LeeG

Suggest running Tools | Diagnostics | Event Viewer at LEVEL 3 during tests. The event trace will show any external Insteon activity that could affect Keypad button.

If the Error LED was the only LED blinking it usually means a problem with the SD card. Power down the ISY and disconnect the cables. Reseat the SD card. If the symptom comes back Appendix C in the ISY User Guide has the procedure for replacing the SD card.

Great news.

Thanks for the update. A delay in either the Sensor or Relay response is easy to see in the event trace regardless of On or Off state. Appreciate the effort in trying to recreate. Post back if anything develops.

The Relay is already reporting as On which should not be possible in any of the Momentary modes. It looks like the relay Options the ISY is tracking for the I/O Linc do not match the device. Comm looks good in the trace. It would take some intermittent interference to be causing a comm issue. \\Wed 03/18/2015 10:24:07 PM : [iNST-TX-I1 ] 02 62 2F 79 00 0F 19 00 Wed 03/18/2015 10:24:07 PM : [iNST-ACK ] 02 62 2F.79.00 0F 19 00 06 LTSREQ (LIGHT) Wed 03/18/2015 10:24:08 PM : [iNST-SRX ] 02 50 2F.79.00 25.28.20 2B 00 FF (FF) - Relay On Wed 03/18/2015 10:24:08 PM : [std-Direct Ack] 2F.79.00-->ISY/PLM Group=0, Max Hops=3, Hops Left=2 Cannot tell from this trace what happened at 3AM Query. Since the Insteon products are not Security oriented an EOL resister cannot be used to protect a NO circuit. At this point I would not change any Option to try to compensate without positive trace information, except turn the Relay Off. If failing often then have the event trace running at 3AM. Otherwise wait and see what happens.

Insteon does not allow multiple Scenes to run at the same time. Insteon will cancel one of them. Discussed briefly in the insteondetails.pdf doc on insteon.com. Insteon has always worked that way. Adding a small adjustment to Sunrise or adding a Wait of a few seconds to one of the Programs should resolve if due to running multiple Scenes.

Also the initial post says the lights turn On at 5AM but the posted Program shows 4:45AM. Is there any chance there is another Program involved. Also you ask if the firmware was updated. There is no automatic update so if it was updated you would have done it manually.

Run Tools | Diagnostics | Event Viewer at LEVEL 3. This will show (or not show) the Scene Off at Sunrise. There is an Icon on the bottom of the event viewer popup, next to last on right side that Copy to Clipboard. Paste the event trace at Sunrise. It is likely the loads but it could also be something else that is running at Sunrise. You are on an older ISY image. Suggest moving to 4.2.28

That is the "All On" problem but seen with an Off command which has been reported but infrequent even compared to the infrequent All On.. Has to do with the commands issued by the Program overlapping the commands from the Keypad. NOTE: pressing any button/paddle actually runs an Insteon Scene. Devices cannot issue direct On/Off commands.

DavidJ101 Works for most of the mag switches Smarthome sells as they are primarily Security devices. If the mag switch is purchased from a larger online firm that sells both Security and Industry devices it requires some analysis of how the device actually functions. I was burned many years ago by that mistake.

The Motion Sensor does not have separate On/Off nodes. The Motion Sensor would be set to Sensing mode: checked and On Only mode. That way the Sensor node can turn a Scene On for quick responder reaction and use a Program(s) to turn the responder Off.

That is not an issue. Lighting control is not used heavily (not like turning On/Off devices every few seconds throughout the day). The Hidden Door Sensor can be used to create that arrangement. The On node (Open) can be linked to the light control switch for near instantaneous On with the Off node (Closed) controlling an ISY Program to determine when to turn the light control switch Off (door closure or time).

Yes, if the Wait has not completed. The If Condition can be True or False, if the Wait (or Repeat) has not completed the Wait (or Repeat) is terminated and the clause (Then or Else based on If evaluation) restarted.

Kush Finally had some time to find examples that show the difference in nomenclature. This link describes a NO switch (one that closes when magnet near switch) https://www.inventables.com/technologies/magnetic-switch-normally-open This link describes the same operation, circuit closed when magnet near, as NC switch. http://www.smarthome.com/seco-larm-sm-205q-w-enforcer-nc-magnetic-contacts-with-15-inch-leads-white.html The difference being one is for security, the other for industry. Same switch operation, circuit closed when magnet near, but one is described as NO and one is described as NC. Just one of those things that can add to the confusion about how a switch operates.

Stu Camera would be a better solution. Plus it adds an additional layer of security.

Same with Dual Band Keypads. Not everyone needs Dual Band but the non Dual Band are no longer sold. Smarthome normally stops selling older models.

Kush We have a cross in terminology. The magnetic switches are closed when the doors are closed. They would not work in a series loop otherwise. Whether you call them NO or NC (security applications refer to them one way, industry applications refer to them the other), the I/O Linc Sensor is On with the doors closed. The Relay is Off when the Query is done so the response is the opposite to the door closed state reported by the I/O Linc Sensor. Therefore it looks like a door has moved when it has not. Sorry for the confusion regarding NC versus NO. The solution is to correct the comm issue (assumption without seeing event trace). The alternative is to use magnetic switches that operate opposite to what you are using now. That way the Relay and Sensor are in the same state so if the Relay response is slow to arrive it will be the same as closed doors so no false message.

The switch is weatherproof but my preference was to mount the switch on the door header at the top of the door opening. Then mount the magnet on the top edge of the garage door itself. I also have a magnetic switch that is activated when the door is fully open. It is activated by the same magnet that controls the switch indicating the door is fully closed. I chose to two switches because the position of the detached garage does not allow viewing of the door from the house.

The magnetic switch is mounted at the top of the door, not on the floor. The magnet is mounted on the top edge of the door. As Stu indicated the gap can be wide with the combined N/O N/C magnetic switch.

Kush The I/O Linc Sensor and Relay are queried when either Sensor or Relay is queried. The Relay query response can be delayed due to comm issue such that the relay response comes back after the Sensor is queried but before the Sensor response comes back. It makes the Relay query response look like the Sensor response. Since the Relay is Off, it looks like the Sensor has changed state to indicate an open door (since NC magnetic switches are used). Suggest running Tools | Diagnostics | Event Viewer at LEVEL 3. Issue a Query for the I/O Linc Sensor a few times and post the event trace. There is an Icon on the bottom of the Event Viewer display, next to last on right, that Copy to Clipboard. Paste the event trace to a forum post. The problem may be time dependent. If the comm issue is difficult to solve, using NO magnetic switches wired in parallel will bypass the comm issue because the Relay Off and Sensor Off have the same response.

Of course anything is possible but it seems unlikely a PLM less than a month old would fail with capacitor issues. What does Tools | Diagnostics | PLM Info/Status show? What does Help | About show for Firmware Level ? UI Level?

The button LED turns On under control of the Keypad. Could be a non toggle mode setting for the D button or a Program is turning the D button Off which would not affect the Scene the button turned On. Note that LED Backlight settings apply to all KPL buttons, not just the button on the BackLight command.

The Restore Device should have restored the links. Since it did not I assume the current ISY information does not indicate more link records are needed. This is a SwitchLinc with a paddle, not a Responder only device? The MH Porch Lights (assuming it has a paddle) must have a Controller link (starts with E2) that causes the device to notify the ISY of paddle operations. Also there is no Responder link for an ISY Scene (starts with A2 and has PLM address as Controller). We've gone through all the steps that should have restored the device link database and PLM for that device. Between the low record count in the PLM and the missing link records in the device, something is not current with the current ISY information. I suggest opening a Ticket with UDI as there is nothing more I know to do.

The next step is to issue a Restore Device for the SwitchLinc. If that brings back the required links there is a File | Restore Devices that will restore all the devices. I'm concerned the ISY Backup file used was not current. With the ISY powered down when the old PLM was installed, yet a Restore Modem (PLM) did not bring back the required links, the current ISY information does not appear current. A Delete Modem (PLM) would create this problem but it was indicated that the command was not used.

The device is missing some required link records associated with the PLM. When was the old PLM installed? Was the ISY powered down when old PLM was connected? Was the Wiki procedure for replacing PLM followed? Link to Wiki instructions for replacing PLM. http://wiki.universal-devices.com/index.php?title=ISY-99i/ISY-26_INSTEON:File_Menu#Replace_Modem_.28PLM.29