Saturday, October 7, 2017


Attention!  Debbie is home, safe and sound...

Paradise ponders: discombobulation edition...

Paradise ponders: discombobulation edition...  Well, of course everything went wrong with Debbie's flights home yesterday.  It's a long and uninteresting story, but the end result is that she's booked onto new flights today.  In fact, as I write this, she's in the air from Cincinnati to Chicago – and amazingly enough, that flight is looking like it will arrive early.  The poor lass ended up staying in a hotel last night, right next to the Cincinnati Airport.  Assuming her connection in Chicago works out all right, she should be on a flight to Salt Lake City in about 90 minutes, and arriving here at 11:15 am.  Yay!  In my last texts with her, though, I discovered that she's feeling like she's got a cold coming on.  Dang it!

With Debbie's change of schedule, our plans for yesterday got blown up.  I did get a coat of paint on the new trim work and siding where the new deck was attached to the house.  It already looks a lot better, though it needs at least one more coat of paint for completely coverage (I'm painting white over a dark primer).  I was also able to make progress on what has been an unexpectedly complicated undertaking.

We created a beautiful fenced back yard for our dogs about a year ago.  This back yard has five gates, three of which are not visible from the back door or deck.  On multiple occasions we've had people leave a gate open, including a couple of times after I put the nice little signs up asking people to keep the gates closed.  This is worrisome for us, as we're afraid that someday a gate will be open, we won't know it, and we let our dogs out.  They'd discover that open gate within seconds, and then they'd likely scatter to the winds – we have, unfortunately, seen this action before.

So I got to wondering if there was a way to put sensors on the gates and have an indicator of some kind in the house to let us know if they were open.  Wireless (radio) sensors of some kind would be the easiest to install, by far – running wires to all those gates would be a serious pain in the patoot.  So I started researching that, and quickly ran into wireless gate sensors (made by Honeywell) that were designed to be part of an alarm system.  I started checking that out, and I found online discussions about using a product called “Tuxedo” with them.  Some of those discussions mentioned how easy this was.  I liked that!  So I ordered five gate sensors and that Tuxedo.

When I received them, I unboxed the Tuxedo and a sensor with the intent of testing things out to see if they worked ok.  I quickly ran into a mystery: the Tuxedo had no power supply, or any obvious place to connect one.  WTF?  The only electrical connection to it was a 4 position terminal block.  I was forced to go read the manual – and in doing so, I discovered that the Tuxedo was actually just a peripheral unit (“keypad” in alarm-speak) for an alarm panel.  Said alarm panel was required in order to operate the Tuxedo.  Sheesh!  So after some research I located a relatively inexpensive alarm panel, the Vista 21IP.  I ordered it, all the while thinking to myself that I was being sucked into some kind of money pit.  I just want to know if my gates are all closed!

When the panel arrived, I sat down to do that testing again.  This time, the Tuxedo had power (after I wired it to the panel), and it booted up and started working.  The reviews rave about how easy the Tuxedo is to use, as a method of programming the alarm panel.  Something I failed to understand until I actually used it is that those reviews are rating it's usability in relative terms.  By any modern user interface standards, the Tuxedo's user interface is somewhere between horrible and criminal.  However, compared with the alternative – using the “standard” arcane and tortuous numeric code programming – it is wonderful.  The standard scheme looks like something out of the '50s, or transcribed from a medieval torture manual.  So I suffered through enough stuff (Googling often for help) to prove that the Tuxedo was actually working.

So at that point I broke out one of the gate sensors to see if they were working.  Here a misunderstanding on my part came up and slapped me in the face.  The discussions I'd read online about the Tuxedo raved about its “Z-Wave” (a wireless standard) capabilities, in particular how simple it was to make them work.  That's the part I really liked reading.  Those same mentioned how easy it was to make the Tuxedo work with wireless sensors, and mentioned the gate sensors.  So I assumed that the gate sensors were Z-Wave sensors.  They are not.  They use a completely different wireless standard, not even on the same frequencies as Z-Wave.  That means the Z-Wave receiver built into the Tuxedo can't receive the radio transmissions from the gate sensors I bought.  After a lot of research, surprisingly difficult in the Google era, I finally figured it out.  I needed another RF receiver, one that is a peripheral to the alarm panel.  The money pit vortex strikes again.  I ordered one of these.

Meanwhile, I tracked down two annoying “faults” that kept popping up on the Tuxedo.  Both of them, it turns out, were actually faults within the alarm panel being reported on the Tuxedo.  One of them was a “103 LNGRNG RADIO” error, and I figured out that this was related to the cell phone interface built into the alarm panel.  That interface is intended for reporting back to some central monitoring company (like ADT).  Of course I had no interest in that, since I'm not using this as a burglar alarm.  With some additional research, I discovered that if I entered a magic series of digits (“411280056**00*”  - I kid you not) on the emulated keypad, I could disable the cell phone completely.  I tried it, and (it's magic!) it worked.  One problem down.  The other error was the panel complaining about a low battery voltage – but there was no battery.  WTF?  Turns out that the panel requires a rechargeable battery for power backup.  So far as I could tell, there was no way to disable this requirement.  Naturally, the battery in question is a non-standard one, so I had to order one of those as well.

Yesterday the RF receiver arrived.  I connected it with some trepidation, wondering what new part was going to be required in order for this to work.  Once I had it connected, I researched how to “enroll” a gate sensor (all that means is to pair the gate sensor with the panel).  I expected this to be something like Bluetooth pairing, but oh, no, it's nothing at all like that.  Pairing on of these involves manually entering the gate sensor's serial number.  That involves strings of numbers that makes the disabling of the cell phone look like child's play.  It took me a half hour of frustrating interaction with what is surely a candidate for the worst user interface ever invented to get the first gate sensor enrolled.  But then, the good news: it actually works!  In under a second, any change of state on the gate sensor (open-to-closed or closed-to-open) gets reported back to the alarm panel and then to the Tuxedo.  Holy crap, this is actually going to work!

Now that my test is done, I have to wire all this stuff in.  The Tuxedo will go on the wall next to our back door, so it's convenient to check before we let the doggers out.  The RF receiver (quite small) will likely go on the wall above it.  Those two things can easily be wired together.  The giant, utilitarian-ugly alarm panel will go in our mechanical room, in the basement.  That means I get to run a four-conductor cable between the basement and our kitchen, while keeping the wiring completely inside the walls in the house.  That's gonna be a bit tricky, and I'm still working out how exactly to accomplish that.  The battery should arrive with a couple weeks, and by then I should have it all hooked up...