Solve the Garage Door Opener Puzzle !

Old style diode with the package made of glass? EMC pickup from the fluorescent lights?

Vladimir Vassilevsky DSP and Mixed Signal Design Consultant

Nope.

Bingo! That was fast.

One was burned out and I didn't feel like making the trip to pick up the rugged heavy filament bulbs for the purpose. Other CFLs in the garage, but they do not cause any problems.

Best regards, Spehro Pefhany

Going with your "hint" I'm supposing "some minutes" is "long enough for the light to go out?" and connects to the already posted glass-pack diode idea.

On one more like 4 years old we got some utterly whacky behavior from a nearby lightning strike screwing up the logic board - one of those nice modern "replace the whole thing" types, since swapping chips doesn't help when they need to be full of the right code, too.

Said behavior was potentially dangerous - driving the door down into the ground with no attention to either the limit switch or the "force sensing" "reverse if it hits something" function. If anyone had gotten between it and the floor while it was affected (it still opened and closed happily, it just wanted to keep closing forever) they would have been hurt. New board did fix it, but cost half what the whole thing did originally. I refer robust relay logic for things that should keep working.

Maybe ;-)

Two things affect downward motion: (1) A torque setting (pot on drive train board) and (2) The photocell loop at the bottom of the door opening.

Both effects will usually exhibit a jerk attempt to go down, then stop.

Replace your Genie with a Liftmaster belt drive... I did ;-)

...Jim Thompson

That should not have happened (failure of a safety-critical system in a way that was not safe), and I would have thought the manufacturer would replace the board for free both to get it for analysis and lest they be subject to liability.

Best regards, Spehro Pefhany

This one's old enough it doesn't have the nanny state photocell protection, just the reversal on obstacle.

...

If it ever actually fails, but it looks like enough of a hassle that I'd probably get someone to install it.

Best regards, Spehro Pefhany

I got fed up with "nurturing" my Genie openers, and called Liftmaster. They did the install and I haven't had a bit of trouble... knock on wood ;-)

...Jim Thompson

...which just indicates that you, like I are old enough to remember when stuff was actually supposed to work, and manufacturers were supposed to care. This was of the "our whopping 1 year warrantee is up, and we'd be happy to charge you twice the price of the whole thing to send someone to come look at it - and you shouldn't use it if it does that", with the latter being all the seemed to think they needed to cover their butts liability wise. Repair parts were not on offer.

I had to search across the internet to even find someone selling parts - it was a hardware-chain house-brand that turned out to be made by Chamberlain, and failure of these boards is evidently common enough that there are several sources of supply once you know what to look for. In point of fact I don't know exactly how long it was this way, having only noticed it when I was inside the garage and closed the door (not very common, as it's detached) and noticed that it seemed to be straining to go further than it should.

Same experience, but it only took a few minutes to figure out from the symptoms, aided by the internal lamp time-out period There are going to be places compact flourescents should not go, even in newer equipment.

Assume the drive train problem was chewed nylon worm gear, also....

RL

Too bad, I liked being able to get more light out of thing without exceeding the wattage limit.

Yes. BTW, thanks for making me look up the proper name of those gears. I was under the misimpression that the 'worm gear' was the mating part (the worm, rather than the worm wheel).

Best regards, Spehro Pefhany

Yes. And a separate issue.. I'm also aware of some of the firmware and design safety standard requirements for this type of product, which are intended to keep the likelihood of things like that from happening ever in the life of the product extremely low. Could have ended up as a mandatory recall situation.

A bad kind of safety failure indeed. If it just quit it would be a lot more acceptable from a safety pov. Imagine millions of these things out there. And a subtle issue.. some people will act differently (less safely) when they believe the protection is in there for them, so the protection had better work reliably.

There are deaths regularly (and injuries), mostly of children, in the US every year from garage doors. Eg.

Too bad they redacted the mfr/product information.

Best regards, Spehro Pefhany

I'm not sure that the use of relays, in your case, would have produced a more satisfactory failure mode.

Improved immunity to I/O and supply transients affecting the control section would seem more appropriate. Failure of some functions might also be detectable by less exposed subsections of a controller - there is usually some kind of fault detection attempt in circuits I'm familiar with, even if only a simple timing fault response, for a powered mechanical event.

RL

The 'force sensor' is the interface and control section that would normally be depended on to avoid potential injury.

The fact that this failure was visible immediately upon failure could be considered a good thing. You wouldn't want to have to wait for a time when it's actual intended safety function was called upon to do it's job, to find out that it wasn't going to work.

RL

Lamp failed shorted and pulled the internal voltage down too low for the receiver to operate.

One the light timed out, the voltage recovered and all was well.

All the ligts that I've seen on garage door openers were 120 VAC and ran off the AC line.

Yeah. Right.

Maybe aluminum house wiring, about to fail and burn the place down?

...Jim Thompson

Did you put a CFL light in it?

Did you put in a higher power light?

A CFL can be making RF when on as could a solid state switch turning on a higher powered light.

Yup, you got it. But Vladimir Vassilevsky got the right answer less than ten minutes after I posted the puzzle.

Yes, the RF overwhelms the receiver front end.

...through some skinny little PC board traces, common to the logic power supply.

So it turns out that it was r.f. noise from a cheap CFL. Probably still aggravated by the impedance common to both components.