XTB II Review

I want to thank Bruce for his wonderful review. Thanks also to everyone else who provided me feedback on the XTB-II.

There was a recent firmware enhancement that added mode options to best configure it for each installation. I'm working on a few additional options to deal with some special cases. These additional features can be added to any existing XTB-II in the future with just a PIC swap.

To everyone who has asked for a 240V 50Hz version, the firmware has been modified, and units should be available as soon as I receive feedback on how well that unit works in an actual installation.

Thanks to everyone who has supported me in this XTB project. While the XTB-II is probably as far as I can go down this road, I hope to offer additional devices in the future to address specific issues.

Have a great new year!

Jeff

My vote would be an RF signal strength meter with some sort of display that reads X-10 signal strength on an LED bargraph. Now that the XTB has solved the signal strength problem, I need a solution to the all-too-frequent "stuck" button problem. It happened again today. Although I ran it to ground as soon as I applied deductive reasoning instead of rote troubleshooting, I just *hate* wasting time that way.

A little meter with some sort of directional antenna would do the trick. Wouldn't need a bargraph led, either, just one that blinked faster with stronger signals. I can't be the only X-10 user in the world that's been plagued by "stuck" buttons. (I'm using quotes to try to conceal that it's not a fault in the equipment, but in the operator. The majority of stuckees have been me mislaying a Palmpad and something else on top of it.)

Today, following my "X-10 Protocol Disaster Document" that I was writing mostly for my wife but which has become quite useful for me and my senility, I:

1) plugged in the Monterey, saw an endless stream of BSCs (but very POWERFUL ones, thanks to the XTB!) 2) Began unplugging TM751's one by one. This sometimes gives directional cues as to the location of the transmitter since any one RF transmission probably *won't* reach all of my TM's. No such luck. No location seemed to be revealed. (It turned out that the stuck transmitter happened to be about equidistant from each TM-751.) 3) When I unplugged the next-to-the-last TM-751, I began to see a steady stream of B6 OFF signals 4) I then wasted a lot of time looking at all the switches that control B6 - the dog room. It turned out to be just a randomly stuck button in the long run, but the clues pointed elsewhere so I followed them. In circles. (-: No stuckees. I even hauled out the AM radio. Nada. Zip. Zilch. 5) I began to curse X-10, since I now had to disable all RF transceivers to restore PLC functionality. Even though the transmitter was sending just B commands, it was also jamming all other housecodes. 6) Then I though of F1 racing superstar Michael Schumacher, who never wastes a moment on anger and instead turns that energy toward solving the problem, I considered the human factor. Where was I between the last known good X-10 command and the failed one? Ah! I was taking care of a doggie disaster!!! Little Rocket began leaking rocket fuel and I scrambled to get a clean-up kit. I had started moving stuff around on the bench and had inadvertently put something on top of a Palmpad that I had been replacing the batteries in.

Interestingly enough, I discovered that the Control Linc Maxi's default to a "polite" mode and would not transmit while the TM-751's were picking up the stuck key. I believe there's a setting for that which I will change. If I used a lot of macros, I would leave it at "polite" but I would rather have a local PLC controller be able to "blast" through any interference and operate the lights in that room even if there was noise on the line. When something like this happens at night, it's more than a little annoying.

Though I hate to admit it, this is at least the 6 or 7th "stuck key" problem. (And before the "X-10 nah nah" crew cuts in, let me remind them that ANY control system is vulnerable to something leaning on a control button!)

Build it for less than $50 and I'll vote for you as Home Automation Man of the Year for both 2006 AND 2007!

-- Bobby G.

Google on "RF field strength meter".

snipped-for-privacy@yahoogroups.com

Interesting suggestion that wouldn't be hard to do using a receiver with a signal level output. There may not be enough demand for something like to justify the effort for a finished product. However, it could be an easy do-it-yourself project.

I wonder if one of those RF sniffers used to find bugs and other RF sources would work?

Jeff

What could be a more basic specific issue for X10 transmission than phase!

How about a phase detector: A plug-in transmitter module sends a short burst of RF on the positive AC half-cycle. A plug-in receiver module displays a green or red LED depending on whether the RF is received on its positive or negative AC half cycle.

These modules could be combined with an X10 test signal transmitter and X10 signal strength meters.

Regards, Charles Sullivan

Sounds like a useful widget, but there may not be enough of a market to pursue. People have become used to buying devices made in large volume in China. Printed circuit boards can cost me $10 or more in small quantities. The first XTB boards were $25 each delivered. Just components for this in small quantity would be over $50. I don't think many would invest that much for a troubleshooting gadget.

It is easy enough to identify phases just by flipping breakers and checking outlets with a small plug-in lamp. Standard panels alternate phases on adjacent breakers so a 240V breaker plugged in anywhere will straddle both phases.

Thanks for the suggestion though.

Jeff

DIY? Who, me? Why do you think I buy XTB's fully assembled! Phat Phingers!

Dave has suggested that as well. I'll have to look into them although I was hoping that something could be jury rigged out of surplus X-10 gear. Too bad the Firecracker is transmit only. If it received as well, it could form the core of a cheap detector. A battery-powered COTS X-10 receiver would be something even I might be able to work into a DIY project.

This weekend I was about to drag out the 100' extension cord, an old 12" parabolic reflector, a TM751 and the Monterey analyzer to make a "sniffer" that might work directionally. I managed to find the stuck Palmpad before I assembled the gear. It sure would be nice to have something like that in a very small, portable package.

I found one site that talked about reading X-10 signals but it pointed out that the signal strength is so low that meters have to be very close to pick up the signal:

I guess I'll look at Ebay to see if any of the units Dave mentioned are available cheaply. The stuck button problem has happened often enough that I am pretty sure it's going to happen again. This time my wife wasn't home so I dodged a bullet but I haven't been that lucky in the past nor do I expect to be in the future. Since she's never caused a button to stick, it's something that's a black mark on my record when it happens and X-10's already made enough of those!

-- Bobby G.

Nice price, but probably beyond my assembly skills. )-:

Eek! $300!! Although I paid about the same for the Monterey, I'm not sure I can justify that sort of expense for an RF meter.

Only 49 pounds for thre 315MHz units. Not sure what that comes to in American money but I suspect it's outside my price range again. I've found some used meters on Ebay for cheap (under $50) but with very little documentation it's hard to tell if they work in the proper frequency range.

Thanks again, Dave!

-- Bobby G.

It might be possible to build one for under $15 but I'm not sure how it might be affected by being handheld while searching for a transmitter.

The RF receiver modules that I specified the BX24-AHT and roZetta have a linear output pin which can be used as a signal strength indicator. You would want an inefficient antenna, a battery and you might get buy with a multimeter as the readout. More likely you would need a PIC and an LED or some type of visual indicator.

snipped-for-privacy@yahoogroups.com

An Oldie-But-Goodie National LM3914 (linear output) or LM3915 (log) 10-LED driver would fill the bill nicely (at only $2.50 or so).

... Marc Marc_F_Hult

>

I googled some more and found this after plugging in the specs of the UK model you found:

"Priced at only $80, the portable RF Meter is a low-cost signal-strength measurement tool for designers of RF radio systems. Two models are available to detect signals at 315 MHz or 433.2 MHz. Radio-transmitter signal strength is indicated by a 10-digit LED display with a one-second peak hold function to simplify measurement readings. The unit operates from a 9-V flat battery, and it has low-battery indication and an auto-shutoff feature that activates after 60 seconds of no operation. RF sensitivity is -110 dBm. The IF bandwidth is 600 kHz. Housed in a 4.33- by 2.56- by 0.94-in. ABS enclosure, it operates from 0°C to 60°C."

I can live with $80. The only problem is the above turns out to be an older article.

The vendor site:

Lists it currently at $114. If this site that tested the output of various X-10 transmitters was right,

there's no guarantee that this meter will even be able to detect X-10 transmissions at more than a few feet. That may or not be enough to work as a stuck key finder. I guess I'll keep looking - someone might have had a pallet load of them made in China and are selling them for less. If I

*knew* it would find stuck transmitters, I'd make the purchase but $114 for what could turn out to be totally useless gear is not an attractive prospect.

Side note to Jeff: "nevermind!" (-:

-- Bobby G.

(much snipped regarding an X10 RF sniffer)

The problem is that the economics just doesn't make sense for a low volume item. Elsewhere was mentioned the National LED driver. Yes, that's a cheap chip. But a finished product requires a printed circuit board and some sort of enclosure. That enclosure probably needs to be custom machined for the readout and to provide a few holes for controls and antenna. The XTB-II case costs $15 each at 25 pieces with its simple machining - mostly setup charge.

At low volume, parts alone for a X10 RF sniffer would be over $50, including a printed circuit board and a machined case. I just don't see the benefit for investing hundreds of hours into a project that may only sell 10 units.

One project I do have in mind is a power status indicator. Many people have reported difficulty controlling power to remote toggled ON/OFF devices, such as TVs. I have in mind a simple module that just transmits the status when the power goes above or below a threshold. It would also respond to a status query. There might be enough volume to produce something like that in the $40 to $50 area.

Other people have reported similar problems. We have several palmpads floating around ourselves, but I don't recall ever having a problem here with a stuck button. We have seen our cat send X10 commands though. ;-)

Jeff

The key spec is "RF sensitivity" which is -110dBm (typical). Off the top of my head, the small receiver modules I mentioned are -105dBm so the sensitivity is OK. However, I suspect these are European made so they may not be tunable to 310MHz. In the past, OKW has responded knowledgeably to technical questions like this so you might ask them whether it might be tuned to 310MHz.

I would put no faith in the results posted by Señor Fernandez. There are so many uncontrolled variables that his tests are meaningless.

I will test whether the small superregenerative receivers will work with a multimeter as readout. If so, it's a cheap way to fill what should be a rather rare need. I have a handful of boards with coin cell battery holders (Two CR2032 or BR2032) that might work for this.

snipped-for-privacy@yahoogroups.com

The test and my memory both failed. Oh, well.

The few remaining boards have solder pads for connecting +5V & GND but do not have the battery holder.

As I feared, it requires a bit more sophistication than a multimeter. It could be done on a breadboard with one of the 315MHz RF receivers (tuned to

310MHz) and a PIC12F675 to measure the RSSI pin and vary an LED's intensity accordingly. Cost:

Insteon RF phase bridges ("SignaLincs") have that capability - you plug one in, press a setup button, then plug the second one into outlets until its LED blinks the right way (a fast blink, IIRC). You get a pair of them in each of the various Insteon starter kits, and they sell individually for $40.

- Dennis Brothers

Using a PIC would require programming which eliminates many folks that could otherwise do a one-off DIY implementation without waiting (perhaps forever)for someone else to do the programming.

The output could be handled by an off-the-shelf National LM3914 (linear) LM3915 (log) or LM3916 (VU) Dot-bar graph driver chip which takes analog input and drives up to 10 LEDs (or 20 or 30, or more if you cascade LM391x's) with linear, log or VU-meter response.

So a one-off X-10 RF signal strength sniffer/meter could indeed be constructed *without* programming for ~$15 or so (depending in part on what is already in your electronic junk box).

1- Radiotronix RCR-315-RP $5.66 from Part 509-RCR-315-RP tuned to 310 Mhz as Dave suggests. 1- National LM3915 analog-in 10-step log LED Bar-Dot display driver There are also LM3914 (linear) and LM3916 (VU meter) versions which may provide better matches to the RF receiver's non-linear analog output which is non-linear. No good reason one could have both linear and log outputs. $2.64 from part LM3915N-1-ND 1- 5-volt regulator such as LM78L05 or equivalent $0.36 from part LM78L05ACZFS-ND 1- 9 volt to 12 volt battery with holder/clip 1- 4.7 ufd tantalum cap 1- 0.01 ufd ceramic cap 1- 10 ufd electrolytic cap 10 LEDs 1- 1.24K 1/4 watt resistor 1- 8.06K 1/4 watt resistor (or substitute 10k multiturn pot for above. See LM391x data sheet.)) 1- General purpose project pcb (eg Radio Shack) Case if wanted. Antenna materials

HTH ... Marc Marc_F_Hult

While I agree a DIY version without a custom PCB or case can be put together cheaply, I don't know of any electronic distributers that offer free shipping. That can become a significant factor when buying from multiple sources.

I had used a receiver some time ago that included an analog RSSI output (received signal strength indicator). That was useful to determine how to dynamically configure the RF network for best performance. The RSSI output could be monitored by a VOM. However, because of the pulsed nature of X10 signals, some type of filter or peak detector may still be needed.

Jeff

Jeff,

Don't get misled. I've been using these receivers for several years, buying the RWS-374-2 direct from Wenshing in Taiwan until they raised their price beyond where it made sense. The Radiotronix receiver is made by another Taiwan company.

The Wenshing datasheet has always labeled the two outputs "linear" and "digital". I see they have recently started silk-screening "RSSI", on the module itself, for the "linear" output but this is misleading as it requires further processing to use it as RSSI.

In the absence of a signal, there's a continuous noise output on both the "linear" and "digital" pins so just measuring the voltage tells you nothing. When I tried a VOM yesterday, there was a negligible difference between signal and no signal which is why I then wrote that a PIC is needed.

A PIC is required to separate the wheat from the chaff. The PIC analyzes the signal, validates it as an X-10 signal and then measures the difference between the final pulse and space which is a good indicator for signal strength. This is the way the BX24-AHT did it although it merely output the ADC value rather than use it to drive an LED.

It will be necessary to spend an extra $0.25 on a PIC12F683 as opposed to a PIC12F675. The PIC12F683, as you know, has hardware PWM which can be used to synthesize an analog voltage proportional to the signal strength. You could do this yourself in software but it would be discontinuous since you have to take time to capture and analyze the signal.

The first screenshot on this page...

shows you what the two outputs look like under signal and no signal conditions.

snipped-for-privacy@yahoogroups.com

Yep ;-) IIRC both Mouser and Digikey offer inexpensive shipping at USPS postage cost if you meet the minimum (which is yet another gotcha unless you have the patience to combine requirements/orders). Arrival to Greater Cincinnati is 2-3 days. A light package with a few chips (assuming that they don't overpack ! ) might cost $2-3.

What bumps up my purchase cost as much or more than shipping costs is my inclination to buy multiples of components. Who wants to get into a late night session only to wonder whether the problem is a particular IC and not have a spare to substitute? I'm usually too impatient to wait days for a replacement/substitute and so a single component that costs X$ often ends up costing me 2X$ (or more).

So as a practical matter, the first one of most/many things costs significantly more -- depending on home much is in my junk box/room. I could for example (depending on what RF receivers I find) build the first few of these sniffer/meters with what I already have. I know I have the LM391xs and all the other components.

But $15 is about the price of one CD. Does it really matter if you save the cost of golf fees (or whatever you call them -- I don't golf) by spending the equivalent of one or two CD's on a less expensive hobby?

... Marc Marc_F_Hult