Keep Them Bulbs On!!! 

"Simmering" the lamps like this also reduces inrush current substantially.

I work in the entertainment industry We call that preheat. It is used in high power stage lightning which is almost always hooked up to some sort of dimmer anyway. If you command the dimmer to be off, it actually sends some (adjustable) amount of power to the light This has not only the advantage of longer lasting bulbs It also limits the inrush current drastically which can be mandatory for running several lights on a single circuit that would otherwise trip the breaker Also it makes the light come on faster which is desired with audience blinders for example One downside other than the extra amount of energy needed even when the lights are of is that the filaments can start vibrating with the dimmed voltage. That "lamp singing" is audible and can be disturbing

Old " instant on" vacuum tube TVs did that...the filaments stayed on. It didn't noticeably shorten tube life. Some had a " vacation switch" to shut the filaments off. In 1976 I built a nightlight that was two 7 watt Christmas tree bulbs wired in series, house in a wooden bait box designed for crickets. My aunt had it for 40+ years, now a cousin has it in a garage. Not counting power failures, it's been on since Fall of 1976...

Note made by others from other tech contexts, but here's my addition. I know that in some countries, for safety reasons, the bulbs for railway and some traffic signals are hooked up so that they are in series with a relay coil to ground/neutral when in the "off" position. The relay is selected so that it pulls when in series with the bulb, but doesn't make the bulb light up. This means that if the bulb goes open (fails or is removed), the relay in series with the bulb will drop when in the "off" position, which means you only need simple relay logic to get an alarm output which says that the bulb needs to be checked up. I learned of this design trick a couple of years back, and up until then, I'd always wondered why some traffic lights and railway signals fade in and out rather than going hard on/off It's because they cool down slower because of the remaining current through the relay. And it wouldn't surprise me if bulb longevity is also a reason why they went this route, as it means that the bulb is always preheated and won't experience the same level of thermal stress. Posted as yet another application of the design trick, and maybe for some people to keep it alive in their designs.

Light dimmers also extend lamp life. I have a grandfather clock that has incandescent bulbs in the base illuminating the pendulum. The original bulbs are still in it. They have been running over 20 years at 1/3 brightness. They will probably last 100 years

Fran, infra red does not pass glass. You're measuring the temperature of the glass bulb, not the filament.

On DEC gear, the bulbs were run between "just slightly red" and "glowing enough to see". They never turned them off and they never took them up to the rated voltage and the change wasn't fast. As a result, they lasted for decades. You can find videos of people with PDP-8 computers with the bulbs all working.

I’m 64 years old, so I cut my teeth on batteries, switches and bulbs, and watched in awe as transistors (my transistor radio was about the size of a ladies handbag), LEDs (ooh! Soo pretty), and Integrated Circuits (witchcraft, sheer witchcraft) took over; but, I never knew this about lightbulbs! Love your vids - they are so calming and comforting!

Nice and simple. Makes me wonder if those old pinball machines had bias resistors too.

I have heard the early vacuum tube computers in the ENIAC era did this too. They left the heater supply always on at a low level to preserve the tubes.

Old security code interfaces for doors in buildings used to do this with incandescents the filaments were still luminous in their "off" state. Last one I have seen was used in the old MDW airport terminal before demolition.

Covers 2-3 areas actually, lowering thermal stress by preheating, lowering inrush current, and maybe under-running a little as well

Ah HAAA! I have definitely learned something new. Thank you!

This is really interesting and helpful! Think imma build/modify exactly one table lamp and put that circuit inside (maybe in addition to a battery for when it's unplugged given the low current requirement) just so I can have some nice incandescent glow whenever I want a break from all the rgb LED's and not worry about if I can find more bulbs. It's not like white LED's aren't fine 99% of the time (assuming they have a smoothing capacitor), but I'm starting to miss the old ways and have found myself seeking out older tools and electronics to fix or augment. They objectively don't make things like they used to

I figured that out when I saw that some pinball machines really whack some bright bulbs (they get far too bright to last at all)... except that the filaments are always on a little. I had no idea the extent to which this can be used. This is amazing to me! It would be good to install light bulbs where that would make some people uncomfortable. Especially for replacing those obnoxiously bright blue LED's found in electronics of the previous era. Perhaps a wifi or ethernet router, maybe a smart watch would be the worst. I don't have any that small but I've got a bunch that don't fit anything I've got. Next time I need to make an indicator panel I want to use always-on light bulbs. So thank you.

Just something I have noticed. I have 2 LED (incandescent replacement) bulbs that I have on all the time. They last much longer than those that I turn on and off. I understand it is completely different than what Fran is showing here. I do put the date on the bulbs when I replace them... yeah, nerdy.

Another advantage of keeping bulbs somewhat preheated during off cycles in an applicatio like Las Vegas signage is that your electric bill will be lower than if you're pulsing them from completely off to fully on and back again, because the filament has very little resistance when it's cold and only develops a high resistance once it's glowing red hot, and so the current rush at turn on into a cold lamp is enormous.

Atari's pinball machines had such a "Keep Alive" feature for the controlled playfield lamps.

The Omega Indicating Light brochure says, ". Lamp life increases (or decreases) with the 12th power of voltage reduction (or increase) while light output is affected only by about the 3.6 power of voltage reduction so a substantial life increase can be achieved with a small decrease in brightness". The power consumption does not decrease very fast with decrease in voltage, though, because the change in current in the incandescent region is small compared to the change in voltage. For indicator lamps, the power consumption is typically a minor consideration compared to lifetime. The inrush current to a cold filament when the lamp is switched on can be a substantial stressor. The addition of a fixed resistor in series with the lamp can substantially reduce the inrush current. It looks like the indicator panel uses this technique also.

For flashy signage like that you could totally leave it glowing a little, you'd hardly notice the difference when there's a hundred other bulbs right next to it. I'm sure some signs did that just to boost the overall brightness too.

I have also observed this with flasher relays for turn signals in vehicles. The old ones flash using a heater and bimetallic switch. The heater is in series with the bulb which provides the "keep warm" current. Newer solid state ones for motorcycles and ebikes I have observed to have a lower pwm duty when the bulb is "off" and full duty when on so when I used LED bulbs, they went bright dim instead of on off. On my car, however, it still used relays so the lamps are fully on or off and I do remember having to replace bulbs until I converted all to LED.

Learned me somethin' new today. Thanks, Fran!

Shows that we used to live in a time where wasting energy was less of an issue than having to occasionally swap out bulbs..

On UK fruit machines/ slot machines did exactly the opposite by over driving to get bigger flash. The other reason was they wanted to limit life & keep the high turnover of machines with the game changing in the cabinet. In the UK the machines regularly had new bulbs with wiring harness & front glass!

I read doing something like this for “numitron” style displays when driving them from a CD4511. Known as a preheating resistor

Also later when companies did start to purchase computers, if they were big enough they would also have a technical maintenance workshop that would repair parts even if they paid for a maintenance contract to have a technician come out and replace the parts. The reason being that having a machine down for the day or so it takes to get the maintenance tech from the company out to fix the problem would cost the company a lot more than it cost to fix the problem yourself right away. Sure, they probably voided the warranty for that part, but again, paying for a new replacement is usually going to be worth the cost compared to the downtime. And it's still done today. We've still got a technical maintenance workshop at my work. In fact I was down there last night and there was a tech repairing a panel light that blew because of a shorted LED. Chopped out the dead LED and brought the panel back to life. A quick and easy repair to get the panel back up and running. Maybe not as good as new, but cheaper than the cost of replacing the whole panel, and it gets the crew back out on the road with a full kit.

The resistance of tungsten changes by temperature quite radically: - at room temperature (about 300 K) 5.5 micro-ohm cm - at 500 K 10.5 - at 1000 K 24.3 - at 2000 K 55.7 - at 3000 K 90.4 - at 3500 K 108.5 So, from room temperature, the resistance up to essentially melting point, the increase is nearly 19-fold, or closer to maximum normal full operating temperature of 3000K, it is some 16-fold. Calculating the opposite way, the inrush current can be about 16 * the rated operating current. The bulb wire probably has some part that has lower resistance, and some with higher than average. In any case, there is a harsh jolt (thermal and also magnetic/mechanical) at the high inrush current. You may have noticed that light bulbs almost always fail at the moment of flipping the switch to ON. This is the reason for it. If you keep the filament pre heated to 1000K (just below incandescence), your inrush is limited to about 4-fold. Now consider that the electro-magnetic jolt is rather proportional to the square of the current, you may enjoy only 16x rather than 240x jerk at the time of turning full power on. Finally, you can benefit even more, if you design the operating full temperature to be lower than the example 3000K.

Running the bulbs at less voltage than they are rated for sure. Example>>traffic light bulbs rated at 130 volts. They last quite awhile.

Used to have a vintage red book from Audel's New Electric Library that featured electromechanical lighting display controllers that were gear driven by a motor. Audel is part of Howard W. Sams.

Big reason why the Centennial Bulb has lasted so long, it never gets cycled on/off.

Gene Simmons-type dude says: "The government cannot handle our bulbous bulbuosity!"

So interesting! Thank you for recording and sharing this.

I may use this idea to extend the life of my Numitron clock display. I already run the tubes at 3V not 5V but this may help them last longer. Numitrons are really expensive.

I have a theory that says keeping your computer on all the time has similar longevity benefits

In the case of signage, I wonder if flashing in fractional to single digit Hz, if the filament has time to cool enough to contract and cycle fatigue.

This was new knowledge to me, thanks Fran!

This is a fascinating explanation, thx Fran!

Fran, are the Minitron/Numitron seven segment filament displays always-on too?

I had no idea. Thanks for a very interesting video Fran.

Thank you!

Depends on the lamp type. the high powered theatrical FLK/FEL/ect all need to be operated at high temperature. At low output levels, the filament vaporizes metal and deposits it on the glass. At full temperature, the filament metal gets vaporized off the bulb and redeposited onto the filament.

I've seen many long hours small bulbs, both still working and dead: they all got significant darkening on the internal surface of the glass - condensed material evaporated from the filament.

Now there is something I did not know. Huh. After all these years I now know why tungsten lights burned out so quick.

I would guess this "almost on" state would also reduce the response time to get to full brightness and therefore allow faster flashing of those animated signs.

Thanks for the TIL!

2:25 I guess that's only regarding the filaments in tubes? I've heard that cathodes wear out much faster with always-on heaters, and every vintage TV youtuber I watch warns about sets with "instant-on" CRTs.

Regarding the conspiracy of putting limit on the lifetime of lightbulbs, it's not entirely evil. To replace a 60w "short life" bulb you would need a perhaps 75w "long life" to get the same amount of light. That extra power would cost you more than the savings in replacing bulbs. It would also cost the power companies more in infrastructure. So the lifetime is the best compromise for the consumer. For photography, sometimes the lamps are run at perhaps 120% when the picture is taken to get extra light, then turned low in between shots.

😮😮 Cool

Thanks!

what about appliance bulbs, like in the fridge? turned on & off a lot & also cold/hot, yet they seem to last a long time

Bin A View Hey Fran!!! Love that telco wire. I have a bunch of that.

Ah, the days of cheap power and no regard for the pollution of those coal power plants!

I'd just like to point out that it is redundant to say that "it self-limits itself" is redundant. 😛

Interesting… Thanks Fran. I remember the IBM 360 computers had a panel with hundreds of tiny lights that flashed and flickered with different intensities as the computer ran programs. You could tell how the program was progressing by watching the light patterns change. I never thought about the issue of light bulbs burning out. Never saw that happen.

did they use this in cars? I have a Volvo from the 60s and as far as I know most bulbs are either original or changed last in the 80s and still working fine

❤

Neat! Does standby mode for a guitar amplifier work in a similar way?

That signage was from Reno, not Vegas.

I Watch you videos thruly but I skip the patreon list finale. I wonder whether RU-vid thinks my watching Is just a swipe watching or a complete one

Well. This vIdeo makes up for not firing it up.

Should this video have come with an epilepsy warning? 🤔 (I'm kind of disappointed that "no light bulbs were or harmed in the making of this video". I probably need to watch Big Clive or Electroboom for that!)

Can I make a request? Can you use any other multimeter but that one? Or if not, do a content warning at the start? Thanks.