VHF Tungston bulb noise

Oh, sure... Blame that on on W also.

You do realize that as a metal becomes up some electrons break away from it to be reabsorbed(due to the dipole created)?

When a conductor is heated the electrons can gain enough energy to "break away" from the metal but generally will be reabsorbed. Tubes are based on this principle but use an external voltage to prevent reabsorption. Without an external field the electrons will create an AC field due to the continuous escaping and recapturing. The effect is generally random and very negligible due to cancellation.

When the filament is wound as a coil though an asymmetry is created making a resonate system that favors a certain frequency ranges.

So there are essentially two parts: First is the fact that radio waves can be transmitted due to the electron cloud around the filament. Second is the RLC network created by the coiled filament.

hard to believe that there is any RF emission from a Tungsten filament lamp. I've never seen any emission on my spectrum analyzer (10MHz - 1GHz). For that matter, I've been at several EMC certification facilities. The shielded chambers, Faraday cages and 10M test sites all use filament lamps. None of them were shielded.

Now, having worked with Xenon arc lamps, THAT is a wideband RF emitter. Mecury arc lamps were fairly quiet by comparison.

There was an article on the subject in Wireless World in the early

From memory, they tracked the problem down to a partially-conductive layer deposited on the inside of the glass during the later stages of the lamp's life, from filament evaporation. The diode action between this and the filament charged the layer to the +ve peak of the mains voltage.

When the voltage of the filament reversed, the electron cloud was repelled by the charged layer and set up Barkhausen oscillations in the potential gradient around one end of the filament. The action was similar to the Barkhausen oscillations which occurred in power tetrodes being operated with a pulse waveform and a 'ringing' inductive load (which caused early television line output valves to emit VHF interference).

In the case of light bulbs, the carbon filament vacuum lamps were found to be the worst offenders; gas-filled lamps did not suffer from it.

That has been observed in the past with low wattage bulbs that use a vacuum filling rather than Argon.

You get thermionic emission from the heated tungsten an electron current flows between the parts of the filament at high voltage and the parts at low voltage (i.e. it creates a biased vacuum diode). If conditions are right you can get Barkhausen oscillations that can then radiate. These oscillations are modulated by the AC input causing the observed effects on the TV screen.

kevin

Now _that_ is a riot. Thanks, Adrian.

Of course modern bulbs all have argon fill, so it isn't likely to be that. Tungsten bulbs are about as quiet as it gets, at least above the first few dozen harmonics of 60 Hz.

Cheers

Phil Hobbs

ak

on

the

ing

can

mp.

ded

Nah, that's nothing. Take a look at a standard metal-halide lamp with an aging transformer.

What are Barkhausen oscillations? Wiki says when you have enough gain in the feedback loop... But that describes many types of oscillators.

TIA, George H.

t -

Virbrations can couple into filament bulbs... but that's also low frequency.

George H.

Electron transit time oscillations due to an instability in collisionless plasmas. See e.g.

Cheers

Phil Hobbs

s a

he

t -

Thanks Phil, I'm not much on tubes... Transit time oscillations, remind me of a reflex klystron. (changing repler voltages)

George H.

I'm not a tube guy either--I had to look it up myself.

Cheers

Phil Hobbs

Oscillations at VHF/UHF in power tetrodes/pentodes, employing high voltages, such as television horizontal deflection tubes, due to electron transit time effects. Sort of like an unintentional klystron.

Not to be confused with "Barkhausen Effect" in magnetics.

I once had occasion to examine a 500W halogen lamp. At the low voltage used on the ohms range of a DMM, it would appear open circuit. Using a curve tracer, it became evident that it behaved as two series-opposed diodes, with a forward voltage of a couple of volts. Applying rated voltage made it behave normally. This was traced to the method of attaching the filament, crimping, rather than spot welding, creating a point-contact diode at either end. Since the problem was related to production testing at low voltage, that's as far as it went. One day, I might repeat the experiment and look at things with a spectrum analyzer.

Transition metal tunnel junctions do funny things like that. That's also why wiggling a BNC connector will change the contact resistance in series with the shield.

Cheers

Phil Hobbs

I worked my way through high school in a TV shop in the late '50s. Every now and again we'd get a set it with a ripper of a Bark oscillation on some channels.

Ion traps (simple clip-on magnets) were quite common on the CRT to keep ions from burning the phosphor. One of those suckers clipped to the horizontal output tube and then rotated until the Barks went away really worked quite well.

Jim

Not a few TV receivers of the 1960s incorporated "anti-Barkhausen" chokes in the anode connector of the horizontal output tube. A microhenry or two, air cored, attached to the top cap connector. They didn't always work.

The common element here is the socket / lampholder, that is what i would look at.

?-)