[Amps] Arc distance

[Will Matney]

Dave,

If the path length was of no concern, they wouldn't need to set the gap on 
air capacitors due to the voltage they could withstand. The gap distance
is what determines the voltage rating, or where they will not arc under power.
Neon bulbs have a set gap to where they'll ignite at the voltage applied to them.
Spark gaps work in the same manner also.

Best,

Will

*********** REPLY SEPARATOR  ***********

On 7/28/06 at 1:19 PM Dave Haupt wrote:

>Ian, I always learn something from your longer
>postings.  Often it's a summary of bits and pieces I
>already knew, more often it's new data.  Thanks for
>this.
>
>I would concur with the notion that the path length is
>of very small concern when it comes to arcs.  During
>my "day job" I have occasion to be involved in the
>design and fabrication of thick-film circuits -
>precision attenuators, IC carriers and the like.  In
>our ceramics processing facility, we perform
>sputtering, both DC-enabled and RF-enabled.  It is not
>unusual to have flash arcs occur in the sputtering
>tank, whose walls are of translucent glass.  We can
>clearly witness the arcs occurring, and they seem
>never to occur where the two conductors come closest
>together.  We also have never actually seen an
>arc-mark afterwards.  It must depend on how much
>material is transferred during the arc.  It is also
>interesting to note that there is no gain in a
>sputtering tank - we have high voltage, but no
>amplification going on.  So, this is purely an arc in
>a high vacuum environment.  Generally, we "train" the
>tank each time we use it, allowing it to arc many
>times before it settles down.  Only after tank
>"training" as we call it, do we apply heat to the
>plating material, so it can vaporize and plate the
>ceramics which we are processing.
>
>Interesting bit of physics, all told!
>
>73,
>
>Dave W8NF
>
>
>==== in reply to ===
>
>There are two long-term sources of gas in tubes. One
>is a leak to 
>atmosphere, and the other is "outgassing" of
>structural materials.
>
>Leaks can be large or small, but they are always a
>one-way trip, with 
>only one end. Outgassing is more complex, and not
>necessarily fatal.
>
>Impurities in metals are the main source of
>outgassing. When the metals 
>are refined from ores, they always have built-in
>impurities. Materials 
>for use in vacuum tubes are extensively refined, by a
>combination of 
>chemical processing and heating in a vacuum furnace...
>but still some 
>impurities remain in the atomic lattice structures.
>
>There are also gas molecules chemically bound onto the
>exposed surfaces 
>of metals, ceramic and glass. These are largely
>removed in the later 
>stages of vacuum pumping, by heating the whole tube
>far above its normal 
>operating temperature while continuing to pump. The
>better the 
>materials, and the longer the time for which the tube
>is pumped, the 
>better the vacuum will be... but they can't pump and
>bake forever, so 
>eventually the tube has to be sealed off.
>
>At that point, the getter is activated. The getter is
>a chemically 
>active material that has been placed inside the tube
>to act as a kind of 
>'fly paper' for stray gas molecules that might appear
>in the future. The 
>getter in receiving tubes (and small glass
>transmitting tubes like the 
>807) is typically barium metal, which had been left
>inside the tube in a 
>little tray. On the production line, an induction coil
>heats up the 
>tray, evaporating the barium onto the glass as a
>slivery-looking film 
>with a highly reactive surface. This will continue to
>mop up stray gas 
>molecules for the life of the tube. (If that film has
>turned white, it 
>means there has been a gross air leak - the barium
>metal has turned to 
>oxide, and the tube is done for.)
>
>Transmitting tubes are different because they run much
>hotter and 
>operate at higher voltages, and a volatile metal like
>barium would 
>evaporate from where it had been deposited, and then
>condense in all the 
>wrong places. Instead, the getter materials are
>typically zirconium or 
>tantalum, which are non-volatile but *need* to run hot
>in order to 
>operate effectively. That is why the main getter in a
>glass tube like a 
>3-500Z is located on the metal anode (the grey surface
>finish is the 
>zirconium getter) and in a metal/ceramic tube it is
>located on the 
>heater (the next hottest location). Most transmitting
>tubes actually 
>have multiple getters to mop up the various kinds of
>gas molecules, 
>using different materials in locations at different
>temperatures.
>
>Immediately after manufacture, the vacuum will
>probably be about 10^-8 
>mmHg, which is really quite good for a routine
>production-line process, 
>but no great shakes by the standards of a vacuum lab.
>At this standard 
>of vacuum, a typical tube may contain anything between
>a million and a 
>billion gas molecules! (PV=nRT... work it out)
>
>Most of the time, a "vacuum" tube operates perfectly
>well in spite of 
>sharing the space with very large numbers of gas
>molecules. But 
>sometimes you need to remember that the tube is also a
>reaction vessel 
>for some complex low-pressure chemistry.
>
>Coming back to impurities... immediately after
>manufacture, the vacuum 
>is probably pretty good because all the surface
>impurities were flashed 
>off. However, impurities that were trapped deeper
>inside the metal can 
>continue to diffuse to the surface over the lifetime
>of the tube, and 
>can be released as gas into the "vacuum" space causing
>a small increase 
>in pressure.
>
>If the getter is active, it will mop up the impurities
>within typically 
>a few seconds (determined by the time it takes for the
>molecules to 
>bounce around until they strike the getter surface,
>and by the 
>probability that a molecule will hit a chemically
>active spot that can 
>form a strong enough bond to make it stick). But a few
>seconds is far 
>too slow to prevent an arc, which can strike within
>microseconds if all 
>the other conditions are right.
>
>This explains why tubes can arc for no apparent
>reason, but if you try 
>again a short time later, the tube goes back to normal
>as if nothing had 
>happened. (Obviously this requires an amplifier with
>good HV surge 
>limiting and fast shutdown protection. If the arc is
>allowed to persist 
>at high current, it will damage both the tube and the
>amp.)
>
>It also explains why transmitting tubes generally need
>to be pre-heated 
>after a long period out of use. The process of slow
>diffusion to the 
>surface of the materials means that gas will probably
>have accumulated, 
>and the getter needs some time at a high temperature
>in order to do its 
>job.
>
>-- 
>73 from Ian GM3SEK
>
>__________________________________________________
>Do You Yahoo!?
>Tired of spam?  Yahoo! Mail has the best spam protection around 
>http://mail.yahoo.com 
>_______________________________________________
>Amps mailing list
>Amps at contesting.com
>http://lists.contesting.com/mailman/listinfo/amps