IndyMike

@SirParadox , your device B isn't capable of "pushing back" a link to device A. The only way for device B to write links in device A would be if you Manually linked the device. If that were to happen, the ISY would have no record of the links (it would show up as a mis-match) and you could easilly restore the device. If you believe the link tables in both devices are incorrect, you will need to delete, factory reset, and re-add both devices.

@SirParadox, the device 3D.96.1C appears to be a PLM. It is listed as a controller for group 0 of your family room overhead device. It is possible that this is an OLD PLM (check your current PLM address to verify). In any case, the ISY believes the links to be correct. Restoring this device from the ISY will simply re-write the same links. If you are sure that the links are incorrect, the quickest way to rectify is to proceed as @apostolakisl indicated in his Option #2 above (delete device, factory reset, re-add to the ISY).

@Tmc, I'm in agreement with @paulbates. Since it appears that you have having problems communicating with numerous devices, it's likely you have a noise source or signal absorber near the PLM location. Look for PC's, chargers, printers, UPS's that are common culprits, and remember that electrical devices do go bad. Another option could be your phase coupling. It's interesting that your system was working with the 2412S which is not dual band. The 2413S is dual band and would presumably perform better in a noise environment. How were your phases coupled before? Note that the 4 tap test (beacon test) that Paul suggested is an excellent way to detect phase issues.

@oberkc, I'm thinking you are using the PG3 HUE plugin to communicate with a HUE HUB - correct? It sounds as if @greensha is attempting to connect directly to the HUE bulbs by enrolling them as Zigbee devices.

@pjjameso, your timer program is marked "Run at startup". Can you see it running in the "program summary tab" (you've probably already looked). If not, try a manual start (Run Then) to see if things run and continue. If everything looks good with the timer program, try re-writing and saving the "Airthings AQ Normal" program. I have seen instances where the "text" of a program is correct, but the saved XML is not. A simple Edit/Save will not correct this.

@AD8BC, as you probably already realize, your PLM is only linked to 4 devices. Something is seriously wrong with the restore. If you have a recent backup, try restoring the EISY then follow with a PLM replace procedure (EISY needs a reboot to recognize the New PLM).

@Tmc, sorry for the delay in responding - I was on the road. Answers: I "save" the PLM table as an XML file and then Import to Excel as an XML table. it allows be to sort columns and select individual device addresses for inspection. Unfortunately the values in the XML file are in decimal. You have to convert to Hex (+dec2hex) for them to make sense. Curious - my file save with an .XML extension. Most browsers can open it and it appears as you have shown above. In order to format the data, you can open the file with Excel or use a XML viewer. The link 325 that is shown is a "responder" link. As I indicated earlier, I believe the ISY leaves these. As @Techman indicated, the PLM link read isn't perfect and can be interrupted. I run the process 2 - 3 times in a row to make sure things are consistent.

@walkman9999, Had the chance to look at the files you up loaded. Your _Northgate program has a lot going on. Many Device Direct Fast ON commands in succession, followed by a scene command, and then more Direct commands. The rapid succession of these commands are bombarding the PLM and may be confusing it. In my experience, the PLM is rather good at handling stacked commands (back to back). But then I use the older ISY994. The EISY is far faster, and may be capable of overwhelming the PLM. I've posted a section of your event viewer below the program listing. Your fist Insteon command is a Fast ON to device 29.18.D7 Highlighted yellow). The normal sequence would be: Wed 02/12/2025 16:46:26 : [INST-TX-I1 ] 02 62 29 18 D7 0F 12 00 (Fast on Command to PLM) Wed 02/12/2025 16:46:26 : [INST-ACK ] 02 62 29.18.D7 0F 12 00 06 LTON-F (00) (Acknowledge from the PLM) Wed 02/12/2025 16:46:26 : [INST-SRX ] 02 50 29.18.D7 71.1B.41 2B 12 00 LTON-F (00) (Acknowledge from the device) I've highlighted all 3 in the event viewer listing below. These would normally be in sequence. As you can see, you have many other commands in between, as well as the curious "Link for responder XX.XX.XX not found". I believe this is the PLM informing the ISY that it doesn't contain the link for these devices. The thing is, I looked at your PLM listing and it most certainly does contain the links for these devices. As I said, confused. In general, the communication response activity I see in the event viewer log looks rather good. Two or three hops remaining fall all devices (Including older I1 devices). You may have some communication issues, but they were not apparent in the log. That's good. You may have "isolated" issues on some circuits, rather than wide spread issues due to a problem PLM location. Suggestions: Insert some short waits between some of your Insteon commands. If you suspect a scene issue, insert a longer wait before the scene command. Your programs were attached in an XLM format. Very few forum members can interpret it (and I have trouble). If you would please right click on the program and select "copy to clipboard", you'll be able to paste a "text version of your program. Many far more experienced people than I will be able to assist with suggestions. It looks like the Eisy is still using the deprecated Peek/Poke method of restoring your older devices. Did you try switching the Eisy to the "Device Reported" mode under "Advanced Options"? XLM Program Listing <?xml version="1.0" ?> <triggers> <d2d> <trigger> <id>50</id> <name>_North Gate</name> <parent>53</parent> <if> <and /> <status id="ST" node="55 18 EA 1" op="IS"> <val uom="51" prec="0">100</val> </status> <and /> <status id="ST" node="1A B3 C1 1" op="IS"> <val uom="51" prec="0">0</val> </status> <and /> <schedule> <from> <sunset>-1800</sunset> </from> <to> <sunrise>1800</sunrise> <day>1</day> </to> </schedule> </if> <then> <var id="3" type="1"> <op>EQ</op> <val prec="0">1</val> </var> <cmd id="DFON" node="29 18 D7 1"></cmd> <cmd id="DFON" node="49 B2 50 1"></cmd> <cmd id="DFON" node="29 18 D7 1"></cmd> <cmd id="DFON" node="1A 61 8D 1"></cmd> <cmd id="DFON" node="1A B3 C1 1"></cmd> <cmd id="DFON" node="29 18 D7 1"></cmd> <cmd id="DFON" node="29 1 8A 1"></cmd> <cmd id="DFON" node="2A B 38 1"></cmd> <cmd id="DFON" node="1A B7 F 1"></cmd> <notify content="1">1</notify> <cmd id="DON" node="11711"> <p id=""> <val uom="51" prec="0">100</val> </p> </cmd> <wait> <seconds>30</seconds> </wait> Event Viewer Log (Partial) PLM Table Sorted for Group 19 Scene Members

@walkman9999, unfortunately I am traveling through Friday. Can't help with the link tables at the moment. I did look at the restore log. You encountered an error writing to a device 1b.b0.cc. This looks like an older device that may be powerline only. As @oberkc indicated, this device would be a prime candidate for a restore. I'll try to have a look at the rest when I return.

As usual, I could not leave this alone... Anyone who has read my posts knows that they are rarely short, and never very interesting. Read on at your own peril... I ran another test deleting a KPL from my PLM. I compared the PLM link tables before and after the delete. Observations: The ISY deleted the control link (E2) for group 0 at address 14H from the table. This makes sense - leaving the link would allow the PLM to try to interrogate the device even though it didn't exist. Think PLM timeouts, retries, and the whole mess. The ISY DID NOT delete the responder links (A2) for groups 1 though 6 (main button and secondaries). I suppose these are viewed as harmless links. The can't do any damage if the device isn't there to communicate with the PLM. With the "deletion" of the Link @14H, all of the following link addresses shifted up by 1. The PLM now shows 1 fewer total links (497 vs 498). The above makes sense to me except for Item 3. There is no way the ISY had time to delete the link @14H and REWRITE ALL OF THE FOLLOWING links. Anyone who has restored a PLM knows it takes a significant period of time. This operation took seconds. This leads me to the conclusion that the ISY DID mark the record @14H as "available" (22). During the subsequent PLM link table read, the record @14H was IGNORED since it was marked as "available". The PLM Link read operation is very different from the device link read operation. The device read operation is handled by the ISY through read commands for each link location. The ISY displays all Link table information including "available" records. The PLM Link Read operation is handled by the Insteon "Get Next ALL-Link Record". This command apparently skips the records marked as "available". I now yield the soap box... IM

@Tmc, my apologies - I was wrong. I just deleted a device from my system and the PLM responder links are still are still there. I'm scratching my head at the moment... Event Viewer: Mon 02/10/2025 06:22:32 PM : [58 C9 B0 1 ] Start : Removing device from ISY Mon 02/10/2025 06:22:32 PM : [58 C9 B0 1 ] Finish : Removing from ISY was Successful PLM Table:

@Tmc, you are correct that there is a record with your device address still in the PLM. If you look at the flag column you should see a "22". This is the sign that the ISY has marked the record as "Unused". In order to completely remove the information, all of the links following the record would need to be removed. Simply marking the record Unused gets the job done quickly. If you find that you have many "unused" records you can perform a PLM restore to rewrite the entire PLM. The ISY will remove the unused records. The same technique applies to devices. The "22" is used to mark records unused, and a Device Restore will rewrite and remove the unused records.

@walkman9999, your Switchlinc Dimmer appears to be a 2476D that is powerline only. Communication with the device isn't great and that can be a problem when trying to restore a device. It might be possible to install a dual band LampLinc or other RF device electrically near the Dimmer to help with communications. It's also possible that the noise or signal absorption is temporary due to a problem device in the vicinity. Try inspecting for typical problem devices (PC's, chargers, UPS's, etc) in the vicinity. Another thing that is not in your favor is that the ISY is using and old I1 programming mode with this device. The Event Viewer communication below is from your Restore sequence. The Red section is where the ISY attempted a I1 write sequence that timed out after 3 tries. The following green section is a successful write that was performed using I2 (much faster). You could try to "force" the ISY to use I2 communication by going to Link Management/Advanced Options and selecting "Device Reported". The I2 protocol allows the ISY to transfer data in far fewer command sequences than the I1 mode. The communication is still susceptible to noise/absorption issues, but there are far fewer opportunities for failure. Sun 02/09/2025 14:11:28 : [ 1B B0 CC 1] Preparing Device 'Shed Bathroom' for Restore Sun 02/09/2025 14:11:28 : [ 1B B0 CC 1] Device 'Shed Bathroom' ready for Full Restore Sun 02/09/2025 14:11:28 : [All ] Writing 28 bytes to devices Sun 02/09/2025 14:11:28 : [INST-TX-I1 ] 02 62 1B B0 CC 0F 0D 00 Sun 02/09/2025 14:11:28 : [INST-ACK ] 02 62 1B.B0.CC 0F 0D 00 06 (00) Sun 02/09/2025 14:11:30 : [INST-SRX ] 02 50 1B.B0.CC 71.1B.41 27 0D 01 (01) Sun 02/09/2025 14:11:30 : [Std-Direct Ack] 1B.B0.CC-->ISY/PLM Group=0, Max Hops=3, Hops Left=1 Sun 02/09/2025 14:11:30 : [1B B0 CC 0 ] Calibrating engine version Sun 02/09/2025 14:11:30 : [1B B0 CC 0 ] May not fully support i2, reverting to i1 Sun 02/09/2025 14:11:30 : [1B B0 CC 1 ] Link 0 : 0FF8 [A200711B41FF1F01] Writing [A200711B41FF1F01] Sun 02/09/2025 14:11:30 : [INST-TX-I1 ] 02 62 1B B0 CC 0F 28 0F Sun 02/09/2025 14:11:30 : [INST-ACK ] 02 62 1B.B0.CC 0F 28 0F 06 SET-MSB(0F) Sun 02/09/2025 14:11:39 : [INST-TX-I1 ] 02 62 1B B0 CC 0F 28 0F Sun 02/09/2025 14:11:39 : [INST-ACK ] 02 62 1B.B0.CC 0F 28 0F 06 SET-MSB(0F) Sun 02/09/2025 14:11:48 : [INST-TX-I1 ] 02 62 1B B0 CC 0F 28 0F Sun 02/09/2025 14:11:48 : [INST-ACK ] 02 62 1B.B0.CC 0F 28 0F 06 SET-MSB(0F) Sun 02/09/2025 14:11:52 : [1B B0 CC 1 ] Link 0 : 0FF8 [A200711B41FF1F01] *Failed Writing [A200711B41FF1F01] Sun 02/09/2025 14:11:52 : [1B B0 CC 1 ] Memory : Write dbAddr=0x0264 [0C] cmd1=0x2E cmd2=0x00 Sun 02/09/2025 14:11:52 : [INST-TX-I2 ] 02 62 1B B0 CC 1F 2E 00 00 03 0C 00 00 00 00 00 00 00 00 00 00 C3 Sun 02/09/2025 14:11:52 : [INST-ACK ] 02 62 1B.B0.CC 1F 2E 00 00 03 0C 00 00 00 00 00 00 00 00 00 00 C3 06 (00) Sun 02/09/2025 14:11:54 : [INST-SRX ] 02 50 1B.B0.CC 71.1B.41 27 2E 00 (00) Sun 02/09/2025 14:11:54 : [Std-Direct Ack] 1B.B0.CC-->ISY/PLM Group=0, Max Hops=3, Hops Left=1 Sun 02/09/2025 14:11:54 : [1B B0 CC 1 ] Memory : Write dbAddr=0x0032 [FF] cmd1=0x2E cmd2=0x00

@BRMeeke, 26 KPL's is a lot. That consumes a lot of links in the PLM (1024 max) and nodes in the ISY (1000 max). Keep an eye on both.

@richtimpa, conventional wisdom dictates that the PLM should be located at the Panel since it is the central point for your electrical system. It is also normally the electrically quiet location. In your case, the Powerwall flips things upside down. Whether the Powerwall is absorbing signals or polluting the powerline, the Panel is where you do NOT Want your PLM. As @bigDvette indicated, moving the PLM far (electrically) from the Panel should reduce both noise and signal absorption. Your electrical wiring has a characteristic Inductance, resistance, and capacitance per foot of wire. Moving away from the panel will allow the wiring to attenuate noise and buffer against absorption. This also happens with Insteon signals, but each device repeats the communication to keep the signal level above the noise floor. Another option would be to put your PLM behind a filterlinc in an attempt to "force" rf communication.