Strange Com Problem

[pkeberly]

If deleting a device and adding it back resolves the issue (and Restore Device does not) the solution is delete the devices and add them back.  

 

The "device does not support PLM settings" is expected for non I2CS devices.  

 

 

I'm not sure what that is, but I have three 2477(D&S) devices that will do it, and rest will not. 

 

Do I manually delete them all, then do an ISY restore? Or do I need to manually re-link evertbody?

My backup is from sometime during the PLM swap, hopefully that's not corrupted too?

 

Thanks Lee

[pkeberly]

Yeah, I believe you might be walking the path what LeeG has suggested and starting from scratch. This won't be so painful as you have a good back up to restore, correct?

 

I think that's the way it has to be.  My restore might be sketchy, and I never liked my device naming anyway.  If this was a fresh install, how would you go about it? 

 

I've been palnning to move the fan swtiches from the attic, down into the baths, so they are manually operable for siutations like this .

Over the holidays, we consolidated the lighting circuits, and have them all on the same phase.

We planned for incandecents, but switched to all LEDS after we pulled the wires, so I had way more circuits than I needed.

 

Thanks Teken

[LeeG]

Restoring an ISY backup does not actually fix anything itself.   Assuming the backup is good and current a File | Restore Devices is required to rebuild the devices and the PLM.   If the ISY backup is good but not current the Restore Devices will bring the devices and PLM to the older level.   From there any newer devices/Scenes/Programs have to be recreated.   

[Teken]

I would probably just hard reset all the devices from scratch. Once done add each device using the ISY. From there create all the scenes you want and name them accordingly.

 

This sounds like a lot of work but it brings you to a known good state and let's you move forward.

 

Doing all of this does not address coupling / noise issues.

 

Before you undertake anything swap out the lights with incandescent and verify coupling.

 

Order a few filter lincs to isolate your A/V gear.

 

 

Sent from my iPhone using Tapatalk

[pkeberly]

I would probably just hard reset all the devices from scratch. Once done add each device using the ISY. From there create all the scenes you want and name them accordingly.

 

This sounds like a lot of work but it brings you to a known good state and let's you move forward.

 

Doing all of this does not address coupling / noise issues.

 

Before you undertake anything swap out the lights with incandescent and verify coupling.

 

Order a few filter lincs to isolate your A/V gear.

 

 

That's my plan, a hard reset for everybody and start fresh.  

Do you have any favorite naming schemes? 

 

I don't think I have much of a noise issue, at least I didn't before the PLM died. 

It appears my scheduled events worked last night, but device initiated events still fail to trigger programs, strange.

 

I need to do a bit of re-wiring. My bath fan switches are remote, they'll get moved into the baths where they belong, poor planning on my part.

I'm pretty sure I now have all of my lighting circuits on one phase. (across two panels), I just need a feed from that phase into the rack dedicated to the PLM.

 

Thanks again to you and Lee for taking the time to help out, very much appreciated,

 

 

Teken, I love your quote.

[Teken]

 

Do you have any favorite naming schemes?

 

I don't believe there is a right or wrong but there are some best practices outlined in this great thread. Perhaps it will give you some insight and ideas to implement: http://forum.universal-devices.com/topic/15025-organizational-structure-question/

 

 

I don't

 think I have much of a noise issue, at least I didn't before the PLM died.

 

I have found over the years from doing more than 350+ installs to reading various forums about Insteon. You will find either the *Insteon* network is very forgiving or operating on a marginal status. Meaning, most people are seeing 90% instant on and a few second delays in scenes activating.

 

This is due mainly to the great 3 hop re-broadcast Insteon uses in its protocol. Having said this, one should never rely on the system using up all its hops to get that last switch to operate / turn off. In the long run this impacts the over all reliability, consistency, and the persons trust in the Insteon network.

 

If you don't trust the device to perform that means its impacted your view and belief in the product. That translates to poor customer satisfaction and that persons view becomes negative over all.

 

You have already made a huge investment into automating your home so take the extra step and purchase a few Filter Lincs: http://www.smarthome.com/filterlinc-10-amp-plug-in-noise-filter.html

 

As stated before sometimes noise can be noise makers vs signal suckers. The level of sucking vs generating varies with each device and how far it is from the Insteon PLM / Insteon devices. Lots of people are in the belief the *Dual Band* technology is the be all, end all.

 

Its not . . .

 

If a wired device is installed into a metal JBOX the RF prorogation will be extremely limited. This can be from 50-90% in signal drop coming out of the Insteon device. Keep in mind the actual antenna is in the back of the device so placing any Insteon device into a box is going to limit its RF potential.

 

This is why having a few plugin Insteon devices scattered around the home proves useful and beneficial in keeping the RF strength up. I normally install either a Range Extender or plugin On/Off, Lamp Linc. Using the latter two provides dual use and can be moved around until the coupling / bridging via the 4 tap (beacon) test is confirmed green in at least four corners of the home.

 

As you indicated before there were quite a few places were it was red. This is where you have lots of the same gear on the same side of the electrical feed. The goal is to couple the home from top to bottom and from all four sides of the house. Most people start by coupling as close to the service panel.

 

This provides the best signal prorogation and from there it really doesn't matter where you put other devices so long as it mkaes sense and is going to be used.  

 

 

Teken, I love your quote.

 

I love the quote too!

 

It really does express life in general and this may aptly apply to your case when you figure out whats going on and the answer falls before you.

 

Keep us all in the loop as to your progress and should you need more help lots of folks will be jump in and help.

[paulbates]

I find it hard to beleive my whole system get's knocked out by noise, at exactly the same time my PLM fails?

 

Paul, just catching up on this thread.

 

It can be cause and effect: enough line noise over time can kill a PLM. I've had this happen with a furnace motor that over time caused a PLM to lock up intermittently, and then eventually die. The new PLM started exhibiting the lock up behavior. I put a filter on the furnace and kept the plm for 3 years and recently sold it after converting to ISY (USB PLM)

 

I would consider Teken's filtering suggestions from earlier in the thread.

 

Paul

[pkeberly]

Thanks Teken, Lee, and Paul.

 

I'm about 2/3rds done with the resets, but my efforts got preempted by a weekend task.

 

I fully agree that I shouldn't rely on the hops, and that I should install filters.  I'm not against it at all, I just can't expend the time right now.

We're coming up on our 4th year of building a new home.  The system is stable enough for us to get by for now.  I WILL put in the put in the effort into making it right, someday...

 

Paul Bates, is your furnace motor the newer ECM type or the older PSC type? 

How did you determine it was the furnace motor, what clues did you see?

 

Thanks again guys, I'm off to stack fireplace (fake) stone. Hopefully this will be my last weekend at that task.

 

Paul K

[paulbates]

.. is your furnace motor the newer ECM type or the older PSC type? 

How did you determine it was the furnace motor, what clues did you see?

 

Paul

 

The one I wrote about was a 25 year old PSC motor.  It was replaced last fall with a new Carrier unit with an ECM motor. I got my filter back from the HVAC contractor from the old unit. I did not reinstall the filter and no noise problems re-appeared.

 

How I determined it is some deja vu. Someone on a board suggested that I look for big noise sources near the PLM. The age of the furnace made it suspect, and it was only a few wire feet from the panel/PLM.  I had someone loan & send me an X10 signal / noise analyzer. It showed a lot of noise when plugged in where the plm went and that furnace kicked on.

 

The cost and/or availability of power-line signaling tools makes it hard to access them. To get around that, there are some general guidelines: Don't put the PLM near motors, big electronics setups or transformers, especially low voltage lighting transformers.

If you have no choice,  filter those things to protect the PLM from damage. I recommend filtering those things in general for good communications with everything on your network.

 

Paul

[pkeberly]

Just an update on progress.  Over the weekend I wrapped up circuit consolidations, and rewired 3 of the 4 bath fans to local control. Also finished adding the rest of my devices, and re-built the scenes and programs.

It works good, much better than before.  Out of our 55 devices, I only have 11 that show less than 3 hops remaining when doing a query all.  It got too late to investigate the weak devices, but I did remove the passive phase coupler during wiring changes.  I'm sure installing that, and repeaters should help a lot. hopefully this weekend.

 

I did have an interesting observation.  Everything was back to normal until last night, it ALL crashed again!  I had NO receive messages at all, but the ISY could transmit...again!  

 

I can pin down the time to about a 2 hour period.  I was doing some some minor code changes and cleanup right before the issue was noticed.  The only changes made; to add a prefix to all of my state variables, and delete two scenes that had large device counts.  

 

To help testing, I created 2 scenes, "good" & "poor".  If a device had 3 hops remaining, it went in the "good" scene, else it went to "poor" (as a responder)  I stopped doing that at 11 in poor, and about 22 in good.  That made it easy to test, just query the scene, and scan for hops.  It feels like it all went south right after I deleted those 2 scenes?  I did a PLM & ISY power cycle, still nothing.   Tonight I did a restore PLM, and it all came back working.

 

BTW, message filtering in the event viewer would be a spectacular feature, along with device names.

 

Case closed, I love my ISY.

Thanks again for all that helped out.

 

Paul

.