The extreme ends of the sintered ingots don't get as hot as the rest of the bar because they're held in water cooled contacts. So the ends are discarded because they aren't as dense as the rest. My customer approached me once to build a tool to snap the ends off instead of whacking them with a hammer as they'd been doing for 50 years. I never did do it, I think because I convinced them it would be the civilian equivalent of an $800 hammer.
Oh ... I wondered why that wasn't on RCM. Thunderbird changed its user interface and "Reply" now means "to the sender", whereas "Followup" means to the group.
Anyhow, here's the reply that I meant to post:
I think Ned is saying "How so would limiting the inrush current prolong lamp life?" If so, I agree. The current doesn't matter - what matters is the filament temperature. You only have inrush because the filament is cold, if the filament is cold its life is not being shortened. As the filament heats up, the current goes down.
Unless the temperature overshoots because of inrush, but I never seen a claim for that. I suppose that it's possible due to the lag between current and temperature (it takes a finite amount of time(?) for the temperature (& resistance) to increase after a current increase). This should be easily modeled, but not easy for me.
I suppose that really high currents could have damaging effects from magnetic forces, but that's also something that I've never heard claimed, nor that I can explain.
Bob
[ ... ]
[ ... ]
Another thing which really shortens the life of an incandescent is to bump it while it is running. This joggles the hot filament, causing two adjacent loops to contact each other and weld together, causing the overall resistance to be lower, and thus the current to be higher, and thus evaporate the filament much more rapidly.
What is it -- bulb life is something like the inverse of the fourth power of the applied voltage? Short out part of the filament and you have the equivalent of higher voltage.
O.K. Worse, based on this site:
expected-life = ( (nominal-voltage/applied-voltage)^13.1)*1000
I can imagine that.
Enjoy, DoN.
"Michael A. Terrell" on Thu, 09 Jan 2014
01:26:12 -0500 typed >> Michael A. Terrell wrote:Cool. I've been thinking about building a combination Captain's Bunk and "Dutch Cupboard" sleeping space for years. Have many sketch in many notebooks.
-- pyotr filipivich "With Age comes Wisdom. Although more often, Age travels alone."
Ned Simmons fired this volley in news: snipped-for-privacy@4ax.com:
You don't read, do you, Ned?
NTC! Not PTC! Geesh!
"Jim Wilkins" fired this volley in news:lan72l$127 $ snipped-for-privacy@dont-email.me:
Yes, and that's how Halogen-Cycle bulbs last so long. The tungsten is caused to re-condense ON the filament, rather than inside the globe. It still does so somewhat unevenly, so they eventually burn out... and almost always at cold inrush time.
LLoyd
Bob Engelhardt fired this volley in news: snipped-for-privacy@news1.newsguy.com:
No, Bob. Thin spots (like at the crimps, and where it's thinned with age) get hotter than the rest, and more quickly (the real key), so that those hot-spot's resistance increases more quickly than the rest of the filament.
That causes the thin spots to sort of 'run away' thermally, dissipating more and more share of the total power until they open. If you can heat the filament more slowly, you can minimize that effect.
My only commercial experience with this is with the small incandescent lamps used in old DEC and DataGeneral computer front panels. We were _forever_ replacing lamps in those darned things until one guy in the shop came up with the idea of putting an NTC in series with each lamp (labor being more expensive than the lamps or the NTC gumdrops).
Since those bulbs were constantly being switched on and off, that was a good solution. It extended the average life by about 3X in that application, without any significant reduction in brightness (at static- state full-on). They were quite a bit dimmer in the 'run' state, but also meaningless in Run, so that wasn't a problem.
Lloyd
An NTC has to drop some voltage to heat up and maintain its low resistance, until heat from the bulb reaches it.
I designed the liquid level sensors for an ink jet printer with NTC thermistors, and then dodged a competitor's intellectual property lawsuit by saying that the low-fuel light in my car used one the same way. Thermistor app notes show how, if you know to look there.
"Jim Wilkins" fired this volley in news:laouhm$dm3$1 @dont-email.me:
Of course, they always keep some small resistance, but properly installed, it's minimal. But that's not what Ned Sed. An NTC does NOT act like a "plain resistor", unless he's re-invented the whole science.
Lloyd
"Jim Wilkins" fired this volley in news:laouhm$dm3$1 @dont-email.me:
Sure... there are even 'true level' detectors that use either an array of NTCs or a long, thin one, partially or fully immersed.
LLoyd
No I read the article. The NTC limiter still has a small resistance when hot, but, considering that the life of the filament is inversely proportional to voltage^16, even a small voltage drop will substantially increase the lamp's life. The inrush has little to do with life except when the filament is about to fail anyway, so a low value resistor will accomplish the same thing as the NTC limiter.
I didn't say a resistor would behave the same as an NTC. I said the extended life of a lamp in series with an NTC would be due to the hot resistance of the NTC, not the inrush limiting. Thus a plain resistor would accomplish the same thing as the NTC in this case.
The tradeoff is that the bulb gives out less light. For non-critical, easily replaced bulbs in reading lamps etc it's more economical to use a lower wattage bulb at full voltage, since they consume maybe 5X their price in electricity during their life.
In college I experimentally shut off the steam radiator and used a
200W bulb for heat. Although electricity was more expensive than fuel oil, the radiant heat from the bulb kept us comfortable at a lower air temperature. There were too many unknowns to quantify the difference. jswNed Simmons fired this volley in news: snipped-for-privacy@4ax.com:
That's not the case, at all, and I have professional experience to the contrary. If installed close enough to the lamp to be heated by it, a properly-selected NTC has minimal hot resistance. I-lamps almost always fail at inrush time, until very, very old.
Lloyd
Yes, that's why filaments are designed to run at a relatively high temperature. It's not a conspiracy by the mfrs to get us to buy more bulbs. The most economical operating point may not be the same for a bulb that's easily replaced compared to one that's hard to access or critical for safety.
Same with a wood stove. We spend most of our waking hours in the room with the wood stove. The amount of wood we burn is considerably less than the BTU equivalent in oil it takes to keep the place comfortable.
Bullshit, you have anecdotal experience that you've misinterpreted to reach a false conclusion. If we're comparing bona fides, I have professional experience working directly in the production of incandescents, including working closely with the filament designers.
Show me a credible reference that the life of properly designed and manufactured filament in a conventional lamp is reduced significantly by inrush.
I'm curious, Pete: Why would anyone want a bed in a box? I'm more comfy sleeping in the open, myself. Besides, this just doesn't look like _you_:
Yes, but it costs considerably more to get the 200x amount of pollution from the wood stove out of the air. Screw wood heat.
Larry Jaques on Fri, 10 Jan 2014
17:23:51 -0800 typed in rec.crafts.metalworking the following:Two reasons. One, making the bed in the morning is as simple as closing the door. Two: easier to heat a smaller space like that with body heat, than an entire bedroom. Three: neato making factor.
-- pyotr filipivich "With Age comes Wisdom. Although more often, Age travels alone."
OK, point there. But repudiation lies here:
True, if you like small, stuffy, but warm spaces.
OK, but that lasts, what, a week?
Our mileages vary.
PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.