I think the hacking community online has a bit of a File Drawer Problem. This is a term I’m cribbing from Real Science™ which refers to the fact that very few scientific studies with negative results ever get published. Scientific journals tend not to find them as interesting, and researchers tend not to submit them, all for obvious reasons. Thus, negative results languish in file drawers where nobody ever sees them. However, knowing what doesn’t check out experimentally is actually just as important as knowing what does. This is part of why, for example, every preliminary study that suggests some tiny effect gets reported all over the popular media as the next revolution. Nobody is writing pop science articles in Time about the two hundred previous studies that showed no effect in that same area. People (including other scientists sometimes) get a skewed impression of the state of what is known on a particular topic because all the negative data is sitting in file drawers.
In the case of hacking (and also “Making”, though I find that word tiresome), people only write about the successful things they do. That’s understandable, but the fact is a lot of things we try, particularly if we’ve never done them before, completely fail. That’s great! It means you tried something and learned something. Failures are a necessary part of the process. However, because nobody writes about it, I think it creates a false illusion of the ease of success. It can look to an observer like everyone on the internet is happily cranking out all this great stuff on their first try without much effort. When a new person who is interested goes and tries this stuff, and it fails, they get discouraged. How come they didn’t succeed like all those people on Instructables seem to? Because nobody is writing blogs about the fifty times they tried that thing before they got it right.
I’m guilty of this myself, and although I have written about one total failure in the past, it’s high time I do so again. This time, it’s a garage door opener. I seem to have an odd fascination with these things, having written about them on at least one other occasion. Maybe I’m drawn to the science fiction aspect of them- it’s a button that you push that makes a wall go away and come back at will! How is that not cool?
Anyways, the current edition of Dunki Freehold has a very nice garage, with a rather annoying garage door. The previous owners only left me one remote for the very elderly opener, so I went and bought another one online. Like everything else, garage door openers are much smaller nowadays, and come in keychain form.
This all seemed fine- the new tiny remote is destined for my daily driver, and clips nicely on to the visor. The bulky old one lives in the weekend car. The old remote works pretty well, so I wasn’t aware of a problem brewing until I got the modern small one. The problem is that the range on the new remote is terrible. Sometimes I have to be only a couple of feet from the door to get it to open. Not very convenient, to say the least. I had been suffering along with this, but then happened across an article online about a simple way to extend the range of garage door remotes.
It seems that there are two things that can reduce the range of an opener- a metal garage door, and a metal roof on the building. As it happens, I have both of those. It’s not surprising that my garage would be a huge sink for radio frequencies. The new remote is much smaller, with a much smaller battery and transmitter antenna. It stands to reason that it is lower power, and might need some help. This article I found claims you can cure this problem by simply relocating the antenna on the opener to a place outside the building.
Here’s where I need to make clear my complete ignorance on this subject. I’m barely an electronics hobbyist, and I certainly know next to nothing about radio antennas, or the design and physics behind them. Based on what little I have gleaned over the years, however this idea does make sense. It is my understanding that the length of a receiving antenna needs to be a multiple of the wavelength of the signal it is receiving. Making the antenna longer doesn’t necessarily improve reception- this is a very common misconception. Sometimes a shorter antenna that is the right ratio to the wavelength is better. If I’m completely wrong about all that, please tell me in the comments.
So how do you relocate an antenna without making it longer? By adding a shielded section to it. The article that set me on this goose chase suggested simply running a length of coaxial cable from the original garage door opener to outside the building, then attaching the original antenna to that. Again, based on my limited knowledge, this all seemed to make sense. I happened to have a long chunk of RG6 coax left over from when I gleefully disconnected my cable TV forever (with extreme prejudice), and this seemed like a great way to recycle it.
A couple of amusing things about this photo. First, you’ll notice it’s immediately obvious why the light socket on the right (in this photo) wasn’t working. The spade connector has fallen off. Well, that’s an easy win for starters! No doubt it vibrated loose, as garage door openers are pretty much vibration machines. Related to that, notice the field-expedient strain relief on the thick brown/white wire pair in the upper left. Methinks that should have a grommet where it passes through that sharp-edged metal plate, but hey- I’m not a Garage Door Opener Designer. Who am I to say.
Anyhoozle, back to business. Let’s get that digital PCB out of there.
My initial thought was to mount a coaxial connector on the PCB in place of the antenna, so that I can screw a standard cable to it and route as needed. The junk pile coughed up a gem that anyone who grew up in the 1970s will recognize in that photo above- a 75Ω to 300Ω impedance matcher. This was used for things like adapting cable TV to your older television that only had screw terminal connections for an over-the-air antenna (typically rabbit ears). You might have also used one to connect your (then) very modern Atari VCS to your 1968 Electrohome Radiation King.
Note that because this impedance matcher is from Canada, it has the very common (up there) Robertson screws on it. These are not “security screws” as every American insists on calling them. They are in fact a very common format that is vastly superior to the Phillips format that America is obsessed with. Fair warning, if you call it a “security screw” to my face, you will get punched in the pancreas. That is all I have to say on this subject. My soapboxes are weird but very tall. Regular readers know this.
Anyways, this lovely artifact is going to give its life to provide me a nice female coax connector.
Unfortunately, while the coax connector would have fit nicely on the PCB, it quickly became apparent there wasn’t anywhere close to enough room for it inside the garage door opener itself. Clearance in there is very tight. We were going to have to do this less elegantly.
This all seems pretty sweet so far, but I wasn’t done yet. One other thing drives me mad about the new remote. It has those rubber mush buttons that feel like you’re pressing your thumb into the cold bloated thigh of a dead hiker you found in the woods that time. It’s a block of rubber with a conductive pad on the bottom that bridges some traces on a PCB. Whomever invented these buttons is a monster. I’m sorry whoever you are, I’m sure you’re a nice person, but you are a monster. That’s objective fact.
Because of the mushy buttons, it was just as possible that my problems were because I couldn’t tell when I was really pressing the button properly. It was impossible to know because the tactile response is so vague and unhelpful. I decided to attack this problem from both ends by replacing the mush with a mechanical button.
That was an afternoon well spent, right? I have an awesome new remote, and a relocated antenna that massively improved the range? Well, actually, utter and complete fail. The relocated antenna appeared to have no effect whatsoever, and to be honest I have no idea why. It really seemed like this should have worked.
As a hail mary, I reinstalled the original antenna it its original location on the PCB, and tried doubling the length with a piece of the same gauge of wire. Again, referring back to my anecdotal knowledge that “antennas are supposed to be a multiple of the wavelength they are receiving”. This also had no perceptible effect.
I did some more tests at this point, and I’m no longer sure that range is the real problem. There seem to be dead zones and areas where it works well. There’s a spot at the end of the driveway that is pretty far away, but if I hit the button in that small area, it works about 80% of the time. A couple of feet closer or farther, and it doesn’t work at all. There’s something more complex going on here that I don’t understand. If there are experienced radio people in the audience, feel free to chime into the comments with suggestions, or just to tell me how wrong I am about all this. If you’re a dick about it, you’ll get moderated, of course, but if you’ve been moderated in the past, you already know who you are.
The whole adventure wasn’t a total loss. I got to take apart my garage door opener, which was fun, and I did make the remote much more pleasant to use. I’ll keep thinking about this problem, and perhaps inspiration will strike again. In the meantime, here’s one less negative result in the file drawer.