One device fails in a scene with many devices

[jp5150]

Hi all,

 

I have a scene that sometimes fails to turn on either my garage coach lights, or my entrance light. This scene is only triggered by the isy based on sunset time.

 

Below are two scene tests. I was wondering if anyone can see something that could help me troubleshoot this problem:

 

First Scene Test

Mon 01/31/2011 08:05:20 PM : [GRP-RX ] 02 61 2C 13 00 06

 

Mon 01/31/2011 08:05:20 PM : [iNST-SRX ] 02 50 13.FE.7F 13.20.53 61 13 2C LTOFFRR(2C)

 

Mon 01/31/2011 08:05:20 PM : [standard-Cleanup Ack][13.FE.7F-->ISY/PLM Group=0] Max Hops=1, Hops Left=0

 

Mon 01/31/2011 08:05:20 PM : [iNST-SRX ] 02 50 13.FE.7F 13.20.53 61 13 2C LTOFFRR(2C)

 

Mon 01/31/2011 08:05:20 PM : [standard-Cleanup Ack][13.FE.7F-->ISY/PLM Group=0] Max Hops=1, Hops Left=0

 

Mon 01/31/2011 08:05:20 PM : [iNST-SRX ] 02 50 13.FE.7F 13.20.53 61 13 2C LTOFFRR(2C): Process Message: failed

 

Mon 01/31/2011 08:05:20 PM : [standard-Cleanup Ack][13.FE.7F-->ISY/PLM Group=0] Max Hops=1, Hops Left=0

 

Mon 01/31/2011 08:05:21 PM : [iNST-SRX ] 02 50 13.FF.5A 13.20.53 61 13 2C LTOFFRR(2C)

 

Mon 01/31/2011 08:05:21 PM : [standard-Cleanup Ack][13.FF.5A-->ISY/PLM Group=0] Max Hops=1, Hops Left=0

 

 

Second Scene Test

 

Mon 01/31/2011 08:05:22 PM : [iNST-SRX ] 02 50 16.30.54 13.20.53 61 13 2C LTOFFRR(2C)

 

Mon 01/31/2011 08:05:22 PM : [standard-Cleanup Ack][16.30.54-->ISY/PLM Group=0] Max Hops=1, Hops Left=0

 

Mon 01/31/2011 08:05:22 PM : [CLEAN-UP-RPT] 02 58 06

 

Mon 01/31/2011 08:05:22 PM : [iNST-SRX ] 02 50 16.E2.E7 13.20.53 61 13 2C LTOFFRR(2C)

 

Mon 01/31/2011 08:05:22 PM : [standard-Cleanup Ack][16.E2.E7-->ISY/PLM Group=0] Max Hops=1, Hops Left=0

 

Mon 01/31/2011 08:05:22 PM : [iNST-SRX ] 02 50 16.E2.E7 13.20.53 61 13 2C LTOFFRR(2C)

 

Mon 01/31/2011 08:05:22 PM : [standard-Cleanup Ack][16.E2.E7-->ISY/PLM Group=0] Max Hops=1, Hops Left=0

 

Mon 01/31/2011 08:05:22 PM : [iNST-SRX ] 02 50 16.E2.E7 13.20.53 61 13 2C LTOFFRR(2C): Process Message: failed

 

Mon 01/31/2011 08:05:22 PM : [standard-Cleanup Ack][16.E2.E7-->ISY/PLM Group=0] Max Hops=1, Hops Left=0

 

----- Exterior Lights Sunset Test Results -----

[succeeded] Terrace Keypad - 4 (16 E2 E7 4)

[succeeded] Terrace Keypad - 5 (16 E2 E7 5)

[succeeded] Rear Lamp Posts (16 30 54 1)

[succeeded] Entrance Light (8 F5 2A 1)

[succeeded] Front/Side Landscape (13 FE 7F 1)

[succeeded] Rear Landscape (13 FF 5A 1)

[succeeded] Side/Bar Landscape (14 1 48 1)

[succeeded] Coach Lights (16 30 34 1)

[succeeded] Terrace Keypad - 8 (16 E2 E7

----- Exterior Lights Sunset Test Results -----

Mon 01/31/2011 08:05:28 PM : [iNST-ACK ] 02 62 00.00.2C CF 13 00 06 LTOFFRR(00)

[Michel Kohanim]

Hello jp5150,

 

Please make sure you do not have any programs that may be activated when you do your scene test as it would interfere with the results.

 

With kind regards,

Michel

[jp5150]

I disabled all programs before running the test.

[oberkc]

So you have one attempt at a scene test that failed, and one that passed. I take this as less-than-ideal communication. One thing I like to do during scene tests is watch the event viewer. I have found that when communication is good, events occur at quick, regular intervals. If one sees delays, this is a sign of potential problems.

 

Otherwise, it looks as if the scene test confirms your experience. You have one device that sometimes fails to responds. You have a scene test that sometimes fails. If it were me, I would begin the trial-and-error process of fault isolation.

[ELA]

I am slowly getting familiar with the communication protocol (with a long way to go).

When I looked at the log I saw a msg. fail each time that did not apprear to have anything to do with the lights that failed the scene test.

I also thought that there was no response from the lights that the op. said did fail the scene. This much seemed to make sense to me.

 

What I do not understand is why the scene test states that it passes [suceeded] on the two lights that are failing??

[oberkc]

There are often intermittent conditions which could cause communication issues, especially if your normal condition is marginal to start with. There are likely devices in your house which run only some time, such as washers, dryers, refrigerators, HVAC, water heaters, etc....

 

Perhaps you have some noisy fluorescent lights which are not always on. Perhaps this is why one of your devices fails some of the time and not all of the time. If you have insteon devices which turn on reliably, but not off, it is possible that the load is causing problems (CFLs, low voltage lighting).

 

I also could not tell whether your troubled device is part of the scene in which you tested. If not, then run a test on a scene that DOES include the troubled device.

 

I understand the scene test to be less than fully 100% conclusive. Rather it is a general representation of the communication between the PLM and devices within the chosen scene. It does not offer any troubleshooting ability, in terms of identifying problem-causing devices on your electrical system.

 

I don't know your system. I assume you have access points or dual band devices or some means of coupling your electrical legs of your house, correct? Is your PLM plugged into an outlet or circuit that includes lots of other electrical gadgets, such as the computer and related electronics (UPS?).

[jp5150]

oberkc, the devices that usually fail are part of the scene in question.

 

ELA said:

When I looked at the log I saw a msg. fail each time that did not apprear to have anything to do with the lights that failed the scene test.

This is correct and confuses me. It appears that the devices that the scene test is having trouble with are not the ones that are actually physically failing to respond.

 

What I do not understand is why the scene test states that it passes [suceeded] on the two lights that are failing??

This also confuses me.

 

I don't know your system. I assume you have access points or dual band devices or some means of coupling your electrical legs of your house, correct?

Correct. I have at least 8 dual-band switchlincs throughout the house as well as 3 dual-band lamplincs.

 

Is your PLM plugged into an outlet or circuit that includes lots of other electrical gadgets, such as the computer and related electronics (UPS?)

This I don't know. I haven't checked what devices are on this particular circuit. I do however have filterlincs installed on all my UPS devices in the house.

[LeeG]

jp5150

 

The ISY Scene test does not duplicate the commands issued for a Set Scene On/Off issued by an ISY Program. The Scene test is good at detecting conflicts with Program activity, finding gross powerline problems that cannot be overcome by the built in command retry associated with the Insteon Group protocol, and indicating when more than a single Hop is required to get to a particular device (more than 1 Hop does not in and of itself indicate a problem that requires a solution).

 

From your original post that started this topic, two Scene tests are indicated as having been issued but only one summary is shown. It can take several seconds for the ISY to complete the Scene test and produce the final summary showing success/fail. It looks like the first Scene test was not allowed to complete before the second was initiated.

 

Finally, the Program initiated Set Scene On/Off may fail to control a device reliably even though a Scene test summary at the end shows success for all devices. Different command sequences can have different results. A device intermittently failing to respond to a Program Set Scene On/Off is likely noise or signal attenuation on that circuit. Unplugging devices (turning them Off is not enough), observing what devices are running when the failures occur (for devices that cannot be unplugged), adding a Dual Band device close to the device(s) that are intermittent, are all techniques that may be required to find the source(s) of the device response failures.

 

Lee

[ELA]

Lee,

Would it be worthwhile to then run individual on/off commands direct from the ISY to only the devices in question and record using the Event Viewer (set at level 3)?

[LeeG]

ELA

 

The Insteon Direct commands the Admin Console and Programs use for direct device control are automatically retried by the PLM should a device fail to ACK the command. If a Direct command fails to cause the device to respond there is definitely a problem as the PLM has tried multiple times. The ACK from the device will show the Hop count on the command that caused the device to respond. That could show how many devices have to repeat the Direct message for it to reach the device. A Direct command to a device that works reliably compared to a device that is intermittent provides a method for seeing the difference between good communication and bad.

 

The Event Viewer does have to be at Level 3 or Device Communications Events.

 

Lee