one of them.

When I looked at this previously, I attributed the on/off behavior to activity on the power line + the load. When you tested with the "new" extension cord, was the LL plugged into the same outlet? If so, that's rather damning. You might want to give it a little more time to increase your confidence. After that - extension cords are cheap. I would cut the ends off (make sure no one else tries to use it) and dispose of it.

Please perform your due diligence before jumping to a new technology/protocol. You're accustomed to Insteon and it's very different from the "other" protocols. Rather than jade you with my preconceptions, I would encourage you to read up on WIFI, Zigbee and Zwave protocols. They all have their + and - and I use all of them. Insteon is extremely effective for lighting (scenes and the like). Very few other protocols can match it. Other protocols are far better at security and power (battery devices).

Thanks, I'll set up a test rig. I'm curious if this was the same thing that I had blamed on low wattage dimmable LED's. I'm assuming "lamp cord" the the AWG. My LampLinc is pretty old - V.3B/ R1.5 Date Code 0211.

@paulbates, makes a good point. My judgement is likely clouded by my disdain for the 2450. I could not find any entries in your logs for the address 70.7B.5B. That's a very "high" address for an IOLinc. It's newer than anything in my system. Are you sure it's correct? If it is correct, your logs do not show any IOLinc activity.

@arzoo, can you describe how you're using the IOLinc sensor input. You mention that it's "Hardwired" to the alarm siren?? Voltage sensing? Contact closure? Do you have any Zwave or Zigbee devices in your system?

If I understand you correctly, you are saying that you have an EMPTY extension cord plugged into the LampLinc. It's possible that the extension cord presents enough of a capacitive load that it's confusing the LL. I'd call that a bad extension cord. Can you describe the cord (2 wire, 3 wire, length, AWG)? I'd like to run some tests - never tried this previously.

I do hope you've found your culprit. It's difficult for me to describe my dislike for the IOLinc without using language that would get me banned from this forum. I would very much advise that you find a different sensor to use as the trigger. Describe the application and there are many members that can help. As you indicated, I don't see any reason that having the sensor in the same outlet as the PLM would be a problem. The real problem is using the IOLinc ANYWHERE in your system. They are simply failure prone devices. Could you provide the address for the IOLinc - I'd like to check your event viewer posts for activity.

@eric_allman, Sorry, I know it's painful rebuilding scenes (particularly for a KPL). I wish there was a neat trick to help with all the On-Levels and Ramp Rates - I don't have one. I've spent a good amount of time looking through the rest interface and the ISY SD card. I honestly don't know where the ISY stores this information for scenes. You can interrogate the ISY for Scene Members and Controllers using the rest interface : /your.ISY.IP Ad/rest/nodes/scenes This will open a XML file in your browser (you'll probably be prompted to login first). You can open the file in Excel as a table. That allows you to sort device addresses to determine which scenes they are members of. It does not include brightness or Ramp Rate information. The following is the scene listing for one of my 8 button KPL's @15.C6.C7. Notes: All devices are members of the My Lighting Scene Type 16 is a controller, 32 is a responder. The Device Group is what is stored in the PLM/Device Link tables. This listing shows decimal, the Link Tables and Event Viewer will show Hexadecimal. That's about all I can offer. Be patient, methodical, and take your time. I find that drowning bait and killing brain cells reduces the tension. But then that's my answer to many problems.

@arzoo, I agree with @Geddy that it's probably time to submit a ticket. We've not been able to identify an Insteon trigger for your events. Possible that things are being initiated by a node server/rest event. Best to get the experts involved.

@arzoo, you appear to have a program firing every 30 seconds and generating the timestamps. That's a pretty tight loop. You should be able to locate the program by looking at the program summary tab. You also have a REST command executing every 5 minutes. I'm not sure if this is program related or an external communication.