KR Audio - July 2 & 9, 2013
- Written by Garrett Hongo Garrett Hongo
- Parent Category: Company Tours Company Tours
- Created: 07 December 2013 07 December 2013
In the KR Audio glass room, I saw two different steps of glasswork -- the shaping and cutting of the envelope, then the connection of the glass stem to the inner electrode and sealing the tube, making it ready for the vacuum room. I met Ladislav Krouzel, the glassman, a man of around 70 or so, I guessed, and about my height (five feet eight), wearing a dark blue polo, denim trousers, boots, and no work apron.
Krouzel held a four-foot-long tube of glass in one hand and a blowtorch in another, shooting flame at the open end of the glass, heating it until it glowed a light red.
Krouzel heated the glass until it softened enough so he could, through rubber tubing, blow into it to form a rounded, somewhat molten cap.
Then the glassman cut the tube about a foot below the cap and mounted the part he cut onto a lathe, where he put a torch to it again, further softening the glass so it could be reshaped as it spun within a metal die, transforming it into the characteristic tapered shape of a KR 300BXLS.
Next, Krouzel took the glass envelope out of the lathe and walked across the room, where he mounted it on a shaft of metalwork connected to a glass stem -- the inner electrode assembly of the triode tube -- that was sticking up from a vise.
The glassman clamped the envelope from above and then heated it up again, this time from the bottom, blowing pressure into it through a small surgical tube so glass extruded from the bottom. Then he pressed the glass inward with a metal paddle, fusing it to the stem and then sealing it.
Excess glass drooped of its own weight away from the nearly finished neck of the tube, and then he drew the entire tube off with a deft lift of his hand. This closing of the tube was a critical stage, a weak point of manufacturing, and his timing must be perfect, drawing the glass and connected inner assembly away from the vise and extruded excess glass while it was still slightly viscous. Chief engineer Gencev told me there must be no mechanical tension as he does this or the hardened glass of the tube neck could break, causing potential leaks and a tube that has become junk.
Adjacent to the glassworks is the vacuum room, where fully assembled tubes (minus metal bases) are connected via glass straws to an exhausting machine and pumped free of air to a relatively high level of vacuum (less than .01 pascal). The vacuum is necessary, Gencev told me, in order to direct the free electrons emitted by the cathode towards the anode of the tube. Regular atmosphere creates obstructions to the movement of a tube's emitted electrons, but a vacuum eliminates these obstacles (essentially, the electrons of atmospheric gases). KR first uses a rotary pump that creates a rough vacuum, then follows it with a diffuse pump that works by blowing the vapor from silicon oil heated to 200 degrees Celsius. Both housed in one large room of the KR factory, the vacuum room and glassworks together make for a very hot place.
In the same room, each tube goes to a cleaning station where it's attached to an inductive heater that operates via radio frequencies emanating through a metal coil. The coil slips down over the glass bottle of the tube and is then activated, heating the anode inside the tube until it glows red. This is how impurities and excess gases are burned off and the characteristic flashing -- that silver coating around the bottom of a tube -- appears. Gencev explained that the flashing is the residue of vaporized material from within the circle of the getter of each tube. If there is no flashing or flashing that is transparent, he said the tube vacuum is bad.
I didn't observe the steps for cathode activation or the assembly of a complete electrode, but I did tour the chemical, mechanical, and testing/storage rooms. In the chemical room, I saw a woman in a pink cotton tee with the word Hysteria in glitter script across her chest. She was introduced to me as Miroslava Pakostova.
Ms. Pakostova sat bent over a desk, picking at metal parts with tweezers, painstakingly putting together what looked to be cathode terminals -- the small metal fins -- onto a round mica spacer.
Gencev said that they use molybdenum, a silvery-gray metal oxide, to make their grids. Filaments, anodes, and cathodes are nickel, and cathodes are coated with barium.
I saw a tray full of completed electrodes already attached to their glass stems, ready for placement in their envelopes (glass bottles) and then sealing.
In the mechanical room, Gencev showed me the large lathes, presses, drills, crimping machines, and stampers they used to fabricate all the parts of each KR tube. It was essentially a metal shop.
Finally, in the testing/storage room I saw a long row of KR 242 tubes being tested and run in, flanked by a row of ordinary light bulbs. Gencev said the light bulbs, hooked in parallel, were the way they controlled overall electronic impedance to the tubes. Typically, a KR T1610 tube is burned in for two days. By contrast, a smaller 845 tube is run in for about 12 hours.
Also in the room were shelves of various kinds of triode tubes -- a veritable paradise for a tube junkie. Just glancing, I saw scores of 300b, 300BXLS, and balloon 300b tubes, as well as a few 211s, and a handful of 845s. But no Kronzilla tubes. I wondered why.
Contributor, The SoundStage! Network
Most-Read Articles on Global
- CES 2015: Competition is Fierce
- KEF's Killer Compact Monitor: The LS50
- Vitus Audio Masterpiece MP-L201 Preamplifier and Masterpiece MP-M201 Monoblock Amplifiers
- Super Speakers: Results
- Magico A3 Loudspeaker -- the Sound
- More Magico, Less Money: The S5 Loudspeaker
- Cyrus Audio: A Little Philosophy
- Magico A3 Loudspeaker -- the Concept
- Awful Avalon
- The Perfect Power Amp -- Ayre Acoustics' New VX-5
- The Limited-Edition Tribute: Near-Perfect Paradigm
- WideaLab Aurender S10 Music Server
- Wadia’s 321 Decoding Computer: An Affordable Exercise in Style and Substance
- Glorious Giyas: Vivid Audio's New G3
- Hegel: 10 to 11 for 0