The Secret To Everlasting Bulbs - Always On!
Added 2024-02-11 01:24:58 +0000 UTCThere was always a trick to making sure a light bulb would last a very long time. Let's look at how it was done....
Comments
Your thermal image can't "see" the barely lit filament because glass is 100% opaque to far infrared - all you're seeing is the temperature of the envelope (allowing for emissivity). After watching your video, I tried viewing a 6V filament lamp on a switchable near infrared security camera. At 0.5V the filament is just visible to the naked eye and camera with the hot mirror filter engaged (ie. daylight mode), at 0.4V the filament appears to be completely out, even in a darkened room, yet still shows up on the security camera in night-vision mode. The voltage has to set lower than 0.35V for the filament to be invisible to the NIR camera...
Gordo
2024-03-17 17:27:22 +0000 UTCIt's in Livermore fire station, California
Gordo
2024-03-17 17:18:04 +0000 UTCHey Fran I know you said you won't be trying a long-term test but do you think that might change? Because I'd be interested in giving it a shot. I've got a corner of the shop already set up with time-lapse video monitoring so it would be easy for me to shove a few flashing lamps in the corner for observation.
Gia Ferrari
2024-02-16 20:31:44 +0000 UTCThe Centennial Light is the world's longest-lasting light bulb, burning since 1901, and almost never turned off. From memory it's situated in the basement of a fire station. Full disclosure, the first sentence I just cut and pasted.
Steph Traeger
2024-02-14 02:01:18 +0000 UTCGreat video! You explain things so well. Thanks so much!
Stewart Webb
2024-02-12 22:37:45 +0000 UTCFran, it’s amazing to see the lengths we had go to in order to make filament lamps last as long as possible. But here is something I didn’t know about tungsten filaments until I found it in the company component selection guide. At the head of the incandescent bulb section, they had a page dedicated to the difference between AC and DC driving that affects lamp life. Driving an incandescent light with DC power results in the filaments developing a condition they called “DC notching”. This caused weaknesses in the filaments that caused very early filament failure. Conversely, driving the same filament using AC power maintains the integrity of the filament, allowing it to burn for years in many cases. The guidance was to use 5V filaments instead of 28V filaments because the physical filament is not only shorter in 5V lamps, but it is also thicker. The recommendation was to slightly under-drive the 5V bulb and use AC power. This is why the incandescent 5V bulbs used in edge-lit aircraft panels are driven at 400Hz and they literally can last for years. The bulkier 5V filaments are also a lot more sturdy and able to tolerate the vibration of an aircraft better than the long filaments in 28V bulbs where they have to drape the filament over one or two filament “supports” in the bulb. The highest degree of incandescent filament preservation I have ever seen is used in the KC-135 tanker aircraft fuel management panel. The panel is filled with stretched-filament 5V seven segment display digits for sunlight readability. This was back in the early 80s and money was no object. They had no real alternatives to these incandescent displays due to space limitations. So, the designers did exactly what you demonstrated. The filaments were never completely off. They received just enough voltage to remain warm, which kept the filaments pliable and less-susceptible to aircraft vibration and the cold-shock of lighting a long filament segment from cold to immediately hot. So, they did everything they could to be gentle to the incandescent digits. But they took it a step further. The display driver circuitry was able to detect a burned-out filament segment upon power-up of the panel and cause the bad digit to flash the remaining working segments in the display. This alerted the crews, and ground maintenance, that a digit needed replacing. If a single segment of the 7 segment display was out, an 8 may appear as a 0. Not good for fuel management over so many fuel tanks in the KC-135. The fuel management panel, and associated remote computer, were radiation hardened since the KC-135’s original mission was to supply B-52s with their fuel needs at all times. Back in the cold war, they actually had a safe in the KC-135s like they did in the B-52s. (Like in Dr. Strangelove) and a lot of times the KC-135 war plan in the “cookies” they opened commanded the KC135s unload ALL available fuel to the B-52s so their final missions deep into Russian territory could be carried out. The crew could either try to parachute out of the KC-135 (provisions were there for that) or chance surviving a ditching. Anyway, no expense was spared in this critical nuclear link. They built somewhere near 800 KC-135s and there are about 400 still left in service. Anyway, the cost of these incandescent digits slowly increased after the cold war since no new design would use incandescent digits. So only a few suppliers are left and each digit is over $100 each (in quantity) and the bar graph versions, which were custom, were hundreds of dollars each. Considering how many digits were used in each aircraft, multiplied by 400 aircraft, results in the Air Force spending over a million dollars keeping the incandescent replaced. They tried to substitute LED 7 segment digits only to find the “keep alive” current used for the incandescent digits lit up the all the LED segments lit all the time. It also appeared to the display fault detection circuitry like an open filament and thus blinked every LED replacement digit all the time. They tried to replace this fuel control panel several times only to find the replacement would be incredibly expensive for 400+ aircraft. Now they have a bid out on the street for a total replacement of the panel with an LCD touch-screen, or equivalent. So, in conclusion, running 5V bulbs with AC and using the “keep alive” voltage is the best way to deal with incandescent filaments. I’ve had MIL-spec 5V bulbs in those twist-lights running for over 10 years. Not all made it that long, but more did than not. The parts selection guide was right. Your demo was identical to the approach some military incandescent display drivers used. Very cool Fran!
Matt Wietlispach
2024-02-12 02:01:21 +0000 UTCAlways fascinating to learn the reasons behind certain design features, and your way of explaining is always easy to understand. 😃
Motten
2024-02-11 23:03:02 +0000 UTCThat is pretty awesome Fran - I saw this done in a military receiver once. I've also seen this with the "instant on" filaments too - it's a cool trick.
Fred Niell
2024-02-11 13:52:24 +0000 UTCThe tungsten filaments have a positive temperature coefficient, so a cold filament has a much lower resistance (more current) than a hot filament. The inrush current when turning on a cold lamp contributes to the mechanical stress on the filament, leading to the typical pop failure when turning on the lamp. You would need to monitor the current on a scope to catch this.
Erik Blake
2024-02-11 09:26:27 +0000 UTCSome of the tube TV's with "instant on" also kept a standby current on the filaments.
Daniel Bingamon
2024-02-11 03:17:36 +0000 UTC
