Improper Programm Behavior Since Firmware Update.

[Brian H]

Smarthome accidentally had the earlier Insteon PLM Developers Guide avilable for download.

It is still avilable on a few sites.

http://www.aartech.ca/docs/2412sdevguide.pdf

 

Membership in Developers Group. Isn't what it is cracked up to be.

Though occasionally you do get a few tidbits.

Serious consultation requires a fee be paid to Smartlabs.

[apostolakisl]

Thanks guys.

[apostolakisl]

Am I assuming correctly that when an RF signal linc device transmits to another signal linc, and then back to plc, that this counts as 2 hops?

 

If indeed that is the case, then using these devices to connect phases in a house would always be a more than 1 hop event.

 

I would think that a dual band plm would potentially fix this.

[IndyMike]

Hi Lou,

 

I believe it requires 1 Hop for the Insteon RF couple.

 

Using the chart below:

1) Initial transmit requires 5 zero crossings of the 60 Hz AC.

2) Between the Initial transmit and the 1st hop transmit there is one "open" zero crossing. The RF devices communicate during this half cycle.

3) RF devices will then transceive the RF info onto the powerline during the 1st HOP.

 

[apostolakisl]

IM,

 

So, are you saying that from power line, to signal linc, to second signal linc, and back to power line is 1 hop total? I'm pretty sure that is what you are saying. I don't really know how to interpret the graphs since I don't know what device is at each hop.

 

So, if indeed that is the case, then what affect would a dual band plm have?

 

Flow chart (SWL = switchlinc, SGL = signal linc, PLM SB = non-rf plm, PLM DB = dual band PLM )

 

SWL => SGL => SGL => PLM SB ??? 1 hop ???

SWL => SGL => PLM DB ???? 0 hops ????

 

I assume that a "hop" is when the signal gets repeated by a device that is not the target. So if a SWL message goes directly to the PLM, that would be 0 hops.

 

So, according to my above theory (the validity of which is certainly in question), a dual band plm in some situations could get rid of one hop.

 

Lou

[IndyMike]

IM,

 

So, are you saying that from power line, to signal linc, to second signal linc, and back to power line is 1 hop total? I'm pretty sure that is what you are saying. I don't really know how to interpret the graphs since I don't know what device is at each hop.

 

So, if indeed that is the case, then what affect would a dual band plm have?

 

Flow chart (SWL = switchlinc, SGL = signal linc, PLM SB = non-rf plm, PLM DB = dual band PLM )

 

SWL => SGL => SGL => PLM SB ??? 1 hop ???

SWL => SGL => PLM DB ???? 0 hops ????

 

I assume that a "hop" is when the signal gets repeated by a device that is not the target. So if a SWL message goes directly to the PLM, that would be 0 hops.

 

So, according to my above theory (the validity of which is certainly in question), a dual band plm in some situations could get rid of one hop.

 

Lou

 

Hello Lou,

 

Yes - I believe that in a RF coupled system, A device that transmits from the opposite phase will be heard by the PLM on the 1st hop.

 

Here's a bit better chart -

 

 

A dual band PLM could conceivably hear a RF transmitter during the original transmit (HOP 0). This wouldn't buy you anything in terms of system response since no commands are transmitted with a hop count less than 1 (that I'm aware of). The PLM must wait until the HOP count expires before it can respond to the transmission.

 

I also do not know what the PLM (or any RF device) will do if confronted with RF and Powerline information that doesn't agree.

 

Hope that helps,

IM

[apostolakisl]

A dual band PLM could conceivably hear a RF transmitter during the original transmit (HOP 0). This wouldn't buy you anything in terms of system response since no commands are transmitted with a hop count less than 1 (that I'm aware of). The PLM must wait until the HOP count expires before it can respond to the transmission.

 

IM

 

I am trying to get the message to complete in 1 hop. Currently I have some 2 hop situations which are screwing things up. If I can turn the 2 hop into 1 hop, it might help.

 

The problem is that the PLM is receiving a message, but the SWL is not getting the ACK on the first try which is limited to 1 hop. So the SWL sends a second message, which the PLM also receives (causing erroneously program trigger), this time the ACk is limited to 3 hops (which gets through with one hop left).

 

So, I am not aiming for zero hops total, just one less hop.

[TJF1960]

apostolakisl,

 

Out of curiosity have you tried removing some or all of your AP’s and running the tests?

[IndyMike]

Lou,

 

If I understand Lee's diagnosis correctly, this is not a HOP count situation -

 

1) The PLM heard your SWL on it's original broadcast and activated your program. The PLM attempted to acknowledge, but the cleanup command was sent with a Hop count of 1. The PLM had to respond with a hop count of 1.

 

2) The SWL did not hear the PLM's acknowledge, and sent a second cleanup message to the PLM with a Hop count of 2. This activated your program a second time and caused the unwanted effects. The PLM acknowledged a second time and your SWL apparently heard the acknowledge with a hop count of 2 (I didn't try a third time).

 

These are separate communications. The issue is that your SWL can't hear the PLM with a hop count of 1.

 

Noise and signal absorbers act differently on transmitters and receivers -

1) A signal absorber close to a transmitter will normally not prevent devices from hearing it's transmissions. This is because the transmitter is capable of driving the load of the absorber.

2) A signal absorber close to a receiver will draw down incoming signal levels due to the impedance (resistance) of your home wiring.

3) A noise source close to a transmitter will have little affect on it's ability to transmit effectively.

4) A noise source close to a receiver can interfere due to the decreased level of the inbound signal.

 

From the sounds of things, your PLM can listen and transmit effectively. At the opposite end (your SWL) the signal is either being absorbed for interfered with by noise - normally both with exist.

 

If your SWL is within RF signal range, you could try moving a accesspoint to the same circuit. This could improve the signal/noise ratio to the point where the SWL could hear the PLM on one Hop. Not a fix - but it would help to isolate the problem.

 

IM

 

Edit: To answer your question on a 0 hop configuration - The only way that I know to accomplish this is with a passive coupler in close proximity to the PLM. The signal will be coupled to the opposite phase on the initial transmission. This would allow devices on both phases to repeat the signal on the first HOP.

 

Do you know whether your SWL is on the opposite phase from your PLM?

[apostolakisl]

apostolakisl,

 

Out of curiosity have you tried removing some or all of your AP’s and running the tests?

 

 

OK, yes, just did it. I didn't even know how many I had. Turns out I found 5. I think I found them all. Anyway, I querried every device and they all respond. Don't know how it is getting from one phase to the other. Must be a 220 appliance.

 

Anyway, I turned off a bunch of the lights with the program in question and none misbehaved. Of course, sometimes they work perfectly, and then other times they won't. So, if all goes well for a week or so, then I will thank you for the suggestion.

[LeeG]

The last post by IM describes my working theory exactly. Thanks IM. As with most things Insteon when the powerline is concerned it is a theory. What is not a theory is the ability to communicate from A to B does not insure B can communicate with A. In some of the traces even A is having trouble communicating with B as the 2 Max Hop message does not make it from the SwitchLinc to the PLM. In those cases it was the next retry, 3 Max Hops that was successful in reaching the PLM. I’m pretty sure there is an intermittent source of interference. Something that is affecting more than one SwitchLinc as it is reported that this sequence is seen on more than one SwitchLinc.

[TJF1960]

The only way that I know to accomplish this is with a passive coupler in close proximity to the PLM. The signal will be coupled to the opposite phase on the initial transmission. This would allow devices on both phases to repeat the signal on the first HOP.

 

FWIW my plm is 40 wire feet from the panel which has a hardwired passive phase coupler. I have no AP's but do have 3 dualband plug in dimmers thru out the house which may or may not be on opposite legs, never tested.

[apostolakisl]

Lou,

 

2) The SWL did not hear the PLM's acknowledge, and sent a second cleanup message to the PLM with a Hop count of 2. This activated your program a second time and caused the unwanted effects. The PLM acknowledged a second time and your SWL apparently heard the acknowledge with a hop count of 2 (I didn't try a third time).

 

These are separate communications. The issue is that your SWL can't hear the PLM with a hop count of 1.

 

?

 

That was my point. The SWL didn't hear the ack in 1 hop (it took two) which resulted in the SWL message being sent to the PLM twice. On the second ACK from the PLM, it allowed for more hops, so it got through to the SWL. So I was trying to get the ACK to go through in one hop (not 2). Not knowing the path taken, I consisdered that the rf signal lincs may have been involved. If using a dual band plm takes one hop out versus two signal lincs, then that was my idea.

 

Anyway, a full evening and full morning of activity have completed in my home without any signal lincs or rf insteon devices of any kind. Everything is working at least as well as before, possibly better. None of the lights this morning misbehaved as per the original problem. My home has standard two phase power with three subpanels. As far as things connected to both phases, that would include the oven, heat pumps, well pump, air handlers?, two clothes dryers. The oven has not been on, the heat pumps/air handers are probably off (it is actually 68 right now), the dryers are off, and the well is off. So at least one of those devices must have some standby draw that is using both phases, I guess.

[IndyMike]

Lou,

 

2) The SWL did not hear the PLM's acknowledge, and sent a second cleanup message to the PLM with a Hop count of 2. This activated your program a second time and caused the unwanted effects. The PLM acknowledged a second time and your SWL apparently heard the acknowledge with a hop count of 2 (I didn't try a third time).

 

These are separate communications. The issue is that your SWL can't hear the PLM with a hop count of 1.

 

?

 

That was my point. The SWL didn't hear the ack in 1 hop (it took two) which resulted in the SWL message being sent to the PLM twice. On the second ACK from the PLM, it allowed for more hops, so it got through to the SWL. So I was trying to get the ACK to go through in one hop (not 2). Not knowing the path taken, I consisdered that the rf signal lincs may have been involved. If using a dual band plm takes one hop out versus two signal lincs, then that was my idea.

 

My mistake - I thought you were trying to improve communications to the PLM when the problem is actually in the opposite direction. A dual band PLM communicating to a dual band SWL might eliminate one hop. A dual band PLM to a Accesspoint would (I believe) still require one hop. Based on my subjective experience, I can would say that the dual band PLM is not as effective a RF receiver as and accesspoint. I do not know how the two compare when transmitting.

 

ELA and I have been having a sidebar conversation on this and he has brought up a good point. Insteon devices are characterized for 3.2 Vp-p into 5 ohms. If you had a local signal absorber (plug in device with a EMC cap) at or below 5 ohms it would reduce the output of your PLM. In short, please police your PLM circuit for potential absorbers.

 

Anyway, a full evening and full morning of activity have completed in my home without any signal lincs or rf insteon devices of any kind. Everything is working at least as well as before, possibly better. None of the lights this morning misbehaved as per the original problem. My home has standard two phase power with three subpanels. As far as things connected to both phases, that would include the oven, heat pumps, well pump, air handlers?, two clothes dryers. The oven has not been on, the heat pumps/air handers are probably off (it is actually 68 right now), the dryers are off, and the well is off. So at least one of those devices must have some standby draw that is using both phases, I guess.

 

This is really interesting. You may be one of the lucky few that has a low impedance path through your utility transformer (rare but it does happen). I am able to communicate through the transformer "at times". If I stress my system (turn on various CFL loads) my I break the system. My phase coupling is through a hardwired passive coupler (similar to Tim's).

 

I have also been playing with a accesspoint close coupled to the dual band PLM. I've been performing link table scans of some of my KPL's and post processing the Event viewer to determine communication efficiency. I'm still playing with this, but I have a number of test runs that indicate that adding the accesspoint increases PLM retries. I will confess that I do not understand how the accesspoints/dual band features operate and am struggling to come up with a theory why RF coms would result in retries. The best I can come up with is a difference in powerline data VS RF data causing the PLM to throw up it's hands and retry the transmission.

 

Would you mind trying some link table scans on your SWL to see how reliable the comm's are?

 

IM

[apostolakisl]

ELA and I have been having a sidebar conversation on this and he has brought up a good point. Insteon devices are characterized for 3.2 Vp-p into 5 ohms. If you had a local signal absorber (plug in device with a EMC cap) at or below 5 ohms it would reduce the output of your PLM. In short, please police your PLM circuit for potential absorbers.

 

I had tried running extension cords from my plm to various other outlets around the house (100 ft cord was more handy than long cat5) and I couldn't really find any improvement in performance. I actually used a washing machine plug that was on its own breaker. Of course, at the time I was looking at it differently.

This is really interesting. You may be one of the lucky few that has a low impedance path through your utility transformer (rare but it does happen). I am able to communicate through the transformer "at times". If I stress my system (turn on various CFL loads) my I break the system. My phase coupling is through a hardwired passive coupler (similar to Tim's).

 

Possible. I do have over 100 feet from my panel to the transformer, but I guess that isn't that long.

Would you mind trying some link table scans on your SWL to see how reliable the comm's are?

IM

 

I can do that tonight.

 

PS, anyone want to buy some signalincs? Seems that mine are just wasting electricity.

[LeeG]

apostolakisl

 

I'm really not trying to be picky, those are Access Points not SignaLincs.

 

Lee

[apostolakisl]

apostolakisl

 

I'm really not trying to be picky, those are Access Points not SignaLincs.

 

Lee

 

Did they change the name at some point? I see on SH website that signal lincs are the hardwired phase couplers, but I know for a fact that the word "signalinc" is on at least a couple of mine followed by the item number. I will confirm tonight. I think it might actually say "rf signalinc". These guys are fairly old. Heck, I don't even know if they are doing anything at all. The led lights up on them but other than that I have no way to check them. I suppose I could isolate my plm using a filterlinc and see if it still can communicate by putting one of these "rf signalincs" on each side of the filterlinc.

[LeeG]

Do they have a black external antenna? I assumed the units removed were pluggable devices, not the hardwired devices.

[apostolakisl]

Do they have a black external antenna? I assumed the units removed were pluggable devices, not the hardwired devices.

 

Yeah, the hinged black antenna and they plug in. Most if not all of them have the pass through plug, which I see the "access points" on SH website don't have.

 

I'll look at all of them when I get home and see. But I am pretty sure none have the words "access point" written on them. I didn't really study this issue, however, as I didn't know it was worthy of such attention. But I did for sure look closely at one of them before I started posting about them because I wanted to make sure I referred to them by their proper name! Who knew I was sabotaging myself.

[LeeG]

Thanks. The terminology does not matter. The devices that have the black external antenna operate on a different RF frequency than any of the more recent RF devices. They also do not couple Insteon Extended messages. Devices such as Motion Sensors, TriggerLincs, Venstar tstat adaptors, Access Points, the many new Dual Band devices, the RemoteLinc and RemoteLinc2, none of these devices operate on the RF frequency used by the RF coupler with the black antenna. Removing the RF couplers with the black external antenna would only affect RF coupling of Insteon Standard messages on the powerline.

[apostolakisl]

SignaLinc RF #2442 are the units. Manufact dates are 12th week 06 and 24th week 05.

 

And I don't have any of the stuff that uses the extended messages. All of those sorts of things in my house are part of the Elk.

 

As far as Insteon is concerned, I only have SWL's, ApplianceLincs, Lamplincs, and a few icon dimmers. I have about 60 devices total.

[LeeG]

Device to device control is done with Standard messages. Passive coupling must be working as much device configuration is now done with Extended messages.

 

I am amazed your system works at all!

[apostolakisl]

IM

 

I ran links tables on several devices. What did you want to know?

 

And I would say that my system is currently working without error with no RF devices. I have not been able to create the false program trigger that this whole thread goes back to.

 

Lou

[IndyMike]

Lou,

 

You've already answered the question. Reading device link tables is a communication intensive action that takes a fair period of time. If your system was having communication problems you would likely have receive the dreaded "xxx failed to communicate" error.

 

I have a excel macro that can post process the event viewer results from a link table scan. It rolls up statistics on the number of hops required throughout the process. If you're interested, PM me and I can give you instructions.

 

Regarding your PLM - is it a dual band unit?

[apostolakisl]

So far,

 

Still have the rf signalincs sitting in a box. The system works pretty much the same, maybe even better. I am not having any failed commands, but I still get the occasional false trigger of that program related to too many hops.

 

Questions: is there any way to trace the hop path taken. Knowing that might help find a shorter path. maybe

 

Do you think a hardwired signal linc at or near one or all of my subpanels might do the trick? It wouldn't surprise if the need for multiple hops has to do with getting from phase to phase. Although you would think the rf signalincs would have solved that, especiallly with 5. What are the odds that all would be on the same phase, pretty low I think, but I never checked.

 

Do the hardwired x10 phase couplers work the same as the Insteon ones? I actually have one of those sitting in a drawer right now.