UPB support, please!

[AutomatedOutlet]

[rocco]

An honest mistake or a baldface lie?

Either somebody at SimplyAutomated is mathematically challenged, or they like false advertising.

It takes 2 AC cycles to transmit on byte over UPB. The minimum packet size is 7 bytes, or 14 cycles (not counting the Acknowledge). So the minimum packet would take .233 seconds. The maximum packet (25 bytes) would take .833 seconds.

So there is no way in Hoboken that they could get 20 to 40 times the speed. If all packets were minimum, the best they could do is 3 times.

[zoomkat]

The below is from the http://www.go-upb.com/ site. The bps rate could probably be almost anything you want depending on the capability of the hardware implementation.

X-10 speed - 60 bps
UPB speed - 480 bps

[Rupp]

What's the big deal with the speed guys. It's still faster than getting off the couch and pushing the button.

[George]

Rupp,

Lets use the following scenerio:

You've having a party with the people you work with that have heard about your really cool automated smart home. You want to show them how the lights work. You press a button to set a new scene (638x modules) that has about 15 lights in the scene to setup a party lighting mode for your home. You press the button and people are just standing around waiting for something to happen. You know that HS is sending all the device address codes before the command and understand but all the people hanging around are just wondering what's going on and why nothing is happening with all the thousands of dollars you've spent on your home.

Even a conservative 4 times improvement in transmit speed takes 8 seconds of thumb twiddling and turns it into a hardly noticable 2 seconds.

That's one of the big deals.

George

[Rupp]

George,
This is exactly why I choose Z-Wave over a powerline protocol. But on the other hand I do not use scenes and never really understood the need for them. I don't have enough lights to create different scenes.

[rocco]

UPB speed - 480 bps

This number is flat out wrong, although I'm sure it was an honest mistake. The real number is 240 bps.

The mistake comes from the fact that UPB issues one of four strategically placed pulses per AC half-cycle. With 120 half-cycles per second, it's easy to conclude that would be 480 pulses per second.

But only one pulse per half-cycle is allowed, so those four possible pulses actually encode two bit. Two bits per half-cycle, times 120 half-cycles per second gives the true data rate of 240 bps.

Sorry, Rupp, I don't mean to harp on speed. I really don't think the speed is that important, but accuracy in advertising is.

[bdickhaus]

A z-wave scene can happen in less than a second, why bother with powerline? I really do not get it, there are now several RF based and hardwired control based solutions that are far, far superior to powerline based. To me the powerline communications is old technology and so subject to noise from outside interference that can't be easily controlled or dealt with that it is unusable. I really do not understand why anyone continues to go down that path, it has been a dead end for years, it doesn't make sense to go there.

I support Rick in what he said, why should HST waste their valuable resources on a deadend technology? They should be looking ahead (like with z-wave) not spending resources to support something that will die out like all the other powerline technologies. Rich gave the politically correct answer, but I think he's wrong, they shouldn't waste their time on UPB.

Bill

[electron]

Considering several hardware manufactures are already supporting UPB, I would hardly call it a waste of time or a dead protocol. Each protocol has its place (I know plenty of people where RF just won't work in their house), and with ZigBee growing really fast, and Insteon being released by SmartHome, I wouldn't count those protocols out just yet, nor would I place all my bets on Z-wave (look at all the problems they are having). As much as you would like want it, powerline based protocols are far from dead (just look at Broadband over Powerline).

[George]

I think we are are looking at from a rather small perspective. As long as HST wants to remain a 'home automation' package then I tend to agree that X10 and ZWave are both valid solutions. But, for commercial applications both X10 and ZWave have serious and mostly unacceptable flaws. UPB goes very far in addressing these shortcoming for commercial applications. Can anyone really see outfitting a 10 story building with X10? Same question for ZWave. Big problems in both addressing as well as signal issues.

I think Rich is looking a bit farther out then we, myopically, are. If HST is going to be a player in 'Automation' and not just 'Home Automation' then I think it is essential that technologies such as UPB be supported even though there may be something else in the future. If we were all to wait for the next better thing before we bought something we'd all still have B&W TV's.

Regardless, it's never, repeat never, the vendor that determines what's a dead end technology. It's the end user and customers that do that. If a vendor decides to not support something that customers are buying then guess what, that vendor will not be receiving any money from that customer base. Simple, business economics 101.

[AutomatedOutlet]

George,

As far as your senario goes, for scene controls within HS, you are right, powerline scenes can take a little longer. BUT, if you use the scene controls built into the devices, they are very fast.

Take a look at Lightolier scene controls. The intellegence is in the switches. Once the command is given (from either a Lightolier keypad, or from HS using the Compose plug-in), the scene change is not only fast, but all devices respond at once - no popcorn effect.

Smarthome switches also have inherant scene capabilities as well.


Bill,

I disagree. A properly installed X10 system is very reliable. In addition, noise isn't an issue at all with UPB and no noise filters are required due to the nature of the signalling.

-Martin

[Rupp]

Martin,
Are you carring any UPB devices?

[electron]

http://www.automatedoutlet.com/home.php?cat=128 (2 pages it looks like)

[Guest]

Polling no longer needed?

Some of the discussions here are based on the fact that Z-wave needs to be polled. If I understood it correctly, with the newest version of the protocol, the USB-controller can be assigned to a switch as a secondary node, meaning that when the button is pressed (local control), HomeSeer is immediately notified. No need for polling anymore.

Can't jump in on the UPB discussion though, never used any UPB products.

Richard

[AutomatedOutlet]

Richard, I may be wrong but I don't believe that's true. The 2-way issue is not a technology issue, it's a patent issue with Lutron that doesn't allow 2-way communication of RF based light switches.