Hearted Youtube comments on Technology Connections (@TechnologyConnections) channel.
-
1200
-
1200
-
1200
-
1100
-
1100
-
1100
-
1100
-
1100
-
1100
-
1100
-
1100
-
1000
-
1000
-
As a person who worked in a G.E. Lamp Plant for over 20 years making the actual bulbs for the FlashCube, SuperCube, HiPower, MagiCube, & FlipFlash (names such as GE, Osram, Wootan, & several store brands); your assessment of the parts (glass bulb, glass beads, zirconium/magnesium foil, tungsten filament, oxygen, primer, etc...) were spot-on. The most dangerous part of the process was the use of the primer for MagiCube, which the post was dipped into. (A safer & different primer was used to dip the tungsten filament & electrical posts for the other flash bulbs. That primer wasn't as pressure sensitive.) The cup that held the primer was changed every 3-4 hours. As long as the primer wasn't dried out on the inside of the dip cup; which happens as the level in the cup drops, it was safe. BUT dried primer was extremely volatile & accidents have happened. As a side note... back in the heyday of chemical flash; our plant was just one of several that ran over 100 machines; each machine producing 2000-2500 bulbs/hour, 24 hours a day, 6+ days a week, for about 20 years. THAT'S A LOT OF BULBS!!
1000
-
1000
-
1000
-
1000
-
1000
-
1000
-
994
-
986
-
963
-
934
-
906
-
896
-
880
-
871
-
860
-
858
-
854
-
852
-
845
-
841
-
840
-
Numitron backing is grey because RCA used what they had in spades, the sheet steel with aluminium coat that was used to make anodes and internal structures for thermionic tubes, and this was proven to survive the glass sealing and gettering operations. Thus they used the standard tools they had in the tube plant, the flat anode sheet, slightly formed to be a stiff backing, and punched out the holes needed to hold the filaments. Then used the technology they had to make glass beads with wire in them, and sealed those into the holes, making the filament supports, and then simply used a flat section of that steel wire that was bent over to hold a length of thoriated filament wire, also a common item in the tube shop. Length and diameter calculated for the brightness needed at the applied voltage, and then simply placed in location, the ends folded over, then spot welded together to trap the tungsten wire under slight tension. Then at the rear spot welds to a lead frame attached to a standard off the shelf 9 pin glass base, and you have the complete unit. Glass top attached, and then evacuated with the standard roughing pump, and as a bonus because of the low voltage, and no need to maintain an ultra low vacuum, the roughing pump and the heat sealing of the tube is all that is needed to operate, no need for a getter to be installed, and no need to flash it, just a RF heating during sealing to get a high vacuum, then seal.
Incidentally there were small versions made, the same size as your common 7 segment LED displays, and they were very popular, as they ran off 5V, and interfaced with logic. They worked best using CD4049/50 CMOS level shifting buffers, as those would source or sink 50mA no problem. Using a buffer/inverter per lamp, and a BCD decoder or counter per digit allowed those displays to be bright, and as bonus you could also use the blanking input on the drivers to use PWM to dim them.
Project to replace those displays with LED ones worked, just that it really did not drop display current use, it was still 5A of current at 5V, though it was good in that at least you had a display that now was available, using a tiny HP 7 segment red display. Do one conversion and you had 16 numitron displays to use to fix others, so we only converted 3 boards to the LED version. Biggest problem was the resistor value selected was too low, so the LED displays were running way too bright, so had to be dimmed. Rather than destroying the cordwood board made to fit them, I simply used 2 6A silicon diodes in the common line, to drop the voltage seen by the LED displays down from 5V to 3V8, which made them dim to exactly match the old displays. Those 2 diodes were hard to fit in the limited space left on the display board. Users liked the new crisp displays, the bright version got complaints that it was so bright it was unreadable at night with dark adapted sight, and it lit up the entire cockpit. Display dimming had to match the other display, and that board used unijunction transistors, and had a disconcerting habit of the power transistor unsoldering itself from the wire leads, it ran so hot. Base lead unsolders itself, transistor is still conducting, runs hotter and lamp blows. Select spare lamp and it also blows, unless enough time for transistor to cool below 200C junction temperature. Would have been nice to have had some of the more modern mosfets that can handle 50A, but not at the age of that design.
830
-
827
-
808
-
806
-
806
-
799
-
774
-
773
-
772
-
770
-
766
-
765
-
745
-
743
-
736
-
710