Insteon Sucks just a little less than X-10 did :(

[fitzpatri8]

An Insteon troubleshooting tool already exists--the Diagnostics section of HL2 used with a laptop PC. You can use HL2 diagnostics to test point-to-point communications from anywhere in the house, to repeatedly ping a particular device or all your AC-powered devices overnight looking for intermittent communications, to test for both standard and extended length messages, and to determine how many hops a signal normally takes to get to a particular destination. There's no 'signal strength' readout per se, but voltage isn't what you are really looking for in a mesh repeater network--you want to know how reliably signals get from point a to point b, something you can determine by looking at success rate and hop count.

[oberkc]

An Insteon troubleshooting tool already exists--the Diagnostics section of HL2 used with a laptop PC.

 

Good point. I have been thinking of getting houselinc2 just for that purpose alone. The only hesitancy I have is uncertainty about the usefulness or whether it is a more positive test than the ISY scene test. I am hoping for a device that gives more than a gross, go-no-go, determination and I have been simply uncertain if HL2 provides this. Do you view the HL2 as more useful than the ISY-99 scene test?

 

I would like to be able to measure gradual degradation in communication, even if not to the point of affecting hop count. I perceive that many times the problems we face are the result not of a single interfering device, but the sum total of many devices. Where a single device may not cause problems, several can. And, with the HL2, I am afraid that I cannot measure the effect of a single device.

 

Perhaps I am asking too much, or that such a device would cost more money than most of us is willing to spend.

[ELA]

Please keep in mind that since this is a "tuned" device, it will represent signals in the 120 KHz range rather accurately. Out of band noise (CFL's @40 - 80 KHz) will be attenuated. At one point I ran a frequency response test on the coupler to characterize it. If you're interested, PM me and I'll see if I can dig it up.

 

Thanks for the status update,

IM

 

Yes it is a dual tuned circuit - I removed it from its case and installed in a box with a fuse plus room for a future additional circuit if needed.

 

I ran a freq. generator sweep to assure it coupled ~ 1:1 at ~130Khz.

[ELA]

I agree that the ultimate tool would both provide signal strength and decoding of messages -along with other features...

 

In the mean time I believe the best bang for buck could be knowing the signal strength at 131khz.

I envision a device that is cable of detecting Insteon Messages in terms of confirming a valid bit stream.

 

"Knowing" the the 131Khz burst is a message then record an average (on peak values) of the bit stream for each message. Then present a table of relative voltage levels for each message , whether a hop or repeat.

 

Sure technical types would want all the features but how about a signal strength meter at least?

 

-or- maybe throw in a spectrum analyzer for the RF portion debugging

[ELA]

I cannot say for sure if this would be of value but looks like it might be?

 

This $199 device and a laptop may allow you to look at your 915Mhz RF levels and any possible interference in the nearby spectrum.

 

http://www.metageek.net/products/wi-spy

[johnradams]

An Insteon troubleshooting tool already exists--the Diagnostics section of HL2 used with a laptop PC. You can use HL2 diagnostics to test point-to-point communications from anywhere in the house, to repeatedly ping a particular device or all your AC-powered devices overnight looking for intermittent communications, to test for both standard and extended length messages, and to determine how many hops a signal normally takes to get to a particular destination. There's no 'signal strength' readout per se, but voltage isn't what you are really looking for in a mesh repeater network--you want to know how reliably signals get from point a to point b, something you can determine by looking at success rate and hop count.

 

I have a copy of the HouseLinc2 software, which I used for a while before switching back to ISY. I have been advised against using both ISY and HL2, but what I hear you saying is that there is a way to use HL2 for diagnostics only.

 

Should I use a different PLM from the one that I use for ISY? Which PLM do you recommend for this purpose?

 

Are there certain HL2 functions/features that I should be careful to avoid lest I cause conflict with the ISY?

[johnradams]

Thanks, Mike, for your useful and instructive feedback.

 

Your poolhouse installation sounds like a very challenging install. With a separate 200A service to a 14 x 20 structure you must have some large pumps/heaters/filters that are switching on and off.

 

What type of loads are you using with your Insteon units in the Poolhouse?

I somewhat misled you on the nature of the poolhouse. The poolhouse is a small house for people, with a sitting area, bath, kitchenette, etc. The loads are as noted elsewhere: ceiling fan, bathroom vent fan, some lights, etc.

 

The pool equipment is located some 20 feet away next to a shed, and fed from a 100 amp sub-panel (which is in turn fed from the aforementioned 200 amp panel). There are 2 other sub-panels downstream from the 100 amp panel; one for a well pump, and the other for the pumping system for the fish pond. BTW, the waterfall on the fish pond can be turned on and off via Insteon (which is fun when it all works).

 

What I am trying to say is the the poolhouse itself doesn't have any particularly weird loads -- the motors and such are all "downstream" from the 200A panel, on sub-panels.

 

I'm a hardwired fan, but that simply isn't an option for your remote building on a separate service. I am curious about how your communications failures exhibit themselves:

 

1) Nuisance messages that a device won't respond (while the device appears to operate OK)?

2) Total loss of communication to devices in the poolhouse (can't command either scenes or direct devices)?

3) Physical devices lockups (requiring an airgap)?

 

The behavior is a mix: some the devices have had "pending writes" sitting in limbo for many months. I added the devices to ISY using the "add by address" scheme, as this is a lot easier (for me) than running around the house pushing buttons. The ISY would generally detect the device OK, although sometimes I might have to try to add it a second time. I don't recall specifically, but I think that that the ISY probably had to do retries just to even detect the device in the first place.

 

Programming multi-switches and such has been very iffy, as reflected by the pending writes.

[IndyMike]

Hello John,

 

Your "pending write" devices are a problem. I'm assuming these are in the pool house. Until these are resolved, you cannot be sure that the devices are programmed properly. Once again, moving your ISY/PLM to the poolhouse should allow you to "restore" these devices and ensure they are properly programmed. Then you can step back and assess your communication.

 

As a side note, scene tests (don't know if you've been running these) will not operate properly with "pending writes" on devices.

 

 

The behavior is a mix: some the devices have had "pending writes" sitting in limbo for many months. I added the devices to ISY using the "add by address" scheme, as this is a lot easier (for me) than running around the house pushing buttons. The ISY would generally detect the device OK, although sometimes I might have to try to add it a second time. I don't recall specifically, but I think that that the ISY probably had to do retries just to even detect the device in the first place.

 

Programming multi-switches and such has been very iffy, as reflected by the pending writes.

[ELA]

John,

You mentioned earlier that Lee of Smarthome had reviewed your installation?

 

I am curious ... did he monitor the signal levels at the Pool house?

 

I really believe at some point a person needs to be able to monitor signal levels rather than just use a trial and error method.

 

The fact that you have a 200A service with two downstream 100 sub panels makes for a lot of potential signal suckers.

 

If I read correctly you have RF coupling via a WAP outside the house to one just inside the pool house. You then have a hardwire coupler at the service in the pool house?

Is there a second WAP in the pool house? I wonder if a second WAP ( on the other phase) in the pool house might help?

 

When I had my recent issues I diagrammed the house circuits and all the device locations. This helped me visualize where a weak point might be.

 

How far is the one WAP in the pool house from the service?

 

Is it possible to completely turn off the sub panels for a quick test? Again a device to monitor signal levels during such a test would be helpful.

[johnradams]

Thanks, ELA and others, for your input and insights.

 

John,

You mentioned earlier that Lee of Smarthome had reviewed your installation?

 

I am curious ... did he monitor the signal levels at the Pool house?

We didn't get that far, as there were many other issues which occupied us that day (v35 devices, filterlincs, WAP placement, etc.). We did measure signals in the main house, and things are pretty reliable here.

 

I really believe at some point a person needs to be able to monitor signal levels rather than just use a trial and error method.

Amen.

 

 

The fact that you have a 200A service with two downstream 100 sub panels makes for a lot of potential signal suckers.

Actually, the main house has 400 amp service, and the poolhouse has 200 amp service. The main house has 3 main panels (all connected directly to the cutoff switch); each of these 3 panels has one or more sub-panels. Furthermore, 1 of these panels is connected to an automatic transfer switch / generator setup.

 

The 200 amp poolhouse panel has a "cascade" of subpanels. Picture this: a 200 amp panel serves the poolhouse (lights, fans, fridge, etc.), and also feeds a 100 amp subpanel located some 20 feet away which powers pool equipment. In addition to powering th epool equipment, this 100 amp subpanel also feeds a small panel for a well pump, plus another small panel for a waterfall and pond system.

 

Why are subpanels considered "signal suckers"?

 

If I read correctly you have RF coupling via a WAP outside the house to one just inside the pool house. You then have a hardwire coupler at the service in the pool house?

Is there a second WAP in the pool house? I wonder if a second WAP ( on the other phase) in the pool house might help?

Yes, there is a WAP outside of the main house intended to "relay" to a corresponding WAP in the poolhouse. Hardwire coupler in the poolhouse, plus a 2nd WAP in the poolhouse (on the other leg).

 

When I had my recent issues I diagrammed the house circuits and all the device locations. This helped me visualize where a weak point might be.

Ugh.

 

How far is the one WAP in the pool house from the service?

 

Is it possible to completely turn off the sub panels for a quick test? Again a device to monitor signal levels during such a test would be helpful.

 

The WAPs in the poolhouse are all close together, as the poolhouse is single-story, and has a footprint of perhaps 16 by 30 feet.

 

If indeed subpanels are "signal suckers", then I suppose I can cut them off to try and isolate. But I need those subpanels to be energized; what would/could I do if they are indeed "sucking" the signal?

 

On another thread, I posted a query about using HouseLinc2 purely for diagnostics purposes (I still have my copy from when Steve Lee was here; we used HL2 for debugging as he wanted to stick with the SH product). Should I sue a separate PLM if I want to use HL2 for diagnostics? Any tips/tricks I should know about? As you can appreciate, I am keen on diagnostics (and anything that UDI and SH can do to improve reliability). Features can always come later. Without reliability, gadgets are, well, just gadgets.

[ELA]

John,

I did not mean to imply that sub-panels were always going to be signal suckers. All I meant was that you have so many panels that the potential was there for a lot of signal suckers.

 

Whether or not they suck signal depends upon the loads on that panel.

 

My suggestion was strictly as a debug tool to turn off all subpanel loads all at once (but only if not too big of a hasstle). This is a temporary test. If their were signal suckers on that panel then you could further isolate and filter them.

 

You have a large and complicated electrical layout. I can see my suggestion of diagramming may be out of the question in your case

 

Wish I could be of more help. Best of luck to you.

[johnradams]

Thanks for clarifying, ELA. Yes, cutting off subpanels is a handy way to isolate problems. We'll pursue this "sometime soon", or as Jerry Pournelle would frequently write in Byte magazine in the 80s, "Real Soon Now".

[wrj0]

I cannot say for sure if this would be of value but looks like it might be?

 

This $199 device and a laptop may allow you to look at your 915Mhz RF levels and any possible interference in the nearby spectrum.

 

http://www.metageek.net/products/wi-spy

Perhaps a little late to be useful, but here's another spectrum analyzer USB device that costs just $39. http://ubnt.com/airview

It is interesting to observe Insteon RF activity - here's an example from my house:

 

[Brian H]

Nice device.

Are you using the Airview9 with the internal antenna or did you use the slightly more costly Ariview9-EXT with an external antenna?

[wrj0]

I went with the Airview9. Some folks on the ubnt forum expressed concern with the PCB connection of the external RF connector. The USB device works well for walking around with a laptop looking for RF sources. While Insteon RF transmissions are a bit bursty for this device, it still provides more info about my local 900 MHz environment than I have without it (e.g., wife's 900 Mhz headphones seem to be far enough away so as not to cause problems with Insteon).