Understanding Scene Test Results

[justin.cool]

My configuration:

 

Approximately 100 Insteon devices:

• Mainly Switchlincs, Lamplincs, Outletlincs, On/Off Adapters, etc etc

• 7 KPL’s

• Triggerlincs (2)

• RemoteLincs (3)

• Controllincs (2)

• ISY-99

• Access Points (5)

• Dualband LampLinc dimmers (4)

• Signal Linc (2)

• Venstar Thermostats with V2 RF interface

 

• House is about 4000 sq feet, two floors, but still pretty spread out.

• One main power panel, 400A service. With three subpanels with breakers.

• Third subpanel is only for the pool and washer dryer. Both main panels each have a signal linc installed.

• All power strips are either filtered via X-10 Pro 20 A in line or 15A plug in filters, or do NOT have any surge suppression or noise filtering. No UPS in the system.

• Numerous other devices are filtered via one of the above or a simple X-10 5A plug in filter or a leviton in line 5A filter.

 

 

I have recently installed two 2477SA1's, 240v Load controllers, for my pool filter and water feature pumps. (Note: these are dual band devices). They are fed from a separate power panel just for the pool pumps and other outdoor power devices for the patio, and are mounted right next to each other.

 

I noticed two reliability issues with the control of these two devices:

 

1. Scene ON/OFF commands were reliable about 75% of the time when done through the remote linc.

2. The ISY did not always have the correct status of the device, ie ISY thought it was OFF, when it was actually ON, and vice versa. This occurred about 1 in 8 to 10 times.

3. One of the pump scenes was also controlled by a KPL button F, and that control seemed to work all of the time.

 

I disabled ALL programs in my ISY-99.

I removed / unplugged all 5 access points throughout the house.

I unplugged all 4 of my dual band lamp linc modules.

Thus, NO RF access into the network EXCEPT through the 240v load controllers themselves.

 

I then ran Scene tests on both scenes and got different results on each. I then removed the KPL from the one scene, so that both scenes had only one device in them, the load controller itself. At that point, as expected, the results of the scene tests were virtually identical for both scenes. I always got one of the two results pasted in below. I got each of the results about 50% of the time.

 

Note: In the power panel for the pool pumps, there are two GFCI single pole breakers for some outlets, but NOT for the pumps.

 

My questions are:

 

1. Are these two different results indicative of a noise problem? If so, suggestions for further troubleshooting would be appreciated.

 

2. Please help me decipher these event viewer readouts. In particular, what is happening in the GRP-RX, CLEAN-UP-RPT, and the Standard-Cleanup Ack? Is something happening that shouldn't? Or something NOT happening that should???? (PLEASE don't hesitate to refer me to a manual or document somewhere, I just haven't seen anything that explains this in detail).

 

3. What does the process message failed actually mean, not recieved? Not acknowledged? Not acted upon?

 

As always, any help with this would be greatly appreciated.

__________________________________________________________________________________________________________________

Thu 09/01/2011 01:01:33 AM : [GRP-RX ] 02 61 4A 13 00 06

 

Thu 09/01/2011 01:01:33 AM : [CLEAN-UP-RPT] 02 58 06

 

Thu 09/01/2011 01:01:33 AM : [iNST-SRX ] 02 50 14.5C.0B 0F.D5.9D 61 13 4A LTOFFRR(4A)

 

Thu 09/01/2011 01:01:33 AM : [standard-Cleanup Ack][14.5C.0B-->ISY/PLM Group=0] Max Hops=1, Hops Left=0

 

----- Pool Water Feature Scene Test Results -----

[succeeded] Pool Water Feature Pump (14 5C B 1)

----- Pool Water Feature Scene Test Results -----

Thu 09/01/2011 01:01:41 AM : [iNST-ACK ] 02 62 00.00.4A CF 13 00 06 LTOFFRR(00)

 

___________________________________________________________________________________________________________________

 

Thu 09/01/2011 01:02:27 AM : [GRP-RX ] 02 61 4A 13 00 06

 

Thu 09/01/2011 01:02:27 AM : [CLEAN-UP-RPT] 02 58 06

 

Thu 09/01/2011 01:02:27 AM : [iNST-SRX ] 02 50 14.5C.0B 0F.D5.9D 61 13 4A LTOFFRR(4A)

 

Thu 09/01/2011 01:02:27 AM : [standard-Cleanup Ack][14.5C.0B-->ISY/PLM Group=0] Max Hops=1, Hops Left=0

 

Thu 09/01/2011 01:02:27 AM : [iNST-SRX ] 02 50 14.5C.0B 0F.D5.9D 61 13 4A LTOFFRR(4A): Process Message: failed

 

Thu 09/01/2011 01:02:27 AM : [standard-Cleanup Ack][14.5C.0B-->ISY/PLM Group=0] Max Hops=1, Hops Left=0

 

----- Pool Water Feature Scene Test Results -----

[succeeded] Pool Water Feature Pump (14 5C B 1)

----- Pool Water Feature Scene Test Results -----

Thu 09/01/2011 01:02:35 AM : [iNST-ACK ] 02 62 00.00.4A CF 13 00 06 LTOFFRR(00)

[LeeG]

justin.cool

 

“Are these two different results indicative of a noise problem?â€

 

No. In the second test an ACK was received from the Load Controller twice. Not uncommon with RF devices

 

“Please help me decipher these event viewer readouts.â€

 

Thu 09/01/2011 01:01:33 AM : [GRP-RX ] 02 61 4A 13 00 06

 

When a serial command is sent to the PLM it echoes the command back to the application with an 06 appended if the command is accepted. The 02 61 is a Scene (Group) Off command that initiates the Scene Test

 

Thu 09/01/2011 01:01:33 AM : [CLEAN-UP-RPT] 02 58 06

 

A Scene (Group) command sequence involves sending Group Cleanup Direct messages to each Responder in the Scene (Group) to confirm it received the Scene (Group) command. The 02 58 Report is sent to the application when all Group Cleanup Direct messages have been sent and ACKed.

 

Thu 09/01/2011 01:01:33 AM : [iNST-SRX ] 02 50 14.5C.0B 0F.D5.9D 61 13 4A LTOFFRR(4A)

 

This is the ACK from 14.5C.08 for the Scene (Group) Off command with an ACK code of 4A. SmartLabs is very stingy with its information, I do not know what a code 4A means. Since it is with an ACK it has some non failure meaning. Could be something like there is no load attached to the device or something totally unrelated to the load itself.

 

“What does the process message failed actually meanâ€

 

Many RF devices send multiple copies of the same message. Motion Sensors send 3 copies of Motion On and Motion Off messages. Where duplicate messages are common the first duplicate is identify as duplicate, the second duplicate is labeled as failed. In this case the ISY is not expecting a duplicate message as the norm so it labels it as failed. It simply means the ISY has already processed the expected ACK and this is a duplicate.

 

There is a very old PLM Developer’s Guide floating around the internet dating from 2007. It describes PLM serial commands 0261, 0262 (many more) and the responses 0251, 0252 (many more) that were spec’d back then. There is a much later version which is proprietary and confidential. Access to the current information requires a developer subscription and an NDA making if very confining as far as sharing information.

 

Lee

[justin.cool]

Lee,

 

As always, I very much appreciate your prompt response. Sounds like the following can be deduced:

 

1. There were no comms issues during any of my scene tests with these devices

 

2. The two different event viewer responses are the result of the RF device under test resending the command, and thus are not an indication of any comms issues.

 

So one last question, why does the duplicate transmission of the ACK get sent only sometimes, and not everytime?

 

In the meantime, I will exercise my system at various times of the day to see if I can get the failures to occur, and THEN run scene test.

[LeeG]

justin.cool

 

It is not absolutely clear that the duplicate ACKs are actually being sent by the Load Controller. That is one possibility and RF devices like Motion Sensors and TriggerLincs do in fact send multiple On messages. With multiple paths for a message to travel, particularly when the source is a Dual Band device itself, it could be the Load Controller sent one ACK message which travels over the powerline and over RF. In theory something in the Insteon mess network should reconcile the arrival of both a powerline message and an RF message, eliminating the duplicate. In practice I see duplicates in my testing.

 

ACK messages are not themselves ACKed so this is not a retry situation due to powerline issues.

 

Bottom line, there is no way I know of to determine if the Load Controller actually sent two ACKs or they are the result of being sent on both the powerline and RF and they were treated as different messages rather than the same message arriving in different ways.

 

Is the ISY PLM a 2412 or 2413?

 

Lee

[LeeG]

justin.cool

 

In a previous post I mentioned the 4A being a code in the ACK which I did not know what it meant. Each Insteon Scene has a Group number associated with it. The Group number is how Insteon relates devices to a particular Scene. That 4A in the Scene test ACK is the Group number of the Insteon Scene. I ran some Scene tests here and realized the 21 in my ACKs are in fact the Insteon Group number of the Scene under test. The 4A is the Insteon Group number of the Scene you are testing.

 

 

Thu 09/01/2011 01:59:56 PM : [GRP-RX ] 02 61 21 13 00 06

 

Thu 09/01/2011 01:59:56 PM : [iNST-SRX ] 02 50 16.3F.93 12.9F.E4 61 13 21 LTOFFRR(21)

 

Thu 09/01/2011 01:59:56 PM : [standard-Cleanup Ack][16.3F.93-->ISY/PLM Group=0] Max Hops=1, Hops Left=0

 

Thu 09/01/2011 01:59:56 PM : [CLEAN-UP-RPT] 02 58 06

 

Thu 09/01/2011 01:59:56 PM : [iNST-SRX ] 02 50 15.4C.0D 12.9F.E4 61 13 21 LTOFFRR(21)

 

Thu 09/01/2011 01:59:56 PM : [standard-Cleanup Ack][15.4C.0D-->ISY/PLM Group=0] Max Hops=1, Hops Left=0

 

----- SceneTest Test Results -----

[succeeded] LampLinc 6 (15 4C D 1)

[succeeded] SwitchLinc Dimmer (16 3F 93 1)

----- SceneTest Test Results -----

Thu 09/01/2011 02:00:04 PM : [iNST-ACK ] 02 62 00.00.21 CF 13 00 06 LTOFFRR(00)

 

Lee

[justin.cool]

Lee,

 

Again, thanks for your response. Now that I have done some reading this makes a lot more sense. However, there is one stand out difference between the scene test results you got versus what I got. That is the ordering. From what I now understand, the order of your events seems to make more sense. Whereas with my scene test, the [CLEAN-UP-RPT] 02 58 06 response came back as the second event. Is this significant? Or is the timing not significant when they all have a time stamp for the same second?

 

Justin

[LeeG]

justin.cool

 

I think the difference is my Scene has multiple responders. As each responder receives a Group Cleanup Direct it sends an ACK which is passed on to the application. The first inbound 0250 ACK is from the first device in the Scene. The PLM does not stack up the ACKs internally. They are passed to the application as they are received. When the 0250 ACK for the last device in the Scene is received the PLM sends the 0258 Scene Report to the application and then sends the last ACK to the application.

 

Sorry for the confusion. I will test this to confirm but I believe the difference in sequence is caused by my Scene having multiple devices and yours having 1 device.

 

Lee