Perplexing Light Bulbs

hard to say, try using your diode mode next time or switch the leads around.. could be galvanic.

"

It's got to be oxidation in a contact. Worst connection is to the screw-form, which is often simply press-fit. Shouldn't be much oxidation going on inside the envelope, as it's evacuated.

RL

On Sat, 06 Sep 2008 22:59:57 +0000, Mike Monett wrote:

: To All, : : I have a perplexing question on some plain incandescent light bulbs. : : I bought a package of four ordinary Sylvania 25 Watt 120 volt : candelabra light bulbs to act as current limiters in a HVDC supply. : The manufacturer's code is 25B10C/BL/4PK. : : Here's a picture: : :

: : After the purchase, I opened the package to check the filament : resistance. All four bulbs were open! : : I applied 120VAC and each bulb lit normally. Afterwards, each bulb : read about 46 ohms on a DVM. : : The next day, three of the four bulbs were open. After applying : 120VAC again, they all read about 46 ohms. : : I examined the bulbs carefully under a microscope. Everything seems : normal. The crimps at the filament look normal. The welds on the : lead wires seem fine. : : I removed the base from one of the bulbs to check if anything : strange was inside. There was nothing. The wires go straight into : the bulb. : : The bulbs all look and behave like any normal 120 Volt incandescent : light bulb when 120VAC is applied. : : But when low voltage DC is applied, the bulbs measure open circuit : until 120V is applied. Afterwards, they all measure close to 46 : ohms, but most go open circuit again after a day. : : It looks like the only place something could be going on is at the : filament crimp. But that seems to be normal and the same as every : other light bulb I have. The others act normally. : : Can anyone tell me what is going on? : : Best Regards, : : Mike Monett

Possibly oxidised lead wires to the insulated center contact or screw base contact points. The high voltage will break down the oxide layer resistance and lower the resistance to around normal. Suck the solder off these points and check, then clean and re-solder.

t =A0120 volt

.

he filament

ulb

applying

s

=A0on the

=A0if anything

into

t

uit

=A0to 46

the

every

I have no idea, but if you care you could measure resistance as a function of voltage and see when they turn on. is it at 2 volts or

100 volts. Also try another DVM.

"George Herold" Mike Monett

I have no idea, but if you care you could measure resistance as a function of voltage and see when they turn on. is it at 2 volts or

100 volts. Also try another DVM.

** Oxidised and sulphated contacts in switches and relays behave exactly like this too.

So you have never see that ?

..... Phil

I am suspecting that these "bulbs"/lamps have aluminum base shells. Aluminum typicaly has a thin oxide coating that may sometimes end up being an insulator to voltages to a volt or a couple, maybe a few volts or whatever,

especially when voltage is less than 2 volts and you make contact gently. You may get more realistic readings if you scratch/scrup/scrape the meter probe that contacts the "shell" part of the base. The "tip" of the base may have a slight need for "digging in" also when voltage is very low to make "full contact".

- Don Klipstein ( snipped-for-privacy@misty.com)

"legg"

** Maybe the maker should advise consumers that such lamps with dodgy internal pressure connections are in fact just like most CFLs - ie

" Non Dimmable"............

..... Phil

I've had just the opposite problem... screw light bulb into socket, doesn't light. Checks good with Ohmmeter.

Cause: Inadequate solder bead on center contact... as in "flat"

Solution: Get out the iron ;-)

...Jim Thompson

--
| James E.Thompson, P.E.                           |    mens     |
| Analog Innovations, Inc.                         |     et      |
| Analog/Mixed-Signal ASIC\'s and Discrete Systems  |    manus    |
| Phoenix, Arizona  85048    Skype: Contacts Only  |             |
| Voice:(480)460-2350  Fax: Available upon request |  Brass Rat  |
| E-mail Icon at http://www.analog-innovations.com |    1962     |
             
     It\'s what you learn, after you know it all, that counts.

[...]

Yes, I suspected the crimp at the filament. But I examined some other bulbs under the microscope and couldn't find any difference.

None of them were welded - the filament was held in place by folding the lead wire over the filament and closing the gap.

Of course, I emailed Sylvania asking them about the problem. As you point out, they have little reason to worry about this problem. I don't even expect a reply:)

There are other solutions, but the light bulb is useful since it lights up when there is a fault condition. So I used a small 25W bulb from a microwave oven instead. It works fine!

Thanks for your help.

Best Regards,

Mike Monett

Did you try cutting off the glass and measuring the resistance right on the filament, or is there something about a tungston filament that makes it hard to make good contact?

--

Reply in group, but if emailing add another
zero, and remove the last word.

Hi Tom,

I was considering that, but it would disturb the connection so much it would be hard to prove the results meant anything. Also, there would be no way to restore the original condition to prove we could re-create the original problem.

As mentioned, I did examine the crimp under a microscope and compared it with normal bulbs. There wasn't any difference that I could see.

The curious thing is all the Sylvania bulbs behaved exactly the same. They are normally open, and applying 120VAC causes them to read the same. They all read 46 ohms within 1/2 ohm. It is hard to see how a loose connection can be so consistent in all the bulbs.

Another problem is there are two crimp connections in each bulb. If it is the crimp, which one is doing it? Then, if it is one, why that one and not the other?

So we have to design an experiment that analyzes something we cannot reach, and tells us why the bulbs act this way.

I was thinking of applying a fairly low frequency, low amplitude signal to the connections to see if a capacitive probe could pick it up. Then trace it until it stops. That would require a narrow-band receiver to track a low-level signal from the probe. So I have to get my HP 8566 cleaned and give that a try.

Best Regards,

Mike Monett

No - my vote is on oxidation at the crimp interface.

This thread lead me to some interesting info on tungsten, funnily enough at:

Hi Den,

Thanks for the reply. Oxidation is a neat word, but it requires oxygen. If there were any inside the bulb, it would glow very bright and burn out. For example, when you break a bulb with power on.

According to Don Klipstein, the gas inside the bulb is probably 93% argon and 7% nitrogen. That adds up to 100%, so there's not much room for oxygen.

However, you are right. There is a problem at the crimp. I describe it in my reply to Tom Del Rosso.

Best Regards,

Mike Monett

Thanks, Don.

You are right. The base is aluminum, but my probes go right through any oxide coating.

I solder small sewing needles on the ends of the probes to make a sharp point. I can push this through wire insulation and measure continuity. It also digs through soldermask, and through any corrosion or oxidation on a contact.

I also use test leads with small alligator clips to grab and hold onto a lead. When the bulb has continuity, the resistance is pretty uniform at 46 ohms, so I think it's making good contact.

Best Regards,

Mike Monett

Hi Mike - I was thinking surface oxidation pre-assembly on either the filament or the support conductors.

I saw your reply to Tom, it's interesting the thought something as seemingly simple as a incandescent bulb can provoke.

regards d.

Tom,

I discovered that gently tapping the bulb on the table would make the connection open up. This gave me some courage to try your suggestion, and I cut the top off a bulb.

It was extremely difficult to make measurements without disturbing the filament. Eventually the connection closed and I could not get it to open again.

But without the glass, I was able to focus in on the crimp connection at max magnification. And I noticed something that escaped me before using low power.

One of the crimps is very tight. The wire folds back and touches itself.

The other crimp is quite open, with a gap the width of the wire.

And if you zoom in real close, you can see a brown powder residue at the crimp around the filament.

I read recently that an arc inside the bulb makes nitrogen or tungsten compounds, and some are brown. If this is correct, it probably explains the fragile connection.

I checked the other bulbs. All had one tight and one loose crimp, and two bulbs had a trace of the same kind of brown residue at the crimp.

So I think that solves the problem. The machine that made these bulbs is out of adjustment, and one crimp is loose.

Probably never happen again in a million years:)

Thanks for the suggestion. I think it led to the solution.

And thanks to all the others for your help.

Best Regards,

Mike Monett