Hardware Feature comparison?

[kwilcox]

That was me who noted the UPS degredation issue. Prior to installing my UPS, I didn't need a UPB phase coupler. Unfortunately, with the UPS plugged into the same plug as my PIM was, I found that I could no longer control UPB devices on the opposite phase. The coupler corrected the problem and I've had no issues since. As a note, UPB installer guidlines always call for the installation of a phase coupler. People who decide to install UPB should make sure they do this as part of the initial project that includes installing a PIM and getting UPStart running.

Interestingly, when I installed the UPB coupler into the same box as my ACT X10 phase coupler (seemed to make sense as this oversize box was right next to the breaker panel) the ACT device started erroring out on most repeated X10 signal attempts. I ended up installing it on the dryer circuit and that corrected the problem.

Also, I misspoke earlier about an aspect of my 1130W switches and that was where I claimed that the second tap of a multiple tap sequence immediately invokes double-tap programming. It doesn't. I was playing again last night, trying to acually see how many times I could tap the rocker before the light activated and I immediately noticed that anything more than 2 taps caused the switch to ignore the input and do nothing at all. Perhaps its becuase the switch was waiting for a 5-tap sequence to initiate programming mode.

 

[PeterW]

BTW: I got some HAI UPB switches from Smarthome.com and the switch bodies are the same size as the older switchlinc's.. if not slightly smaller. They fit in my J-boxes! Score!! But, it just seems so wrong getting UPB stuff from smarthome. (They have Z-wave stuff too FWIW)

Also smarthome has got a really small UPB phase coupler now.

It seems smarthome have HAI and webmtn upb products in addition to the SA ones.. Automatedoutlet? The HAI 6-button and 8-button wall controllers etc fill a void - automatedoutlet's product pages still list the SA keypad faceplates as due mid-june.

BTW, while I think about it.. the UPB passive phase couplers are just another capacitor system, right? Is there any chance that they'll passively couple X10/insteon as well?

 

[PeterW]

Hmm. Well, the "small" SA UPB phase coupler is a fused 12uF capacitor. I've been doing some digging around, and it seems that 0.1uF is the preferred X10 coupler capacitance value. Hmm.

Actually, the X10 passive coupler folks seem to like to use a tuned R-L-C circuit to create a band pass filter to provide an upper and lower bound of the frequencies that they pass from one phase to the other.

If my analog electronics memory from oh-so-many years ago serves me correctly, the capacitor is only acting as a high pass filter. The higher the capacitance, the lower the frequency threshold is that it'll allow past. So, the 12uF capacitor should allow the X10 and Insteon frequencies past, as well as much lower frequencies.

Since the UPB signalling system is pulses rather than frequency carrier, I'm guessing that the capacitor is there to carry as many of the harmonics of the pulses across as possible so that the devices on the other phase see something resembling a pulse.

Even more amusing is another thought. I currently have my smarthome signalinc coupler-repeater and the old UPB phase coupler on the same circuit, about 4 inches of wire from each other. This can't be good. I wonder if the UPB passive coupler is leaking some of the repeated signal back into the original phase? That is, assuming that the j-box mounted couplers don't have an inductor to stop them coupling X10 frequencies across..

 

[upstatemike]

I was re-reading this topic from last month because I am trying to come up with a strategy to migrate from my existing X-10 environment to either UPB or Insteon. I have around a hundred X-10 devices; mostly 1-way modules and switches with maybe a dozen 2-way or transmitter devices.

I have to do this over 18-24 months due to cost and time constraints and it looks like that rules out Insteon completely! It appears that as sonn as I replace a dozen or so 1-way X-10 devices with Insteon that the rest of my X-10 system will lose what little reliability it has and force me to change the remaining modules all at once (which I can't do).

Has anybody migrated a large X-10 installation to UPB? Are there any "signal suck" issues with UPB at all when you get a lot of them on line? Any issue in ganging several together? ( I have some 5 and 6 gang boxes with 1-way Leviton relay switches in them now).

Any "large scale" UPB intallation feedback would be very much appreciated.

 

[fitzpatri8]

I have to do this over 18-24 months due to cost and time constraints and it looks like that rules out Insteon completely! It appears that as sonn as I replace a dozen or so 1-way X-10 devices with Insteon that the rest of my X-10 system will lose what little reliability it has and force me to change the remaining modules all at once (which I can't do).

Now granted, X10 sucks, and 2-way X10 sucks even more. Still, I don't know that I agree with your logic. If you've already got over a hundred devices, you've already sunk considerable cash into HA, and you've already had to deal with signal suck issues. Converting to any other technology is going to cost you, no matter how you do it.

You also have a bit invested in X10 remote controllers, either wireless or plug-in. If you want to upgrade your system gradually, and if you want to avoid throwing out that investment in X10 remotes while upgrading to newer technology, you need to keep the X10 portion of your system reliable so that the remotes can still make things work.

Insteon draws down X10 signals the same way as 2-way X10, but there is no limit to how many 2-way devices you can have in a house.

Signal losses due to 2-way devices can easily be restored by adding a coupler/repeater and/or a booster link on long branch circuits. Neither is expensive, and both can be resold to X10ers on Ebay once you complete the transition.

If you upgrade a circuit at a time, changing over the most far-flung devices from the circuit breaker first, there's no reason you can't overcome the little extra signal suck with minimal hardware. And the Insteon devices can respond directly to your existing X10 signals, so you don't have to keep two remotes everywhere, one for the new stuff and one for the old, or, alternatively, using a computer and two seperate interfaces to receive X10 and transmit another flavor of signal.

A single Insteon device can respond to both an x10 signal and an Insteon one, so as you add new remote controls you can use both the X10 and Insteon remotes to control the same devices without having to set up a computer to intervene.

Tom

 

[upstatemike]

Signal losses due to 2-way devices can easily be restored by adding a coupler/repeater and/or a booster link on long branch circuits. Neither is expensive, and both can be resold to X10ers on Ebay once you complete the transition.

My logic is based on the fact that I already have a lot of coupler/repeaters and boosterlincs just to keep the 1-way stuff working. I have lots of TVs, PCs and UPS's sucking the signal down already so I'm reluctant to add more 2-way stuff. I also have a lot of ganged switches to deal with.

On the other hand I really like the Paddle/ LED options of Insteon so I'm not giving up on it just yet. I appreciate the input as I continue to consider my options.

 

[toymaster458]

The best I can say is I had one location where I installed 4 2-way SwitchLinc Version 1 switches. I could not get them working properly and ended up replacing one of them with a SwitchLinc with Booster and all was well. I recently replaced all 4 switches with the INSTEON version and was able to address them all with X10 with out the need of a boosterlinc.

I am not saying that they will not effect your current X10 signals but I am starting to feel that the new switches do not suck as much X10 signal as the old ones even though Smarthome says they suck the same amount of signal as the old ones.

I now have about 20 INSTEON switches installed in my house with NO BoosterLincs and X10 signals are still working properly.

 

[Dean Roddey]

I've been using Z-Wave, and I'd been having a fair amount of reliability problems and latency problems and such. In my case, since I'm using my own product and I wrote the driver, it was kind of my own fault I guess. I really wasn't able to get much help from Zen-Sys on driver development and have been left to my own devices mostly.

But, after many iterations and explorations, the CQC Z-Wave driver that's going out in 1.4 (very soon now), is really rocking. It's completely solid, with almost instant reactions to user invoked events, and very reasonably low latency keep CQC updated on module status, at least within my small (6 modules now) network. And no errors, either errors enough to show up to the user, or even errors that I retry and recover from. It's been spooky actually.

I just made the key couple of breakthroughs over the last two days. I've been playing with it today and I'm very happy now. I've still got to let the beta testers at it and see it goes for them, but for so far the results are extremely gratifying.

I do have a small apartment, and the controller node is about in the middle of the apartment, so it has good coverage. But, during testing I had it back in the bedroom so it was all the way on one end of the apartment and it was just as steady. There was just slightly more latency on the modules at the other end of the apartment. The spread of modules is pretty good, so there should be good routes always to any node.

Anyway, I just wanted to point that out, since I've done some amount of bad-mouthing of Z-Wave in the depths of my utter frustration over the last few months (and I still am not terribly happy with the company.) But now I think the technology itself for me anyway, and with our product since I can't speak for others, has lived up to its promise finally.

I can't say how it would do in a considerably larger system. One of our new customers has 16 or 20 or so modules so I'm going to be very interested in his results with this new version of the driver.

 

[kwilcox]

@upstatemike:

I'm slowly doing exactly what you want to do. I also have a "delicately tuned" X10 setup with strategic boosterlincs everywhere and am well aware of your trepidation regarding making any changes to the "critical balance" of this tuning. However, I've been installing UPB a room at a time and there are no issues at all between the two (other than my X10 reliability going up as I remove 2-way devices). UPB is noticably faster than X10 too. The only unusual interaction I saw was my UPB coupler causing error conditions on my APC active coupler when both were installed in the same junction box. Moving the UPB coupler to the junction box for the dryer outlet fixed this problem.

 

[upstatemike]

Wow great feedback, thanks to all!! It looks like UPB is pretty safe and will scale well to larger installations. It also looks like the Insteon issues are more manageable than they originally appeared. I wasn't really considering Z-Wave but I will be curious to know how things work out for Dean's customer installing 20 Z-Wave switches.

Maybe a good question to ask is who has, or has seen, the largest installation of non-X10 switches and modules so far, whether it be UPB, Insteon, Z-Wave or whatever? (And how well was it working?)

I also have a lot of X-10 Screw-In Dimmable lamp modules. Does any technology currently have a replacement for those?

 

[Dean Roddey]

I wasn't really considering Z-Wave but I will be curious to know how things work out for Dean's customer installing 20 Z-Wave switches.

Keep in mind that it'll depend a lot on the automation system you use with it. The problem, it seems to me now, isn't the technology itself, but that Zen-Sys has not really invested anything at all in insuring that companies like ours can get the info we need to write high quality drivers for it. In fact, they expect us to pay them to get what info they do provide (and a pretty good amount of money at that.) They've really had an arrogant attitude towards people like us who are working our butts off to help them expand their market.

I had to figure it out by long trial and error, and I'm as experineced in this area as anyone you'd likely run into. Until now, I've assumed (and probably others who have tried to use Z-Wave with products such as ours or Homeseer) that Z-Wave itself is just semi-flakey. I've read through the HS support forum and seen plenty of problems with Z-Wave there, probably for the same reasons we had problems.

Now that I've finally nailed the driver, my opinion of it has turned completely around, for the technology at least, not the company.

 

[upstatemike]

It's good to hear that the technology is sound and I'll keep a cautiously open mind with respect to Z-Wave. I just want to see how it works in a really large installation. In theory more is better in an RF mesh network but I'd still like hear about somebody who has actually done a large install.

 

[AutomatedOutlet]

Lots of good stuff in this thread. I'll go over a bunch of them in no particular order.

Maybe a good question to ask is who has, or has seen, the largest installation of non-X10 switches and modules so far, whether it be UPB, Insteon, Z-Wave or whatever? (And how well was it working?)

Most of our professional installers are all going either with UPB or Lightolier Compose. I would say over 80% are going UPB. We do have a few that have used X10 or Zwave but the number is dwindling.

We have a bunch of UPB integrators that routinely do installations of UPB with over 150 devices in 10,000+ sf houses. If you are doing houses this size I think you either need UPB of Compose. Installers using Zwave run into issues with larger houses because of polling issues, lack of devices, and also the fact that if the end customer moves a lamp module (yes, it happens), the entire network gets messed up.

Has anybody migrated a large X-10 installation to UPB? Are there any "signal suck" issues with UPB at all when you get a lot of them on line? Any issue in ganging several together? ( I have some 5 and 6 gang boxes with 1-way Leviton relay switches in them now).

Excellant question. Since I consider myself my biggest guinea pig, I have done whole-sale swap outs in my own house several times over the past few years. Right now, Most of my house is 2-way switchlincs, some Ligholiter, and a few random things. Over the past few weeks I have been pulling out some of my older 1-way switches and replacing them with Insteon switches. Of course no software supports Insteon so I've been running them in X10 mode.

Yes, they do such X10 signals. I haven't tested how much but it I'm positive that it's at least as much as regular 2-way switches if not more.

In any case, after doing some testing and thinking about it a lot, I think I'm going to migrate my whole whole to UPB. I probably will still retain some X10 devices like wall outlets, motion sensors, etc.

With UPB, as I migrate devices, my X10 signal gets stronger and clearer some I really like a blen is the right answer, at least for me.

If you look at the pricing of Insteon vs. UPB, UPB (believe it or not) is probably a little cheaper once you factor in the fact that for 3-ways you don't have a slave switch available with Insteon. Yeah, the UPB dimmers are $10 more, but the slaves are about $33 less......

I think that the key to any migration is to make the system better and more reliable as you go. Whatever devices you move to have to be suppported by software that is available on the market.

To date, all software manufacturers that I am aware of either currently support, of have announced they will support UPB. So far, some software manufacturers have said that they are looking into supporting Insteon but none have given a date. In fact, you software manufacturer, HAI, has said that they are not even considering supporting Insteon.

Perhaps things will change. If anyone wants a deal on some 2 week old, slightly used, Insteon switches, please let me know. I'm switching to UPB.

For those of you on X10 now, Please don't take my posting the wrong way. I think X10 works great. I have been very happy with my X10 2-way devices for years and will continue to have X10 devices in my house and office building. I will gradually change the stuff out because that's what I like to do. It gives me a chance to get to know a new technology better and be able to better understand it. I like UPB because of it's speed and reliability. I'll sure miss those LEDs though of the Switchlincs but will enjoy the tri-color status LED on the UPB switches and the increased speed.

 

[ChrisWalker]

Okay, here's some real-life "big" Z-Wave experience. I hope this adds significant value to this thread.

Back in the day, Z-Wave experienced its growing pains. There were reliability issues, and there were a small number of devices on the market.

These days, that has all changed and Z-Wave is the _only_ technology we've found reliable enough for large installations. And today there are finally enough devices on the market to complete full homes, including thermostats from two vendors.

I head up design and engineering for a company, ControlThink, which is a Z-Wave Alliance Member. We have worked hard to help Zensys fix the few issues that the Z-Wave stack has had, and things have become very reliable. Even with the first switches and devices which came out years ago (sans the internal fuse issues which were remedied long ago, hardware defects, etc.) And we really beat things up around here. We test things to their limits, and we continue to work closely with Zensys to improve the protocol further. We built advanced self-healing (so you can move a switch, and the network fixes itself) into our protocol stacks, at significant expense.

We even have a home with over 200 Z-Wave switches, a half dozen thermostats, and 15 plug-in modules, etc. We're created technology to bridge multiple Z-Wave networks together, for homes with more than 232 devices or that are phsically too large ("hop"-wise) for the Z-Wave protocol. We're working with installations where a single Z-Wave network is spanning three buildings, reliably. And then we're briding Z-Wave networks via IP elsewhere, and have created redundant, fault-tolerant Z-Wave systems. It's pretty awesome, and it all uses the same Z-Wave hardware everyone else buys on the open market.

Our engineering comes from a long background of developing software and hardware solutions for high-reliability applications (like warehousing and manufacturing systems, military, government, and medical applications, and various automation systems). We spent man _years_ of time developing our Z-Wave implementation, and even we are happy with the reliability (which approaches 100%).

BTW, if anyone would like to play with a reliable Z-Wave implementation that runs from a PC (note: I'm not promising that the PC itself is highly-reliable :huh: ), e-mail me through the site here and I'll be more than happy to re-flash an ACT USB controller for you with our special Z-Wave build (which, combined with our software, fixes most all of those nasty existing polling issues) and send you some software which will show you how easy and reliable Z-Wave really is. We even have a "reliability test tool" which we designed for internal testing and for other Z-Wave Alliance members to use. I'm not sure that we can let that out (because then we'd have to officially support it), but if there are enough requests I can see what we can do.

Additionally, we are doing full-scale analysis of live Z-Wave networks (both large and small) and will be publishing those statistics later this year. At this point, I know of no other technology that provides the affordability, reliability, and scalability of Z-Wave for residential (and even commercial) control systems. And yes, we do continue to test other protocols and hardware.

Chris

 

[electron]

That's some great information Chris, and thanks for the offer! Hopefully this means that there is hope for Z-wave now that some of the basic issues have been taken care of.