Connect with us

LEDs + Lighting

These Pumps Suck!

Selecting the right secondary vacuum pump

Published

on

It’s been several years since the neon industry’s had a hot discussion about nothing. By "nothing," I’m referring to the lack of air, or vacuum, inside a neon tube. Or, specifically, how to achieve the vacuum necessary for a quality neon tube.

Because so many people have entered the adventurous, proud business of turning glass tubing into neon lamps, with little discourse on the subject, it’s time for a short review of neon-pumping principles.

The science of absence

You can fabricate a quality, long-lasting, neon tube by exclusively using a double-stage, rotary-vane pump. I’ve used such a setup, and the tubes are still perfect after more than 35,000 hours of operation. But, the pump was quite oversized and perfectly maintained, so I could achieve the required pumping speed at 1 micron, which is limited by the tubulation’s length and diameter.

The limit is 0.2 liters per second, or 12 l/min. for a 4-in.-long piece of 6mm tubulation (High-Vacuum Techniques, page 129, by James Yarwood, Chapman & Hall, London, 1948). Fig. 1 shows some pumping speeds of mechanical pumps.

At 1 micron (10-3 Torr), a Megavac pump running at 320 rpm is just reaching the speed of 0.2 l/sec. This is insufficient for fabricating quality neon tubes, because every vacuum system loses pumping speed due to dead volume, restrictions in conductivity and other factors.

Further, these figures are only valid for a brand-new, perfect system. Even in the first widespread publication about the industry, 1935’s Neon Signs, the precursor to today’s Neon Techniques and Handling. Samuel Miller mentioned the necessity of a secondary pump to obtain a proper pumping speed at the tube connection.

He wrote: "Very few neon plants today depend on a mechanical pump alone. Almost every neon manufacturer has come to realize that some sort of diffusion pump is necessary for a high-speed pumping system." This statement holds true today, and debate occurs every few years about which type of diffusion pump is best suited for neon.

Premium suckers

From a physicist’s point of view, the best way to achieve a fast, "clean" vacuum without any backstreaming, in the range needed for neon fabrication, 1 to 10-4 mbar, is the molecular drag/turbo pump. However, it’s also the most expensive. Molecular drag pumps that operate at 50 l/sec., including power supply, will vacuum $5,000 or more from your wallet. That’s a big gulp. Further, they’re sensitive to incoming rushes of air and particles, which may scrap them in seconds.

A diffusion pump requires much less investment. However, not all diffusion pumps are created equal; different construction principles create different pumping speeds in various vacuum ranges. As I’ve said, neon needs a major pumping speed at the end of the bombarding cycle, when the pressure in the tube is approximately 0.5 mbar, and has to be brought down to 0.001 mbar, or 1 micron, as fast as possible.

Old-style, mercury-diffusion pumps (Fig. 2) are best suited to the task, but they’re no longer in use due to the risk of spillage if the pumps breaks, and the necessity of a dry-ice/liquid-nitrogen trap to prevent mercury vapor from backstreaming into the manifold. As such, specific oils are today’s pump medium instead of mercury, and they require different pump constructions.

Standard oil-diffusion pumps comprising umbrella nozzles (Fig. 3) are optimized to achieve a high, ultimate vacuum. However, the pump’s action starts between 0.01 and 0.001 mbar, and not at 0.5 mbar, which neon requires (Fig. 4). If one intends to use such a pump for neon (perhaps purchased cheaply from a surplus store), it must be very oversized to generate sufficient pumping speed in neon’s range.

In contrast to umbrella-style pumps, most suitable, neon-ready pumps are jet-style. Their principle is similar to a waterjet pumps; however, oil vapors propel the gas atoms. Multiple constructions have been used for decades; the simplest — and most cost-effective — variety for neon production is the single-stage, straight-updraft pump (Fig. 5).

The boiler temperature, which should be electronically controlled, is key to achieving good performance. Set the boiler temperature to prevent pump-oil condensation in the uppermost part of the cooling chambers. This part can’t be warm when touching the operating pump; there should be almost no visible boiling oil.

Some pump/oil combinations like to "burp" periodically when running (exhibiting a short, violent boiling action), but there’s nothing to worry about, provided no oil or oil foam spills into the intake or exhaust flanges. Silicone oils are less vulnerable to oxidization than mineral oils, but it’s almost impossible to repair a broken glass pump containing silicone oil.

As such, glass-diffusion pumps (especially multi-stage types) are masterpieces of scientific glasswork that should be handled with utmost care.

Advertisement

Subscribe

Advertisement

Most Popular