AMATEUR VACUUM TUBE MAKING
THE GETTER FLASHER
The life of a vacuum tube made under amateur conditions is determined by the outgassing of the insides of the tube after it has been sealed off. Without a vacuum furnace that can get right up to the annealing point of the glass, it is impossible to bake out all of the air. This residual air will leak into the vacuum space inside the tube over time, weeks-months-years, depending on the degree of processing you do to the tube before sealing it off. To make the whole process more practical, a shortcut can be used. This is the chemical getter. This is a chemical, usually a mixture of several reactive metals, that is sprayed onto an area of the inside of the tube where it will not interfere with the electrical operation of the tube. The metal reacts with the gasses normally found residually inside the tube, mainly water vapor, and some oxygen, nitrogen, and traces of other gasses. These gasses go into chemical combination with the getter flash and are thus removed from the vacuum. A relatively small amount of getter material will remove virtually all of the residual gas in a tube for many years. A properly processed tube with a getter will not lose its vacuum for over a hundred years. Many type 01-A tubes made in the early 20's are still completely functional today almost a hundred years later.
Since the getter is made of reactive metal, it cannot be left exposed to air for a long period of time. The new getters are coated with a barrier material that stops the effects of atmospheric air from ruining the getter before it is installed in the tube.
However, in amateur tube making, new getters are almost impossible to find. To get them from the manufacturer, you must meet minimum order requirements that are typically 1000 pieces, at about $2 each. A lot of money to put out in one lump sum!
The answer is to use old getters from old TV tubes. The getter is flashed in these tubes for a few seconds but not all of the getter material is used up. There is typically plenty enough left on the getter ring to service another tube.
There is a problem with doing this. Since the getter has already been flashed, the protective coating is long gone from the ring. Thus, the ring immediately begins to react with ambient air when you break open the TV tube to get it. You have about 2 hours of time to get the getter back into vacuum before it is so reacted that it will no longer work. Thus, you will typically get your tube completely ready to seal together before breaking open the TV tube. The getter is then mounted onto the tube in the desired place and the tube is then sealed together. It is immediately after put on the vacuum system to be processed.
FLASHING THE GETTER
Since the getter is located inside the glass envelope, the only way to flash it (flashing it means to heat it to red hot to evaporate the getter material off the getter ring) is to use a process called induction heating. A powerful high frequency magnetic field is applied to the tube envelope near the getter. This field induces a current by transformer action in the getter ring inside the tube. The power level is high enough to cause the getter to heat to red hot in a matter of seconds.
The power required to do this is a minimum of 300 watts of RF output. A more useful power is on the order of 500 watts, which gives more leeway for the positioning of the getter ring inside of the tube.
Here we have pictures of the present getter flasher. This is old technology from 1997 so it is not as elegant as units available today. A state-of-the-art flasher is now under construction and will be presented here in a few weeks. Thus, I will not go into the construction details of this unit, since it is too critical to the design layout. This unit outputs 300 to 400 watts, depending on which coil is used. It is marginal for some getter positions, where the getter is more than a half inch from the coil.
The size of the unit is 7 x 10 x 6 inches. The size of the handpiece can be seen by the pen. It is just over 5 inches long.
Here is what the insides look like. The circuit is driven by a MC4024 VCO. The oscillator is made variable so the frequency can be adjusted for different load coil sizes.
The driver is two sets of complimentary pair transistors that outputs a solid square wave at very low impedance to drive the gates of the HEXFET output stage. The output stage is a push-pull set of parallel connected 16 amp 400 volt HEXFET transistors. The power supply for these is 60 volts DC from a transformer-rectifier located in the bottom of the box. The current draw is right at 10 amps at 60 volts. The highest spike voltage on the HEXFET drains is about 250 volts, so there is plenty of safety margin.
The two gray cylinders are the matching transformer that converts the 120 volt peak-to-peak voltage from the HEXFETS to the 15 amp circuit to drive the hand piece. This circuit to the handpiece is through a specially made piece of coax cable with 20 ohm characteristic impedance so the current wouldn't generate high voltages that require expensive capacitors. Even at that, the voltage across the work coil reaches 1000 volts peak to peak. The handpiece has a set of mica capacitors that resonates the work coil to 700khz, the operating frequency of the unit. The resonant current through the work coil is over 100 amps. Due to the short duty cycle (typically a few seconds on) the coil doesn't heat up seriously. It does smoke sometimes!
This flasher is not practical to build. It was a first attempt and used designs that are now obsolete. My new getter flasher is a small hand held unit. Plans to build the new flasher begin on the next page.