A Perfect Vacuum... Pump

The previous entry in this blog was 'Screw This!', so I'm surprised this entry wasn't entitled 'This Sucks!'.

If the next blog entry concerns fans or air compressors I'd recommend the title 'This Blows!'

_{/Not that consistency is required; no hobgoblins around here.}

I used to work on high vacuum pumps of all types. The first thing I learned was: 1)Vacuums don't suck and 2)Pumps don't apply force IN the evacuating space!

A vacuum pump is grossly misnamed. What it is pumping into is the atmosphere! What it is pumping out into the atmosphere is less and less gas. The evacuated space is not relevant! What is relevant is the surface area of the evacuated space, the temperature of that surface, and the number of particles of gas present within that space, the ratio of the aperture of the trap to the surface area of the chamber, and the difference in pressure that can be held at the aperture by the trap. The space is empty except for the atoms to be evacuated. And those atoms do what they want in that space, in spite of the "pump". The proper term is "Trap", not "pump".

All vacuum systems trap atoms and then either move them into another trap, (so called pumping) or keep them still and in one place. A cold trap removes velocity from the atoms and "traps" them, reducing the number atomic impacts on the surface of the chamber, and thus reducing the apparent "pressure" of the evacuated space. A "Getter", which is made of baked out (at low pressure) titanium that is given an electrostatic charge, which traps atoms of gas within the matrix of Titanium atoms. A small piece of Titanium about 2" x 2" x 1/4" will hold, at high voltage, about 2 liters of air at normal atmospheric pressure.

A Crooks Tube, uses drops of Mercury and gravity to trap atoms in a glass tube held vertically. The Mercury is kept cool and just above the solidification point so that it remains liquid on the chamber side of the pressure differential that forms when it drops into the top opening of the glass tube. An Oil diffusion pump traps atoms on one side of a sheet of high temperature, high velocity, oil. The Oil is heavier than the atoms it is trapping, and so it is like a freight train changing the direction the atoms are moving. The trap is that the movement is always down and away from the chamber, much like the Crooks tube.
A mechanical pump is the least capable of achieving a high vacuum in the chamber, but it is the most capable when it comes to holding atmospheric pressure. It is mechanically pressurizing the air outside the chamber, and when used in series with Oil Diffusion or Crooks Traps, it will allow for the maintaining of a much lower differential pressure. That is why it is called a "rough" pump. It produces the baseline differential that the more delicate traps need to function. An Oil diffusion pump or Crooks tube will have a difference of pressure that is easier to measure in atoms than it is to measure PSI! There is only a handful of atoms on one side or the other in a trap at high vacuum conditions.
The mechanical pump on the other hand is capable of creating pressure in PSI on the atmospheric side of the system. (I have an old oil-less compressor from an air conditioner that I am going to use as a roughing pump. All I have to do is hook up it's low pressure side to the chamber and/or the trap and let the high pressure side be open to the atmosphere. This is sufficient to get a Neon or Argon tube to light up.)
Measurement of "pressure" at high vacuum is not really possible. Instead, electric charge is usually applied to a region of the chamber space, and the atoms are recorded when they deposit the charge on the measurement devices positive plate. Very accurate low current measurement between the chamber and the positive plate is what is needed to estimate "pressure" in a high vacuum system. Also, if the measuring device is also a trap, then lower "pressures" can be read because of the tendency of the trap to focus the charge carriers, the atoms, onto the positive plate of the gauge.
Remember that the atoms are doing what they want, until they hit an object within the chamber. When they hit, they can be slowed, sped up, charged, or discharged, or trapped mechanically or otherwise, but they cannot be "pumped" from the chamber. They are moving at random within the chamber and have "odds" of hitting where you want them to. The aperture of the trap is a target. The bigger it is, and the more capable it is of maintain a difference in the number of atoms between the high side and the low side, the more efficient and capable your Vacuum system.
Vacuums don't "suck" and Vacuum Pumps don't "pump" anything, they "trap" gas. Pressure is gas in motion, and a Vacuum is the lack of pressure.

"A small piece of Titanium about 2" x 2" x 1/4" will hold, at high voltage, about 2 liters of air at normal atmospheric pressure"

Why the substantial thickness? Does this imply that the titanium is porous, giving it greater surface area for improved adsorption? The electric field won't be present in the interior of course thanks to Faraday self-shielding, so is it present only to attract the molecules/atoms ('particles' in short) to the piece? If so, what ionizes them? Neutral particles won't be attracted. This is an adsorption trap, correct? Once adsorbed, is the high voltage still needed? If not, must the piece be heated to drive off the trapped particles?

Just curious how this works. Two litres is pretty impressive.

Yes yes and yes. It is an adsorption trap, the voltage must be constant, once removed it will start to out-gas slowly, and the titanium atomic matrix is very very spacious.

These types of traps are really large "getters" and they must be baked out under high vacuum to expel the gas collected. The gas is charged by contact with the walls of the chamber and the getter. The getter is held at a high positive voltage and the atoms tunnel into it's surface when they accelerate towards the charge of the plate.

I'm sorry, I meant to say that the charge is received by the atoms from the walls of the chamber and the outer wall of the getter.

The Titanium is porous by it's own nature but using sintered powders increases the collection surface. It is true that the inner areas are shielded like you said, but they are not shielded from the atomic migration of the gasses collected, nor are they shielded from the gasses arriving at the surface. The gas will migrate further into the matrix. Internally, it would seem there is no charge, but then, neither are there many electrons except for the ones just delivered via the gases that just arrived.

"Just curious how this works. Two litres is pretty impressive."

Agreed. Entrained at STP within 1 cubic inch of Ti is 2 liters (about 122 in³) of air when charged "...at high voltage...".

[edit] I got sidetracked while composing, didn't see Deefburger's response until now.

I had a hard time believing it myself when I was told what was going on. But I can tell you that I may have miss heard or understood that figure at the time. These were used on two-liter sized tubes and the Engineer who taught me may have meant two liters at 4 millitor! In any case, the amount of gas that can be held is still quite impressive.

One of my tasks was getting it back out! That way the Ti could be reused. But this took hours at high temp and low pressure and I would swear there was a never ending supply of gas in those little tiles!

The density of dry air at 0 °C & 760 mm Hg is 1.2929 g/liter and so if your initial state were 2 liters ATM at STP, the tile would be about 2.58 g heavier when the trapping cycle was complete.

If it were 4 mTorr, good luck!

Yeah, you are right. In any case these things are used to evacuate large Klystrons and TWT's and have been for years. They seem to work and they never seem to run out of gas when you are recycling them!

I have an Eimac 450TH in brand-new condition and still in its original shipping carton (with horsehair packing). Shipping date: April, 1958.

Curiously, it has no getter. No Eimac tube/valve I've ever seen has a getter. What gives?

I'm guessing they did something clever with the plate structure.

Getters were added to much smaller tubes because the smaller volume allowed a much higher pressure to form when the tube was operated. The larger tubes of old were made with large volumes that would not see much degradation of performace when the cathode was heated and raised the pressure. Also, I don't think Titanium was much available at the time the tube was built.

Very cool piece to have though! One of the engineers I worked with years ago had a CRT that he had built himself on his bookshelf. He built it in a lab in Scotland where they were developing television, sometime in the Thirties! His tube didn't have a getter either but it sure was pretty!

A very wealthy friend in San Antonio, Texas, is an avid collector of all sorts of scientific apparatus - and stuff from movie sets. He owns the original model of the Nautilus used in the filming of 20,000 Leagues Under the Sea, for example. It's about three feet long. Also Farnsworth's first television apparatus, not one, but two Enigma machines and their counterparts, a number of de Forest's early vacuum valves, and other really cool stuff. Not bad for someone who makes $35 million/mo (digital television and image transmission patents, among other things). I showed him my Eimac 450TH. He wanted it. "No way!"

I used to have a couple of ferrite bead magnetic memory from a VAX 1170. It was 16K per board and each board was 1x1 foot card edge connected. The memory itself was a cloth woven from coper magnet wire with a ferret bead in each and every crossing of the weave! It was beautiful! 16KB of beads in one cloth memory. Beautiful and complex but also so beautiful and simple by todays standards. I don't think I could power 2GB of memory made from these cards!