Digger: No, I bought replacements because it is a hassle dealing with them. I gave up quite a while ago - last year some time.
My favourite hassle was when the tech support guy wanted me to reinstall the broken switches so that he could "diagnose" them. I'd had spares from changed plans, so I'd originally just put the faulty ones aside. But no, he insisted that if I wanted them replaced, I'd have to go and take the working ones out, put the faulty ones back in and then do his checks. I don't recall the details, but one was that he wanted me to move the cover plate and see if that fixed it (!). It didn't matter that I'd tried that many times before, including unscrewing and reseating the paddle assembly etc itself. In each case I'd already discovered that pushing the microswitch with a flat-ended round plastic object showed that the micro-switch itself was the problem. (FWIW, a piece of acrylic rod)
I have had rev2.0 switches replaced that had an electrical failure though. There's nothing to complain about there. Its the bloody faulty paddle microswitches that are driving my wife (and therefore, me) crazy.
I'm guessing they used cheap switches and the contacts simply oxidize or get otherwise impaired somehow.
There doesn't seem to be a single cause, in my experience.
My summary:
Master bathroom: 2 failed -relays, about to install 3rd.
Kitchen 1: 2 failed relays, using 3rd.
Kitchen 2: just replaced a rev2.0 -relay that was almost 2 years old, but not quite.
Downstairs Office (adjacent to kitchen): -dimmer failure. Replaced with -relay and CFL's, it failed too. Now on 2nd -relay.
Entry: three simultaneous rev2.0 -dimmer electrical failures. (smarthome replaced these, replacements working perfectly)
Downstairs bathroom light: on third -relay. They seem to last 4-6 months max here. Very high traffic.
Downstairs bathroom fan: the rev2.0 -relay for the fan is dead and needs replacing. Rarely used.
Fountain: rev2.0 -relay paddle appeared to be dead, but wiggling the faceplate solved it for now. This device is in a closet and has ZERO use of the paddle. Its all remote control.
Master bedroom outside light: rev2.0 dimmer just today discovered with flakey 'on' paddle. Needs replacing.
Top of stairs: 2.0 -dimmer 'off' paddle flaked out. Replaced with 1.3 keypadlinc.
Master bedroom entry: keypadlinc 1.0 died (lights went out). Replaced with 1.3 keypad.
Wife's study closet: rev2.0 -dimmer paddle worked, but was "deaf" to insteon and X10.
Several smoked appliancelincv2's: replaced by new revisions, but they still won't work with their aquarium lights. (see below)
Notable survival stories:
Wife's study entry: rev 2.0 -relay. Gets lots of use. No sign of trouble.
Downstairs shower: rev 2.0 -dimmer. Massive humidity doses (no exhaust fan!). No sign of trouble.
There's a bunch of rev 2.0 devices still deployed in places that I've largely forgotten about but they dont give any trouble. Some of the switches that have had failed paddles are actually being used as wire-in appliancelinc or inlinelinc replacements. They generally work fine in that role - especially with the ability to remotely program them. eg: I embed some of the failed -relay modules inside the bodies of fluorescent lights and use them as self-contained remote controlled lights.
To generalize my problems:
- mostly related to -relay devices.
- bathrooms seem to be a good way to kill them, but the worst bathroom environment hasn't caused even the slightest problem
- high traffic areas seem to be a good way to kill them.
- zero traffic areas seem to be another good way to kill them - eg: not touching the switch for 12-18 months.
- appliancelincV2's still seem to suck. I use switchlincv2-relays with broken paddles to replace them.
I don't seem to have much trouble with devices in areas that get light but regular use.
On the plus side, I've never needed to buy any inlinelincV2's. I've still got a stockpile of unreliable switchlincV2 paddle devices available for use. I have more places to put them still.
In case anybody was wondering, no, I generally don't use switchlincV2-dimmer devices as inlinelinc replacements. I don't like the idea of their fins not being in contact with a heat sink. I do have two in this role though, but they're in a metal 3-gang box so there is plenty of heat sinking availability. (they're used as bedlight dimmers.. one "broken" -dimmer on each side and a 2-way power outlet in the center slot to plug the lights into. Plus, they're light loads, 40 or 60 watts. There should be no problem with heat.)
BTW: Ken: I wasn't really kidding about the 'revenge' thing. I've selected the first victims. The first ones are going to die by fire, sledge hammer and then electric drill. I recorded the video of them not working today, all ready for fun on the weekend. I do like the idea of sending some of the faulty ones back to smarthome - but carefully packed in horse or cow manure. Somebody else on AccessHA seemed to think that was a good idea. I doubt that I'll be doing that one though - I could imagine that one landing me in real trouble.
BTW2: The access point thing pissed me off too. I know there are issues with the signalinc-RF (can't say more due to damn sdk NDA), but new, apparently incompatible devices? WTF is up with that? The one good thing about the developer support program is that they don't tell us squat anymore, so there isn't anything to leak. At least you can get most of the docs to make a PLM work from public sources without spending $200. (A tip the the audience.. Instead of buying the SDK, get the $200 out of the ATM in notes and set it on fire. You'll get much more enjoyment from that and IMHO about the same amount of developer help.) What really bites is that the access points don't have a passthrough socket. And I don't particularly want any of those butt-ugly remotelincs so the whole trade-in program is a bust.
Mutter.. grumble.. whine.. rant.. Smarthome's botched handling of the whole Insteon thing just makes me angry. Insteon had so much promise, but when it came to the crunch, it seems like they had way too little quality control and way too much cost-cutting. I don't know what the heck their beta testing program does, but it doesn't seem to find the glaringly obvious problems. And don't get me started on what I think of the amount of planning and forethought that went into their computer interface to the system. (If they'd written houselinc and/or sdm while designing and implementing the system they'd have discovered how badly they botched it before it was too late to fix. They'd have discovered that extended messages didn't work well before shipping the first products.) There's no excuse for a network scan taking hours. If extended messaging worked and/or the powerlincV2 didn't suck so badly, a network scan should take take 30 seconds up to a minute or two at the most. I mean, WTF? The insteon promo material mentions 5ms to send a message. However, and 9600 baud to talk to a powerlincV2 means a minimum of 2ms per character. A 9 or 23 byte packet, plus overheads, plus the silly 20ms character pair delay mean that a computer initiated command takes upwards of 400ms+. There's no excuse for not designing the PLCv2 so that it could get computer commands out onto the wire at near line speed. It wouldn't have cost that much more to put enough cpu power in it to make it work right. The PLM seems to be a significant improvement so far though, but it still seems too slow.
Grrr. I have to stop before my blood pressure spikes again. That's why I had to pull out of the SDK program in the beginning of last year. And I was having such a good day today till I started replying to this and getting angry all over again.