Comm to RF devices bad device link tables after power outage

[junkycosmos]

Hi UDI Gang !

 

Need some advice here. Been running insteon / ISY for a few years now with fairly large setup. All has been nicely solid, well nearly GREAT for a few months now. Have made nearly no changes since upgrading to 3.1.17. However I'm stuck solving this latest riddle.

 

As nature would have it here last we lost power for six hours and had a few blinks when it came back on. Now I do have ISY's power adapter on UPS which ran out before power was back on and the PLM is not on UPS. This does mean the ISY was on for a few hours without the PLM however this was 6 days ago and all other devices appear to be fully functional with ISY.

 

All appears to have recovered with a few exceptions:

 

3 of our 5 triggerLinc have stopped communicating with ISY

2 of our 9 motion sensors have stopped communicating with ISY

 

Here is the strangeness:

With all 5 of these units the RF indicators still blink when triggered or the set button is pushed. I can also see blinks of the LEDs on both the access points and the PLM when RF is sent from these devices.

 

Event viewer shows no activity for any these 5 devices.

 

Diagnostics --> Device Link Table

 

from one of the triggerlincs (13.E6.13/driveway)

Device Links Table shows:

0FF8 : 00 00 00.00.00 00 00 00

 

Compare shows

0FF8 : E2 01 0F.D9.29 FF 1F 01 [Record mismatch]

0FF0 : E2 02 0F.D9.29 FF 1F 02 [Missing this record]

 

****

 

From another triggerlinc (13.E4.66/front)

Device Links Table shows:

0FF8 : E2 01 12.33.2A 00 00 00

0FF0 : 00 00 00.00.00 00 00 00

 

Compare shows

0FF8 : E2 01 0F.D9.29 FF 1F 01 [Record mismatch]

0FF0 : E2 02 0F.D9.29 FF 1F 02 [Record mismatch]

 

****

 

From one of the motion sensors (11.9F.F4 office)

0FF0 : 00 01 0F.D9.29 FF 1F 01

 

Compare shows

0FF8 : E2 01 0F.D9.29 FF 1F 01 [Record mismatch]

0FF0 : E2 02 0F.D9.29 FF 1F 02 [Missing this record]

0FE8: E2 03 0F.D9.29 FF 1F 02 [Missing this record]

 

****

 

From other motion sensor (13.FA.29/kitchen)

0FF8 : A2 01 11.9F.F4 00 00 00

0FF0 : 00 00 00.00.00 00 00 00

 

Compare shows

0FF8 : E2 01 0F.D9.29 FF 1F 01 [Record mismatch]

0FF0 : E2 02 0F.D9.29 FF 1F 02 [Record mismatch]

0FE8: E2 03 0F.D9.29 FF 1F 02 [Missing this record]

 

****

Other strange item:

Tools--> Diag--> Show PLM Links Table:

gives an empty window, clicked the start button and it switches to gray for 1 sec and then back to clickable. Clicking count gives result of zero.

****

 

I have not manually checked every other Insteon device in the house (~200) but everything else does appear fine, especially the other triggerlincs and motion sensors.

 

Appreciate any help here in both solving the riddle of how this happened and fixing. So far I have not taken any action other than the diags above.

 

J

[Michel Kohanim]

Hi junkycosmos, long time no see!

 

I think the problem could be either of two things:

1. If you have a dual band PLM, then perhaps the PLM has gone bad (no more RF)

2. If you have dual band devices and/or access points, then those have gone bad. Do you have any dual band devices that no longer work?

 

My recommendation is:

1. Try moving one of the Motion Sensors or TriggerLincs (whichever one is easier) next to the PLM. If it works, then the problem is #2 above

2. If it does not work, then the problem is the PLM

 

With kind regards,

Michel

[junkycosmos]

Hi Michel,

 

Thanks much for the reply. With the nice summer been working more of the outside projects these days.

 

I too had same initial thought (#1) about RF issue since I've seen where APs etc can be a tad sensitive to power outages etc however all other RF devices were working fine. I checked the few dual band devices and those too appear to be functional. So going with your PLM thought (and noting how the ISY would not pull the PLM device table) I did a full power cycle this morning.

1. unplugged additional power adapter from ISY

2. unplugged PLM from wall socket

3. counted to 30 then plugged back in PLM; PLM and ISY powered up

4. plugged back in ISY power adapter

 

Good news! These 5 RF devices are now working nearly fine. Exception is 1 motion sensor that looks to have a phantom link (office).

 

PLM links table now behaving as expected; took expected time to load and then count worked.

 

PLM is the older non dual band v.85 with a piggybacked AP; hence it powers the ISY too.

 

sequence of prior events:

04:20:20 power went out (PLM outage, ISY still on via adapter powered by UPS)

06:11:52 UPS batt ran out (power to ISY downed)

07:37:27 Power restored; PLM powered up

07:37:50 UPS completed restart and powered up adapter to ISY

After full ISY start up all but 5 of my devices functioned fine

 

My expectation is that the ISY would have powered up at 07:37:27 when the PLM received power even though UPS power was not yet up (so the ISY power adapter was not live yet).

 

This morning after power cycle restored functionality I redid the device link table views (strangeness here)

both triggerlincs and 1 motion sensor show same results as prior (missing and miss match records but they are working fine)

 

The office motion sensor now has extra rows and blinks as if looking for a phantom link but ISY does show status updates for the sensor

0FF8 : E2 01 13.FA.29 00 00 00

0FF0 : E2 02 0F.D9.29 FF 1F 02

0FE8 : E2 03 0F.D9.29 FF 1F 03

0FE0 : 00 00 00.00.00 00 00 00

 

 

-->> What I cannot figure is how 98% of my devices were working just fine (including many other RF devices) but these 5 RF devices stopped working but have started again from the simple power cycle.

 

Suppose I can delete, reset, and re-add the office sensor to clear out the phantom link.

 

Also wondering if I should still look to replace my PLM with the spare dual band one (v1.5) which I have on the shelf although I've been saving for when I get to upgrading to 994iZ.

 

Do let me know if you have any other suggestions or items I should check here please.

 

Thanks again

J

[LeeG]

The number of link records is correct at 3 active records. One covers motion, one covers dusk/dawn and one covers low bat. However, it appears that the PLM was changed at some point as the motion entry has an Insteon address of 13.FA.29 which must be the current PLM or motion would not be indicated. The next two links have 0F.D9.29 which may be of an old PLM that is no longer in use or there is more than the ISY that is managing devices.

 

This device should be put into linking mode and a Restore Device done to fix the two links which have the wrong PLM address.

 

0FF8 : E2 01 13.FA.29 00 00 00

0FF0 : E2 02 0F.D9.29 FF 1F 02

0FE8 : E2 03 0F.D9.29 FF 1F 03

0FE0 : 00 00 00.00.00 00 00 00

 

EDIT: actually my conclusions about which address is correct and which is bad may not be correct. The motion sensor is going to send a motion message sequence which starts with a Group Broadcast which is not addressed to any specific device so the ISY PLM would still show motion detected even if 13.FA.29 is not the current ISY PLM. However, all three link records need to have the correct PLM address and the solution is Restore Device.

[junkycosmos]

Lee, thanks much for the info.

 

I follow on the expected 3 links in the motion sensor (motion, dusk, batt).

Yes the PLM is F.D9.29; I believe it is my original. Could be this motion sensor picked up a link when I was trying to read the device table a few times. I will do restore device.

 

The part I am still a bit perplexed about is that all functionality was restored by doing a simple power cycle of PLM and ISY. However my device tables for triggerlincs are still empty, yet devices are now fully functional.

 

Leaves me thinking (guessing); they are currently working via some group broadcast which was being ignored prior to power cycling OR the show device links function is not fully working correctly (v3.1.17).

 

Regardless it is not a huge deal since the "fix" was really simple however I does leave me wondering if I have a skeleton that I could use to eradicate before it comes sneaking back out.

 

Thanks for any info.

J

[LeeG]

My assumption is the PLM was in a bad state after the quick power cycles. Since the link database reappeared after a clean power cycle it had to be there the entire time, the PLM just would not retrieve it for some reason. Insteon devices are much better at handling clean power drops and restores. A very short drop in power causes the voltage to start to fall off but it does not have a chance to go all the way Off before 120v power comes back. That can play havoc with electronics.

 

Reading the motion sensor link database should not have changed the link record. Do you have a device with the Insteon address of 13.FA.29? (address in first link record)

 

Normally a device will not send any message if the link database is completely empty. If the TriggerLinc is changing state in the Admin Console with an empty link database the TriggerLinc must be sending a Group Broadcast (which does not contain any device address) even when there are not active link records. The ISY is looking for a Group Broadcast message so things appear to be working even though the rest of the Group protocol is missing. The TriggerLincs should be given a Restore Device.

 

No telling how long the motion sensor and TriggerLincs have been this way. Being RF and asleep they would not have been aware of the loss of power.