A Vacuum Tube Is a Particle Accelerator You Can Hold in Your Hand

Pull a 6L6 out of its socket and hold it in your palm.  Let it cool off first!

Glass, a few wires, a small black box of nickel and mica inside.  It looks like a piece of mid-century lab glassware.  It is a piece of mid-century lab glassware.  It's also, as far as I can tell, the same machine as the Large Hadron Collider.

The 6L6 beam tetrode: a source of charged particles, a vacuum, an electric field across a gap, a pair of plates to aim the beam, and a target.  Scale those ingredients up by a factor of a billion and you get CERN, the European lab that runs the LHC.

The unit of the trade

Last post, we worked out that the electrons in a 6V6 with 340 volts on the plate reach about 4% of the speed of light.  That’s a 12W tube in a small tweed Fender.  A Music Man HD150 runs its 30W 6L6GCs at 700 volts: 5.2% of c, almost ten thousand miles per second.

The speeds are fun to think about, but what really matters is what the electron does when it lands: heat the metal, punch through it, or kick out an X-ray.  This is kinetic energy, measured in electron volts.  One electron volt is the energy a particle gains falling through one volt.  At 700 volts on the plate, every electron lands with exactly 700 eV.  The plate voltage you read with your DMM is also the energy gauge.

At CERN it's protons instead of electrons, but the unit is the same.  The LHC delivers 6.8 TeV per beam for 13.6 TeV total collision energy—nearly fourteen trillion of the same little units a 6L6 hands out seven hundred at a time.¹

Once we have that unit, every accelerator lines up on a beam.

The tubes in your amp sit at the small, quiet end of a very long line.

What holds it together

From a 6L6 to CERN, the same three parts do the work.

The vacuum.  At atmospheric pressure, an electron moving at 4% of light speed wouldn't make it a millimeter before colliding with a nitrogen molecule.  The glass envelope on your tube holds the inside at roughly 10⁻⁶ to 10⁻⁷ torr, a hard vacuum by most definitions.  The LHC beam pipe goes much further: about 10⁻¹⁰ to 10⁻¹¹ torr, three to five orders of magnitude harder.² The difference comes down to time and distance.  An electron in a 6L6 only has to cross a few millimeters of gap, so the vacuum just needs a mean free path longer than that.  A proton in the LHC circulates for hours, covering enough distance to reach Pluto, and any stray gas molecule in the beam pipe is a collision that degrades the beam.  Both machines need some nothing inside them.  The LHC just needs a lot more of it.

The cathode.  A research electron gun heats a piece of specially coated metal until electrons leave the surface by thermionic emission.  So does your 6L6.  The orange glow inside the tube is a filament warming the cathode—a nickel cylinder—to something like fourteen hundred degrees Fahrenheit until electrons boil off on their own.  A klystron or a CRT does it the same way.

The accelerating field.  A few hundred volts in a guitar amp, a few hundred million in a research linac.  Put a static electric field across a vacuum gap and any charged particle in it gets pulled across.  The tube holds its whole accelerating voltage across one short gap. A big accelerator never has that much in any one place; it reuses a modest kick over and over instead.

The part SLAC doesn't have

A research electron gun needs a tight, coherent beam aimed at a fixed target.  Your guitar amp doesn't.  It requires a beam that can be turned up and down by a few volts of signal, very quickly, with a clean response.  That's what the control grid is for—the fine spiral of wire sitting in the path of electrons leaving the cathode.

A few volts on the grid changes how many electrons make it through on their way to the plate.  More negative grid voltage chokes off the stream.  Less negative lets more through.  That's a tube amplifier: a few volts of audio on a wire, steering tens of milliamps of high-voltage current behind it.  Voltage in, current out. That's what makes it an amplifier and not just an electron gun.  (Rectifier tubes skip the grid entirely—they're pure accelerators.)

SLAC, the electron accelerator at Stanford, shapes its beam with two miles of magnets and RF cavities.  The 6L6 has no magnets, but it isn't entirely without beam optics.  The beam-forming plates that give the tube its name sit on either side of the electron stream, using electrostatic repulsion to squeeze the electrons into focused ribbons aimed at the plate.  Just two bent pieces of metal steering an electron beam through a vacuum.³

In a research electron gun, the kind used in electron microscopes, or as the injector for a particle accelerator, the part doing the same job is called a Wehnelt cylinder: a negatively charged, cup-shaped electrode surrounding the cathode, squeezing the beam through a central aperture by electrostatic repulsion.  Different geometry, far more beam energy, but the same idea: a charged piece of metal near the cathode deciding where the electrons go.⁴

X-rays on stage

If you accelerate electrons fast enough and crash them into metal, some of their kinetic energy comes out as X-rays.  It’s how every X-ray tube works.  The threshold for useful medical imaging is around 20 to 30 keV.

My Tektronix 544 oscilloscope runs 24,000 volts on the CRT anode. The standard 544 ran 10 kV, but mine has the factory 108G fast-sweep modification that brought the accelerating supply up to 24 kV.⁵  It also came with a leaded glass faceplate.  Tek didn't spec leaded glass on a piece of test equipment for aesthetics.

A typical guitar amp maxes out around 500 volts.  An electron that falls through 500 volts arrives with 500 eV, and when it slams into the plate, the most energetic X-ray it can throw off is 500 eV—one photon carrying everything the electron had.  (Physicists call this ceiling the Duane–Hunt limit.)⁶  Turning that energy into a wavelength requires an equation, the same one that makes blue light more energetic than red.  A photon's energy is hc divided by its wavelength, where h is Planck's constant and c the speed of light:

E = hc / λ,  so  λ = hc / E

hc is a fixed number, and in the units we're working in—electron-volts and nanometers—it comes out to about 1240 eV·nm.⁷  So a 500 eV photon lands at 1240 / 500 ≈ 2.5 nm, just inside the soft X-ray band.  But at 500 eV almost none of the energy comes out as photons at all, and what little does is stopped within a micrometer or so of ordinary glass.  The envelope and the chassis absorb it long before it reaches you.  You can play a tube amp every day for sixty years and not pick up a measurable dose.⁸

Still, X-rays from a guitar amp! That tells you something about what's going on inside the glass.

The thing in your palm

Take another look at the 6L6.  Inside the glass: a heater, a coated cathode, a control grid, a screen grid, a pair of beam-forming plates, and an anode, all in a few cubic centimeters of vacuum.

When it's running, the cathode sits at fourteen hundred degrees and the plate takes something on the order of 10¹⁷ electrons a second, each one reaching a few percent of the speed of light before it lands.  The plate sometimes glows dull cherry because that bombardment energy has nowhere else to go: thirty-plus watts of kinetic energy turning into heat inside a piece of metal the size of a postage stamp.  The whole apparatus sits an inch from another one just like it, under a steel chassis, inside a wooden cabinet, in the back of someone's car on the way to a bar gig.

You can hold it in your hand.  It's still the same machine they have at CERN. Just smaller, slower, and tuned for making music instead of making matter.

---

Sources

1. CERN, "Facts and figures about the LHC," home.cern

2. "Taking a closer look at LHC — High vacuum," lhc-closer.es

3. O.H. Schade, "Beam Power Tubes," Proceedings of the IRE, Vol. 26, No. 2, Feb. 1938

4. Wehnelt cylinder: Kimball Physics electron gun tutorial, kimballphysics.com

5. "544," TekWiki (w140.com/tekwiki)

6. W. Duane and F. L. Hunt, "On X-Ray Wave-Lengths," Physical Review, Vol. 6 (1915), p. 166

7. NIST, "Fundamental Physical Constants" (CODATA), physics.nist.gov

8. H. Pöttgen, A. Schirmer, M. Port, and R. Nusshardt, "Emission of Parasitic X Rays of Vacuum-electron Tubes with Glass Housings: Implications for the Evaluation of Occupational Doses," Radiation Research, Vol. 201, No. 5 (2024), pp. 499–503