Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[JeffVolp]

I noticed today that the supply of TW523s on eBay has dried up, and only one PSC05 is listed at a very high price.  SmartHome lists the TW523 as discontinued.

As many of you know, the TW523/PSC05 is the X10 powerline interface used by many OEM products.  Some have been discontinued, such as HomeVision and Stargate.  But others like the WGL transceivers and Rain8 Irrigation controllers are still in production.

I wonder if the TW523 will be produced again to support these products or whether I should offer something to fill the gap.

Jeff

[Brian H]

One of the on line X10 dealers has an ever growing set of web pages. Dedicated to Discontinued X10 devices and urging you to voice your displeasure to X10 Wireless Inc.
The PSC05/TW523 and PSC04/PL513 are both on the list.

I believe Smarthome also has discontinued their 1132B that could do TW523 functions.

I just looked at the WGL site and all of their TW523 interfaced devices. Have a bold warning to verify you can get the PSC05/TW523 before ordering. They said after 01/15/2012 the PSC05/TW523 could be in short supply.

Hard to say if an alternate would be that popular.

[JeffVolp]

Thanks for the info on that being on the discontinued list.

The XTB-IIR also provdes that interface, but a smaller plug-in alternative might be worth the effort.

Jeff

[glacier991]

I think I agree with you Jeff.

[dhouston]

Will it, at least as an option, report all powerline activity in realtime?

[JeffVolp]

Will it, at least as an option, report all powerline activity in realtime?

That is a mode option now with the XTB-IIR.  It can output data to the digital port just 1/2 cycle delayed from the powerline input.  That capability can also be included as a mode option in a smaller plug-in unit.  Obviously, that data cannot be error checked to insure it is indeed a valid X10 command.

Jeff

[dhouston]

Will it, at least as an option, report all powerline activity in realtime?

That is a mode option now with the XTB-IIR.  It can output data to the digital port just 1/2 cycle delayed from the powerline input.  That capability can also be included as a mode option in a smaller plug-in unit.  Obviously, that data cannot be error checked to insure it is indeed a valid X10 command.

And cannot be used to avoid collisions if delayed 1/2 cycle.

[JeffVolp]

[And cannot be used to avoid collisions if delayed 1/2 cycle.

I may be good, but I can't anticipate when a bit is going to come.  Per the X10 Spec, the data output must be asserted within 200uS after the zero crossing, but it takes 48 cycles of 120KHz before it can be recognized as a logic "1".  Assuming the burst starts 100uS after the zero crossing (the X10 spec allows 150uS), that is at the mid point of the 1mS transmission window.  If the data output is asserted at that point, devices that sample earlier in the cycle would miss it.  So the data is output at the next zero crossing.  I don't know how to do any better than that.

If you are concerned about collisions, all XTB devices except the original XTB can automatically abort a transmission if they detect a "1" when they should be transmitting a zero.  And in that case there is no echo back to the controller, which informs it that the command was not transmitted.

Jeff

[dhouston]

[And cannot be used to avoid collisions if delayed 1/2 cycle.

Per the X10 Spec...

I'm familiar with the spec but others (e.g. Smarthome's 1132B) reported in realtime. I assume they used other methods (e.g. envelope detector) for earlier detection of a signal.

If you are concerned about collisions, all XTB devices except the original XTB can automatically abort a transmission if they detect a "1" when they should be transmitting a zero.  And in that case there is no echo back to the controller, which informs it that the command was not transmitted.

But the collision has already happened at that point with a high probability that you've already sent a "1" when the other source sent a "0".

[JeffVolp]

I'm familiar with the spec but others (e.g. Smarthome's 1132B) reported in realtime. I assume they used other methods (e.g. envelope detector) for earlier detection of a signal.

Or perhaps they don't wait for 48 cycles to determine a logic "1".  Sure, you can do that, but it would make it more sensitive to noise.  I chose to honor the X10 spec.  Even at that, the original XTB-II did not send out data early enough for some Stargate customers until I advanced digital output to be asserted almost simultaneous with the zero crossing.  At that point the 120KHz burst has not even started.

But the collision has already happened at that point with a high probability that you've already sent a "1" when the other source sent a "0".

Another transmitter can only be detected when it is sending a "1" while the local transmitter is sending a "0".  If collision detection is enabled in the XTB microcontrollers, it will abort the transmission at that point.

Jeff