This is a description of crystal production in 1968.
I began as a Trainee Engineer at the AWA Engineering Products factory in North Ryde, in 1968. AWA had several other factories, Domestic Products were made at Ashfield, AWV making valves, transistors and picture tubes at Rydalmere, and several other smaller divisions, dealing with Avionics, Marine, and Two Way radios. These factories were all in Sydney, but there were several in other Australian capital cities, and also in New Zealand.
I started my traineeship and spent 3 months in each section of the factory, for training and experience. There were many sections, drilling, turning, drawing, estimates, short-order-shop, instrument calibration, crystal manufacture, transformer winding, assembly, and final testing. There were other sections that I did not get to see, plating, painting, colour TV transmitters, design, and accounting.
The crystal room was very interesting. It was one of the few air conditioned areas of the factory, and also had a double door air lock, and sticky floor mat, all to help with cleanliness. We all wore white dust coats, and were not supposed to bring food, or personal belongings inside. It was divided into several areas. The main area was the final test and adjustment. There was the closed area used for crystal electrode plating. At the back was the crystal blank store, and the cutting area.
CUTTING AREA
People were normally not allowed in the cutting area. It was only when the boss was away, that a friend allowed me in, and showed me the process. In a box under the bench were lumps of raw quartz, the size of grapefruit. They were irregular shapes, and still had red dirt on them. In those days, we imported the raw quartz from mines in South America.
My friend then showed the next step, which was extremely important. He picked up a lump of quartz, and immersed it in a small tank, with Perspex sides, and filled with kerosene. There was a polarized light shining through the tank. He rotated the quartz, until he found the crystal axis, visible with the light. The quartz was then marked with a pen, and then checked. The first cut was done with a diamond saw. Then it was checked again. Now the crystal blanks could be cut to the required specifications, depending on the type of axial cut that was required. The crystal blanks were cut to size and thickness, depending on the contract or number of crystals ordered. A few extras were always made, to allow for breakages, or bad quartz, or just to put some in the crystal store. Of course, if there were an appropriate number of suitable blanks already in the store room, then these were used, rather than cutting some more.
COARSE GRIND
The blanks were inserted in a round bakelite holder, the size of large LP record.
It had lots of little square holes, which held the blanks. There was a metal bottom plate underneath it,
then the blank holder, then a top metal plate. The two plates were connected to an oscillator,
and an AWA Argosy receiver (type CR3D) was used to listen to the oscillator.
The plates were started to rotate with a motor, and grinding paste was added.
As the blanks were ground they increased in frequency. The receiver was used to chase the frequency.
At the appropriate frequency, the blanks were removed.
PLATING
The majority of crystals made then, were in the HC6U holder. The crystals were sent to the plating room. They were placed in a metal mask, like a sandwich, which was about 300 mm square. It held about 25 crystals. It had holders for each crystal. There were cut outs in each holder, with a round pad in the centre of the crystal, and a radial grove from here to the edge. The cutout on the other side was the same, but it went to a different edge.
The mask was mounted horizontally in a jig, and there were two 1 inch wide horizontal strips below the mask. These had a dimple in them to hold the plating metal. Solid gold wire, was cut into 5 mm long lengths, was counted out and placed in each dimple, so as the determine the correct amount of plating metal. A large glass dome was placed over the whole jig. A vacuum pump was started, and the air was exhausted from the glass jar. When it reached a sufficiently low pressure, the operator pressed a button. One of the wide strips was energized, it glowed red hot, and the gold was vaporized. The gold was deposited on the crystal through the mask. A lever was pulled and the mask rotated horizontally, presenting the other face to the heater strips. Again a button was pressed, and the other heater, vaporized the gold, and deposited it on the crystal. The pressure was reduced, and the crystals removed, and inspected.
MOUNTING
The HC6U base, had 2 little springs, with coiled ends. These were attached to the pins on the base.
The crystal was inserted edgewise into the two springs, and secured.
They were then put in a tray, and sent for final adjustment.
ADJUSTMENT
This section contained, four long benches, with about three girls per bench. Each operator had a view of a HP5245 digital frequency counter (with nixie tubes), which they could use, by pressing a button. Counters were very new and very expensive, so they only had one to share. Each operator had an oscillator, and they plugged the crystal into this, pressed the button, and checked whether the crystal was in the required range. If so, it passed. If not, there was a small plating bath, and more metal was plated onto the gold, to lower the frequency. The operator guessed how much time the crystal needed to be in the bath, and sometimes it was just seconds. Alternatively, there was another bath containing acid, and the crystal was etched to increase the frequency. Once again, the operator made a guess how long the crystal should be in the bath.
SEALING
The crystals were now sent to be sealed. The metal can was soldered on, and it was filled with inert gas and sealed.
FINAL TESTING
There were two operators for final testing, the boss, and the trainee, me! We each had an oscillator, but we had our own frequency counter, of the same type. First, the frequency was checked at room temperature, to make sure it was within specification. Then the frequency was checked over the temperature range, from the cold lower end to the hot upper end. We had a small crystal oven (about the size of an apple), which had a temperature sensor in it. There was an insulated container next to us, filled with dry ice. The oven was pushed into the cooler, and we watched the temperature fall. When it was at (or beyond) the lower temperature, the frequency was checked. The oven was turned on, and when the crystal reached the upper end of the temperature range, the frequency was checked again. This was commonly 0 to 50 degrees C for commercial crystals, or -30 to 100 degrees C for military crystals.
Two other checks were performed. It had to meet an activity level, in other words, the output level. It also had to have good activity at zero degrees C. If there was water vapor inside the crystal holder, it would freeze at zero, and the activity would dip, or even stop.
This testing took some time for each crystal, as the crystal had to be checked at every temperature in the range. The boss showed me an easy way to perform the test. The oven was placed in the dry ice. When it exceeded the lower temperature, the oven was switched on, and you pressed (and held!) the button, and you watched the frequency counter. You looked the frequency counter with one eye. You watched the temperature readout with the other eye. You also watched the activity at the same time (a third eye was useful). The frequency would change as the crystal heated up. It would move in one direction, then slow and stop, then start moving in the other direction. Provided it stayed within the frequency limits (which you had memorized) the whole time between the lower and upper temperatures, the crystal passed test. You also watched the activity level change, and it should stay above the minimum activity limit, the whole time. Particularly important was when it passed through zero degrees C, there should be no activity dip.
If the crystal was within frequency tolerance, and in excess of the minimum activity level, it was passed, and sent to dispatch.
This was an interesting period in the trainee-ship, much better than operating a turret lathe or a drill press.
Copyright
Ray Robinson