Outdoor flickering “candle” failure

[Hafcanadian]

I’m always having issues with my garden and patio LED lanterns, whether battery or solar powered. The one I’m trying to fix now is two AA powered. The lamp won’t come on at all, even though in the past it and its twin had funky switches. The switch seems okay this time, at least when I check continuity between its posts, and at one point I temporarily soldered in a larger switch that didn’t alter the result.

The board is cleaner than most that’ve been outside a lot, but the lamp won’t work. That said, if I set my LCR tweezers to “Diode” and put the tips across the lamp’s leads, the LED flashes (not flickers), so the LED seems okay. It’s flickering candle imitation must be controlled by a board component, or under normal voltage it may be the lamp’s built-in mode.

I note a “crystal”(?) can on the back of the board.

See photos and please advise what to check.

Thanks.

 

[ahsrabrifat]

Set your multimeter to DC current (mA), in series with the LED if you can.
Or, if not possible, put it across one leg and see if any measurable current is flowing.
Use diode mode on your multimeter to probe between its pins (usually Base, Collector, Emitter).
You should see ~0.6–0.7V drop B–E or B–C in one direction. If you get 0V in all directions, or OL in all, the transistor may be dead.

 

[schmitt trigger]

First of all, I applaud your attempt to fix this and avoid sending another piece of E-waste to the trash.
Do you see that black epoxy blob on the board in there? There is a naked chip below. It is the absolutely cheapest way to purchase and assemble a semiconductor, saving the manufacturer a few micro-dollars versus a SOIC or TSOP.
But the bond between the epoxy and board is NOT hermetically sealed. Humidity can, and definitely will ingress and damage the semiconductor. And it cannot be replaced.

I sincerely hope that there is something different and you can actually fix this device. But garden lights are usually a prime example of fecal-matter electronics devices.

 

[elfcurry]

First of all, I applaud your attempt to fix this and avoid sending another piece of E-waste to the trash.
Do you see that black epoxy blob on the board in there? There is a naked chip below. It is the absolutely cheapest way to purchase and assemble a semiconductor, saving the manufacturer a few micro-dollars versus a SOIC or TSOP.
But the bond between the epoxy and board is NOT hermetically sealed. Humidity can, and definitely will ingress and damage the semiconductor. And it cannot be replaced.

I sincerely hope that there is something different and you can actually fix this device. But garden lights are usually a prime example of fecal-matter electronics devices.

Would it help if such 'blobbed' cheapie boards were treated by covering the blob with varnish or some moisture-resistant sealant when new to make them last longer?

I'm assuming these cheap items work well enough when new and indoors but if used much outside will degrade.

 

[rjenkinsgb]

The blob is epoxy, which should be waterproof - at least initially.

I'd suspect failure is more likely to be from repeated thermal cycling, with different expansion in different materials eventually causing cracks.

If that damages the bond between epoxy and PCB, it could allow moisture in - but it may also result in the microscopic bond wires from the bare IC to the circuit board starting to detach.

Coating the entire board and wire connections in a slightly flexible varnish or resin would help prevent problems due to moisture, but you cannot do much about thermal stresses.

 

[Hafcanadian]

Reckon I’ll just toss this board and see if I can find a timered “flickering” LED assembly that I can fit inside the “candle” and resurrect the lantern.

Like these, only 3”X3”

 

[Hafcanadian]

Shoulda done this long ago.
Searched many options on Amazon and found this from Volnyus. I have two lanterns, one larger than the other. The large one:s switch is still unreliable, but working for now. The Volnyus is a set of two (other counts available), the same 3v power, different “flame” setup but maybe better, doesn’t have an annoying screw in addition to a catch holding the battery hatch closed (why do they do that?), and has the same size candle unit for the small 3”X3” one.

I simply hot-melted the new candle into place in the carriage frame (the old one used small screws that the new one doesn’t). The big advantage, beyond being brighter, is that with the new remote control I can set the candle to be solid, or flicker like the original, can set it to four different timer lengths, and can dim it 4 levels if needed. When the larger unit eventually gives up, I’ll put the extra Volnyus candle in it; it’s shorter than the original, but so what. Last but not least, the set of two was only $17, a darned sight cheaper than the labor hours value I put in to trying to analyze that faulty board.

Thanks for clueing me in to the IC “blob” being an unserviceable likely culprit, so that I was driven to move on.

 

[schmitt trigger]

The screw in the battery’s hatch is a requirement for battery operated toys, to prevent toddlers from taking the batteries to their mouth, and probably swallowing them.

It all started with coin cells, which do make a lot of sense, but extended to any type of battery. And now it seems to any type of battery operated device.

**I didn’t write the legislation, if you wish to write nasty comments, direct them to someone else.