Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!samsung!sol.ctr.columbia.edu!emory!rsiatl!jgd From: jgd@Dixie.Com (John G. DeArmond) Newsgroups: sci.electronics Subject: Re: incandescent light bulb life extender Message-ID: <5297@rsiatl.Dixie.Com> Date: 16 Dec 90 12:04:00 GMT References: <4247@kitty.UUCP> Organization: Rapid Deployment Systems (making go-fast things and things that-go fast) Lines: 270 We all got to see another one of Larry's little mental explosions. Larry's obsession with my life has obviously become an illness. It is unfortunate that such people are allowed about. Nontheless they are so we have to live with it. I won't bore the group with a a point by point rebuttal. I will address only the technical issues the Lid raised and then opinionate only a bit at the end. [An amazing and most likely fabricated investigative quest description deleted.] > Likewise, I had not previously heard of any fires caused by the >UL-approved devices, which is why in my original article I stated that UL >considered them safe because they were confined within the metal barrier >of the lamp socket. I never claimed that the rectifier buttons had caused any fires and I don't believe the other fellow did either; I simply stated that the ones I tried in my fixtures had burned to a crisp. Since there was nothing flammable near the socket other than the lining of the socket itself, it would be hard to imagine a fire scenario. *BUT*, the fact that the buttons burned up - I did save the residue to show people - tells *ME* that there is a potential to start fires. And to that conclusion, I don't give a damn WHAT UL has to say about it. > It may also help readers to know that typical base temperatures >of incandescent lamps in the range of 60 to 100 watts run from 194 to >208 deg F (data taken from NFPA handbook). As usual, Larry backs up his arguments with rote dissertations from a book. Since "typical" rarely occurs outside the printed page, I prefer to make in situ measurements of the actual conditions involved. I'll allow the reader to determine whether "typical" values from a text or measured values are more appropriate in a given situation. > What Mr. DeArmond failed to find in his Digi-Key catalog, because >it was somewhat buried at the upper right hand corner of a page which >otherwise listed varistors, was the one series which actually is suitable >for power inrush limiting. Those thermistors, the Keystone CL-series, >have 18 AWG and 22 AWG wire leads, are quite suitable for lamp or power >applications, and have explicitly specified current ratings from 1.1 to >12 amps. The best part is that at full rated current THEY DON'T GET ANY >HOTTER THAN 25 DEG C. Yes, the Keystone parts are quite adequate, but the panasonic ones are too. I'm not arguing against the Keystone or any other similiar part, I'm arguing based on actual experimental results that the Panasonic part works quite well. I was originally attracted to the 8 ohm part as a result of seeing it used in an inrush application. >He'll just tell >the insurance company that he made measurements with a gen-U-wine >FLUKE 8800A 5-1/2 digit DVM, that while 12 years old and obsolete, >still has NBS-traceable calibration! :-) Is the implication here that one should discard fully functioning instrumentation that meets the manufactur's specifications and has a traceable calibration just because it accumulates some age? Is that what you're really trying to say, Larry? Or are you saying that new is better regardless of its calibration? Do you think that a voltage measurement made to 4 decimal places with an instrument that is within mfr's specifications is bogus because the instrument is old? I would have expected much better for someone who purports to be technical. > A filament short is *usually* blown away because there is no >device in series with the lamp to create impedance and limit current. >A series thermistor, even in the hot state, may well limit current such >that neither the shorted filament nor thermistor itself will have >enough energy to instantaneously open. The actual instantaneous >fault current available from a typical 20 ampere electrical branch >circuit may reach the hundreds of ampere range, with a low resistance >fault within a lamp usually blowing out in several cycles - which is >usually not long enough to trip an overcurrent device. It doesn't take >much imagination to see how even 1 ohm of series resistance can drastically >reduce current such that the energy necessary to open an internal lamp >fault is no longer available. But if your imagination is true, then any lighting circuit that contains a dimmer circuit is a fire hazard. After all, dimmers contain active devices with no more physical ruggedness or surge handling capability than the thermistor in question. Let's look at a hard example. I pulled the dimmer from a fixture in my house and looked at it. It contains a device that cross references to a GE SC-146 triac that is packaged in a TO-220 package. The GE SCR manual rates this device at 10 amps continuous. The dimmer is a 1000 watt unit which is larger than the normal 600 watt units which would probably use the equivalent of a GE SC-241. The data sheet gives a maximum peak one cycle non-rep surge current of 100 amps. In other words, this is the maximum current the device would be expected to handle for one cycle of AC current in isolation from all others. This is the value used to evaluate triacs for inrush and filament short circuit duty. On page 245 of the same manual, it recommends that for a 100 watt bulb, the inrush current design criteria be abtou 17 amps. Thus this triac gives very good margins for inrush but it would fail miserably in Larry's direct short, multi-hundred amp scenario. But the failure rate for lamp dimmers is insignificant. What's wrong with this picture? Could it be that Larry's scenario is incredable? I think so. If we saw any significant instances of long term (defined as many cycles) direct short circuits, the fatality rate for lamp dimmers would probably get the attention of Ralph Nader :-) It would be reasonable to postulate that the vast majority of lamp failures involve filament short currents of less than about a hundred amps for only a cycle or two. > Consider this example, which should be intuitive to many readers. >A 2" piece of 28 AWG wire connected to the 18 AWG conductors of a line cord >will probably blow away in a fraction of a second if the line cord were >plugged into an AC outlet. If an appropriate series impedance that limited >current were now added to the circuit, the 28 AWG wire would first become >red hot (perhaps for a long time) before opening. Also, that series >impedance might itself get rather hot in the process. After thinking about it for a minute, I decided to make another test. I took a PNT-124 8 ohm thermister out to my metal shop and connected it to my DC welder power supply. A Gen-U-wine Fluke amp-clamp was applied to the secondary lead before the rectifier in order to measure the peak current. The volt meter in this case was an IES 2360 (brand new but uncertified - that ought to make the Lid happy :-) because it has a peak hold feature. The thermister was clamped directly between the two welding clamps so as to reduce the circuit impedance. I set the welder control to 150 amps and pushed the ON button... The thermister "flashed" into oblivion. "Flashed" is the best word to describe the action. There was a brief ball of flame similiar to that from the muzzle of a gun. The device pretty much disapeared. The leads were left clamped into the welding clamps. The tinplate on the leads was slightly discolored but not so much as to expect it got red hot. The interesting thing is that there are still shards of solder on the end where the thermistor used to be. That would explain the flash. The solder attachment points failed and vaporized in a ball of fire. The peak current registered was about 230 amps. I say "about" because I don't know the single cycle behavior of this meter and because it is not certified. I'll let the reader determine if this failure mode, which is similiar to many solid state components when grossly overloaded, is hazardous in the context of a lamp socket. I do agree that in the event a person were to place sufficient impedance in a lamp circuit such that the current could only peak at 20 amps or so, that it would be a bad idea to use ANY thermistor in the circuit. >> and because we could speculate that a lamp construction that would allow >> a continuous short between electrodes could never be approved, > > Mr. DeArmond should take apart a few different models of light >bulbs and see how filament support wires can well be deflected against >broken filament sections to create low resistance faults in a lamp failure. >Such light bulbs are all UL-approved. As usual, Larry did not address the issue. A connection established across part of a filament merely establishes a circuit through a filament that is now grossly overvoltaged. This filamnet will fail instaneously and with quite a show of pyrotechnics. Let's assume that a filament could establish contact with the opposite electrode over as little as a fourth of its length. That would mean that its cold resistance would be 1/4 that of normal and the inrush would be 4 times normal. Using GE's figure of 17 amps for a 100 watt bulb, that would give us an inrush of 68 amps. This is still well below the point that even the triac discussed above would fail, assuming the filament blew away in one or two cycles. Since the filament would heat vastly more rapidly than normal, the failure would clear the fault more rapidly than the normal inrush interval. From the discussions in the GE SCR manual, planning on filament clearance times of one or two cycles is reasonable. > Yet, Mr. DeArmond says: "The part will rise about 60 degrees C >over ambient when in operation". I believe he means room temperature >ambient here, but even if he means fixture ambient, the end result is >the same: component failure. And yet they don't fail. Is it magic? I think not. It is the result of careful experimentation and design. Learn how to read a spec sheet, Larry, and then do some experimental work to verify the applicability of the datasheet. Learn what the term "specsmanship" means. Be creative and never take a spec sheet at face value. You might slip up and find a use for something that the manufacturer never imagined. [another mental masturbation deleted.] > The truth is, the 600 mw rating is for *best* conditions when >the thermistor temperature is heat-sinked at 25 deg C. Less severe thermal >conditions, my ass! Maybe this article will cause Mr. DeArmond to grow up >and realize that he cannot always bluff his way through life - especially >when he pretends to have engineering knowledge that he in fact lacks. If you think my data is so bogus Larry, why don't you simply repeat the tests. If you've got enough time to prowl around in my personal life, you've obviously got enough time to set up a light bulb and a digital pryometer and report your own results. We know why you don't, of course. You'd rather masturbate in public and sling mud. There's some growing up to be done but it's not on this end. Tell you what, guy. Why don't we just cut to the chase and be done with it. I'll put my experimental results up for 3rd party verification and make the results available here. Let's see if you've got the guts to duplicate the experiments and do the same. I'll even let you use a NEW voltmeter. ----------------- EDITORIAL OPINION What follows is strictly my opinion. If you don't want to read further, go to the next article. Let's look at the extremes to which Larry has gone in order to satisfy his fettish. He has bought a CD ROM version of the net at no small cost in order to be able to conduct his hate-driven vendetta against me and a couple of others on the net. That in itself is sick. But he goes further. He posts "facts" about my personal life that are just plain out and out lies. Everything from the trivial such as the size of my house and office and what I paid for it to the rather serious such as slandering my company, my financial status and my family. Did you know, for example, that I've gotten reports of his contacting former clients and spewing this filth? It is probably a waste of time to even attempt to address these lies so I won't. But if anyone has questions about a particular instance, please email and we'll discuss it. I suppose it is to be expected that successful people make a few enemies. There are people in this world who are so insecure that they terribly resent others who have bootstrapped themselves up under adverse conditions. They like to rationalize others' successes and their failures with excuses like "If only I had as much {money, education, good looks, balls, etc} as that guy, I'd be successful too". They tend to view others through the distortions of their own character flaws. Thus with Larry. The only thing he has achieved in his life is having Daddy put him through college so he terribly resents those successful people who did not do so. I quite possibly would be more successful than I am now with a full college education but we'll never know. At least I'm honest enough to admit my lack of that credential. One does, however have to wonder about an obcession with something that in my case would have happened almost 20 years ago. When I would have been in college, the 8008 would have been the rage :-) Isn't that scary. I wonder how relevant that education would be today? Perhaps I should get mad and get a lawyer or worse. Maybe I will before it's over. I'll admit that I did get mad several months ago in the beginning of his fit. Normal people just don't act like Larry does so he caught me off guard. But what I feel right now is pity. Pity for a person with such a distorted phyche that he must try to build self-respect by attacking others. I'd imagine that he feels terrible guilt from having had it all given to him by Daddy and anger because he cannot find a way to enjoy it. Daddy was probably authoritarian and this is Larry's way of striking back. I initially wondered what it could have been that sent him off into such a frenzied rage starting over in comp.dcom.telecom a few months ago. I'd never had a conversation, email, net or otherwise with the guy. What I discovered as I began to swap email with other of Larry's victims is this model of his personality that I've described here. The really sad part is that he could probably build that self-respect AND the respect of others strictly on his own. Perhaps it's not too late. Larry, I do hope you get help with your problem. You'll live a much happier life for it. We'll be rooting for you. John -- John De Armond, WD4OQC | "Purveyors of speed to the Trade" (tm) Rapid Deployment System, Inc. | Home of the Nidgets (tm) Marietta, Ga | {emory,uunet}!rsiatl!jgd | "Vote early, Vote often"