How to change a flat tire speech


Wolfgang Scharschmidt
08.Dec.08  1



Expenses of a tube manufacturer for the production of long-term reliable tubes
                                                                                >Part 1:Tube manufacturer's point of view <
Part 2 deals with the topic from the user's point of view
I. construction
The guarantee to develop tubes of the best quality and to develop them in series production has always required a number of fundamental technical and financial efforts. Depending on the electrical requirements, this mainly concerned the construction, the choice of materials and their uses, the production of individual parts, the assembly precision and the testing costs during production and in the quality control area. Before the start of a new development and production, extensive measurements and examinations of previous tube types were required, based on which improvements were determined or analyzed typical sources of error corrected. Specialist engineers from the development and testing departments, production and quality control have analyzed the following main deficiency factors in the past, which had to be taken into account, such as:
a. Reduction in the number of welds.
b. Shortening of the system structure with the use of larger cathode diameters.
c. Narrow punched holes for the passage with a tight fit of the lattice bars with optimized double mica sleeves.
d. For better holding of the grids locking bands in the mica parts.
e. For the tube base, fix the lower mica parts in several places.
f. Fixing of the anode on the mica parts by means of strips welded over the anode tabs, instead of flanging or twisting the anode tabs.
G. Constructive measures that give the anodes greater rigidity and a more secure fit.
Based on these considerations, all systems of long-life tubes, for example, are short and robust, i.e. with box-shaped anodes, short supply lines and fixed by supporting mica, the tubes become rigid units in a "calibrated" flask. For example, the system of the E180F is only 6.5mm long and mounted on the base with very short supply lines.
II. material
The intensive effort required to increase the reliability of the long-life tubes began with the selection and provision of the raw materials. A first requirement was above all the purity of the raw material, since the cathode activity, the insulation, the absence of noise and the vacuum of the tubes largely depend on this.
         1.   Cathode material
For most long-life tubes, active nickel with little silicon and magnesium was used in order to keep the tendency for interlayer formation low. For example, for the tubes E90CC and E92CC with a special cathode without an interlayer, which were once developed for calculating machines, special passive nickel was used, which is practically free from magnesium and silicon admixtures.
       2. Glass                                         * see also pages 166-1 / -2 and 353 to 359
Glass was one of the most important tube materials. It has the advantage of being transparent, chemically invulnerable, electrical insulation, strength and the possibility of being able to melt it in the flame. On the other hand, it is more brittle than a dense material and requires some refinements in the technical process in order to ensure sufficient break resistance. First of all, it was necessary to comply with the tube piston wall thicknesses and dimensions, normatively based on the expansion coefficient Ö
of the glass used in each case. A lead glass was therefore usually used for the foot and natural lime glass for the flask, which pushed the foot slightly through the flask. A table provides information about favorable or unfavorable compounds of the various commercially available glasses, listed in "Proceedings of the I-R-E Volume 40, No. 10, Oct. 1952, page 1166".
* see contributionGsend: Description of a measurement method for determining the aValue
Ö Is a particularly important parameter for the foot glass because it is metallic
Implementations may only change within small limits. Other important glass properties
are their viscosity and insulation resistance. Also these two sizes will be
Checked and measured on receipt of material.
     3. control
For the delivered raw materials, there was a fixed requirement to continuously monitor them using quantitative physical and chemical measuring methods. First of all, it was necessary to adhere to the specified tolerances, to check all materials 100% for dimensional accuracy and to test them mechanically and technologically, such as the flexural and tensile strength of wires, sheet metal, entire system components, as well as some special mechanical tests related to glass .
Furthermore, the selected raw materials or semi-finished products or individual parts had to be cleaned thoroughly by washing, boiling and degreasing, similar to surgical instruments, and the oxide coatings on metals had to be removed by reduction under protective gas. The use of numerous auxiliary materials, cleaning agents for the stamped parts such as (tri-trichlorethylene or per-perchlorethylene), the water, as well as flushing and fuel gas were also subject to strict testing with ongoing analysis in order not to poison the cathode emissions early on with traces of impermissible admixtures. For the chemical production of the emission paste, chemical and spectrolytic examinations were particularly important in order to determine harmful admixtures: The following additional measurement methods had to be taken into account:
a. spectrographic examinations of the cathode material
b. checking the getter activity
c. a gas release measurement on the metals of the electrode systems
The cathode materials for tubes with a long service life required special purity requirements, which, in the course of tube development, became increasingly important and could be implemented thanks to refined measuring methods. Spectral analysis, for example, offered a method that could be carried out quickly for ongoing monitoring. It still made it possible to reliably detect admixtures of around 0.05% or, depending on the substance to be detected, to provide evidence of up to 0.001 ... 0.0001%.
In order to avoid a resistive intermediate layer in the cathodes, precise monitoring of the cathode nickel was of particular significance. All other tube materials, such as nickel or iron sheets, had to be tested for impurities in a comparable way. Even volatile additions of zinc or cadmium could be recognized and eliminated by this test. These impurities, which tended to deposit on the glasses, mica panes or electrodes, were neuralgic weak points for possible later insulation faults and noise disturbances. However, all of the aforementioned material testing and control measures were useless if no precautions were taken against subsequent contamination.
Another important component was the pin wire used for power supply [Finkh], because it is also responsible for maintaining the vacuum. Constant monitoring in crack detection devices and the creation of cross-sections were used to ensure that it had no channels through which air could possibly penetrate the tube during its long service life. So is z. B. A channel with a clear width of 0.01 mm can hardly be seen without high magnification. A tube with this wire quality would fail prematurely. Because of this problem, the wires were rolled before use and the fine cracks were smoothed out.
The behavior of the glass at high temperatures also had to be checked explicitly, as it was repeatedly heated, melted or fused with metals and then cooled again in the course of the production process. The measurements of thermal expansion required for this can be traced back to a softening point of 480-600 ° C, depending on the type of glass, from room temperature. In view of the special glass properties of becoming birefringent under tensile or compressive stress, polarization microscopy was used as a more precise measurement method. Under the polarization microscope, the degree of optical changes in the glass could be used to determine the level of stress directly.
     1. AllesGemein
The high demands placed on long-life tubes made it necessary to manufacture individual parts on precision machines specially designed for these tube types. So it was advisable not to manufacture the components continuously, but only to the extent of a daily production, a simple procedure to limit contamination.
As a further preventive measure, all components were washed thoroughly with perchlorethylene, which could only be used for a few washes without renewal. After washing, the metal parts were annealed in hydrogen or degassed by HF heating. All parts treated in this way then had to be transferred to assembly in a dust-proof, closed box. The parts that were still in stock were kept in protective containers for greater security.
     2. individual parts
The tube failures, which are based on specific cathode properties, were more difficult to influence and control than electrical and mechanical errors, defects such as:
1. Exhaustion of emissivity
2. Formation of intermediate layers
3. Decomposition and evaporation of the oxide paste
4. Poisoning due to the release of gas from individual electrodes or structural parts.
The first two points led to a reduction in gain and the latter to insulation defects and thermal lattice emission.
The listed cathode property deficiencies have less of an effect on the service life, the larger the effective cathode surface or the lower the specific load on the cathode. Compared with normal receiver tubes, the cathode load of which is between 25 and 90 mA / cm², these are kept lower for amplifier tubes for industrial and commercial systems with 10 to 40 mA / cm². With the E180F, the cathode temperature with normal heating is in the range of 680 to 690 ° C [black temperature]. It is around 50 ° C lower than with older types of tubes. The guaranteed service life of 10,000 hours requires compliance with certain heating tolerances, the deviation of which is specified with parallel heating at ± 5% from the nominal value of the heating voltage [with max. Mains voltage fluctuations of ± 10%] and with series heating with max. ± 1.5% of the heating current. A tube can, however, also tolerate larger heating voltage fluctuations, but larger changes in characteristic data can then occur. The overheating causes premature evaporation of the active cathode layer; the underheating leads to the depletion of the active surface substance of the cathode. However, a small amount of underheating is more advantageous, since a lower tube temperature improves the service life, but the cathode can only withstand reduced heating voltage fluctuations. The best service life is obtained with underheating between 3 ... 5% with the highest possible heating constancy of e.g. 1%. The optimal heating voltage, however, depends on the type of tube and the cathode load. The lower limit is always given by the fact that the saturation current must always remain large compared to the operating current.
All cathode sleeves had to be checked for dimensional accuracy before and after spraying with emission paste with gauges and the coverage of each cathode checked, some of which were then subjected to additional quality tests with microscopic surface tests. Particular care had to be taken to ensure that the cathode sleeve was correctly seated in the insulating mica platelets. A tab of the cathode sleeve, to which the cathode lead was usually welded, had to hold the sleeve immovably in a section of the lower mica plate. At its lower end, the connection of the cathode lead to the cathode socket pin was made with double welding for safety reasons. This ensured that, as a result of expansion or shrinkage due to switching on and off, no interruption in the cathode feed nor in the heating cables could occur.
The filaments are usually made of tungsten and covered with aluminum oxide. Each individual heater was checked for interruption of the insulation cover and the quality of the flattened connection points. When inserting the threads into the cathode sleeves, it was absolutely important that the insulating layer was not damaged.
The recurring burner errors are mainly due to three causes:
1. Excessive temperature inside the core
2. Damage to the insulating layer at the end of the cathode due to vibratory movements of the burner in the cathode
3. Beginning of breakage of the torch wire with sharp bends.
The working temperature could be lowered by using a thicker core wire. However, this requires a wire extension with additional turns, which also gave the torch a better fit in the cathode sleeve. Since the burners always have sharp kinks and are prone to breakage at these points, further improvements were required. The flexible double spiral torch was introduced, which made it possible to use a larger wire diameter. The burner temperature could thus be reduced to a safe value, no longer had any sharp ends and, despite later being firmly seated, could easily be pushed into the cathode sleeve without damage. Another improvement was achieved with a burner holder which, attached to the lower mica part, protrudes under the cathode. In this structure, the burner could be inserted and welded with a safe view before the assembly of the lower part in the glass base. In addition, the foot wires could be welded to the torch holders near the mica circumference and thus provided a good hold for the entire tube structure.
The production of the grilles required special care, precision and cleanliness. It is one of the mechanically finest systems in the tube and therefore sensitive. Even for the inspection of the grids, only the optical inspection came into question, since the usual mechanical tools the sensitive wiresA. could have damaged. The test was carried out with a
Projection device at 25x magnification. Otherwise, the manufacturing tolerances would hardly have been controllable, especially since the distances between the holes in the mica insulators were often only 6-8 µm and the tolerance of the lattice bars, such as with the E180F, had to be kept to ± 3 µm.
The high S / C ratio that was required for broadband amplifier tubes could be achieved, among other things, by small grid-cathode distances and thin grid wires. If you placed value on small tube dimensions and low heating power, then with regard to the mechanical stability of the systems with normal grid construction, you reached the limit of what is feasible.
Here are two typical tube examples:
                  1. Type 18042 G / k spacing= 120 µm; Mesh wire D.= 40.0 µm
2. Type E180F "= 53 µm; "D= 7.5 µm
The values ​​of the first example could still be controlled with a conventional grid construction. In the second example the grids could no longer support themselves and therefore had to be designed as so-called tension grids. For this purpose, the tungsten grid wire was wound with high pre-tension on a rigid frame consisting of two drawn molybdenum rods with four welded molybdenum strips. The high mechanical pre-tensioning of the grid wires stabilized the construction as a whole.
All possible precautionary measures were applied to the occurrence of grid currents, mainly caused by:
1. Lattice emission
2. poor isolation
3. bad vacuum.
The first two sources of error could be avoided constructively or through appropriate manufacturing precautions; bad vacuum due to various measures already mentioned.
        A. The wires as an important and sensitive component for the torch and grille are used
Incoming delivery checked for elongation and tensile strength by a materials testing center.
         Procedure GeGen lattice emission
An effective way of counteracting lattice emissions was to dissipate heat well from the grids [primarily g2and G1]. Either through copper bars or cooling fins welded to their ends [heat convection] in addition to receiving or radiation. In this way, the grid temperature could generally be kept in non-critical areas. Gold-plated grid wires were also used [control grid for steep tubes], since gold has a high work function for electrons; E.g. for tubes like CCa, E55L, E81L, E180F, E810F, E288CC and other special tubes. The production of the frame and its lattice gilding took place in separate operations; the gilding of the frames through oxidation and reduction electrolytically only after the surface has been roughened beforehand.In contrast, the grid wire was fire-gold plated by pulling the wire through a drop of gold heated in the RF field. To attach this wire to the bridge, a glass cover was used on the outside of the bridge. With regard to the contact potential, the gold plating was mostly only used in exceptional cases.
         activities GeGen grid currents
1. Careful mica selection
2. Enlargement of the creepage distances on the mica insulators by spraying with magnesium oxide.
3. Extension of the insulation distances between the connection pins of the tube socket.
The design of the tube base as a press plate was a significant advance in tube construction. In the case of critical tubes, the insulation distances could be artificially lengthened by enlarging the glass cone pin feedthroughs, which at the same time provided greater security against leaks or air pullers. For some tube types, for example, mica washers were used in the area of ​​the base system lower edge, which covered part of the press plate in order to prevent barium or magnesium from depositing between the grid connections and the other bushings. Both measures made a significant contribution to reducing grid currents and noise.
IV. construction
The assembly of the various individual parts took place in different separate work steps, a potential source of error, which made additional monitoring of the system assembly with constant checks for correctness and cleanliness essential.
The entire assembly had to take place in clean rooms with artificial, dust-free ventilation. For this purpose, vestibules, air conditioning and the like were used with constant monitoring of the air dust content. The rooms themselves had to be provided with a suitable floor covering as well as walls and ceilings in order to avoid the formation of lint as much as possible. And for the assemblers, certain clothing was determined according to the same criteria; smoking is strictly prohibited in these rooms.
The individual tube parts and systems were to be kept in dust-free containers. The finished systems had to be inserted into the flask as soon as possible after completion so that they could not be damaged by dust until they melted.
In order to protect the tubes from contamination during assembly, it was advisable to work in such a way that the assembly room was completely placed under a slight overpressure and the assembly areas were protected with an additional plexiglass hood, or work tables with attached overpressure cabins were used in which the assembly of the electrode assembly was used took place. The increasing number of tube failures caused by crackling or insulation defects caused by charred fluff were almost always caused by poor cleanliness or disregard of regulations by the assembly staff, i.e. their qualifications and care had a decisive influence on the product quality and the production yield.
Frequent assembly errors resulted from too many and poor welds, although welding was only carried out using assembly jigs. With a reduction in the number of welding points and with training measures for the staff, this problem could generally be resolved quickly.
The requirement to reduce the number of welds was often relatively easy to implement by making structural changes, but generally around 20 welds per tube still remained. Achieving reliable and reproducible welds required a high level of precision, but also concentrated work execution, especially since the setting of the welding resistance was left to the skill of the worker concerned, including subsequent control. Significant improvements and continuity of the welds were later brought about by newer, more sensitive generations of welding machines with limiter devices, which made reproducible welds safer.
Despite this optimization, there was still reason to monitor the system structure and its assembly between the individual production processes using special controls such as microsection examinations or optical tests with a magnifying glass or a binocular microscope. Work that was usually the responsibility of a test department attached to production.
Before forwarding to production processes in accordance with V., was every piece to check for the following points:
a) correct fit of the system,
b) dimensional accuracy,
c) durability of the welds without bending,
d) dust inclusions,
e) mechanical failure.
V. Melting and Pumpen
Both operations cannot be strictly separated from one another, as they continuously merge into one another. The problems that arise in this context are, above all, glass tensions that occur during melting and during vacuum pumping / evacuation.
If possible, the melting process should take place immediately after installation. This includes that the finished systems are immediately inserted into the glass bulbs in order to avoid subsequent contamination by dust or the like. The melting process itself and the subsequent cooling contained numerous sources of error. Most of these were glass defects, the most common causes and preventive measures of which are as follows:
a)thermal SpannunGen
As the name suggests, the stresses that occur are temperature-dependent. For example, they can arise when glass types that do not match each other are fused or when the melt is cooled unevenly.
Two main causes can be assigned to the stresses that occur:
1. different expansion coefficients of the glasses used,
2. Different expansion at melting points to a more distant point.
Here it was essential to reduce tension rings, even to avoid them, which arise when the glass is fused too hot and which have a harmful effect due to a temperature gradient that is too strong. The stress distribution in a stress ring is very complicated. It can be seen as a zone of high tension covered with a thin layer of glass. In the course of its service life, the tube [piston or foot] is stressed by further tension forces, which can lead to the tensile strength of the glass being exceeded. A check is carried out in the voltage tester on the basis of comparative observation with known standard values. [Optical changes due to polarization microscope or discoloration in the glass tension tester] A crack also offers a control option. It is done with a diamond, which leads to immediate glass breakage in the event of high tension.
A "crack" is the point at which a glass forms acute angles between its surfaces at its junction. One skilled in the art can identify the internal cracks by observing the external contours. Cracks in the fusion of the foot in the flask can be detected by immersing the cold tube in boiling water, because any failure to break starts from the inner surface.
               c) LeaksGopportunities
Leaks occur as a result of faulty glass-metal or metal-ceramic fusions. They depend on the chemical reactions during the melting process between an oxide layer adhering directly to the metal and the molten glass. The different reaction stages are accompanied by color changes. Visual observation of the sealant color is therefore an effective test method for vacuum tightness. In addition, the tube can be immersed in a viscous fluorescent liquid, which diffuses into the leak and becomes visible when additionally irradiated with ultraviolet light. Fine leaks, on the other hand, can only be detected by storage and characteristic destruction. Storage in hydrogen at a pressure of 7 kg / cm² has an accelerating effect.
After the fuse wires had already been examined in the manner described under "Material", additional fusion tests were also carried out, taking into account the expansion coefficient.
After melting down, the tubular flasks had to cool down evenly by going through a tempering furnace with three temperature ranges. To protect the systems against oxidation, the tubular pistons were constantly filled with fresh protective gas, just as they were before on the melting machine.
To test for harmful voltages, a number of dummy tubes were produced on the melting machines during the daily start-up of production and then cut open again for closer inspection. In addition, tube samples were taken every hour from normal production and examined in the same way.
The utmost care was taken when pumping out. A long-life tube was evacuated an average of four to six times longer than a normal radio tube. The metals were pre-degassed accordingly, the glass bakeout on the pump and the degassing of the systems were carried out with greater expenditure of time and increased care. Particularly high-quality materials were used for the getter or two highly active getter pills were even vaporized. In order to direct the precipitation of the getter on the upper part of the piston (dome), a cooling air flow was directed over the piston dome during gettering.
The melting of the tube with removal of the exhaust stem turned out to be the most frequent source of errors, which required a general quality inspection of the melted exhaust stem.
In addition, tubes were randomly immersed in boiling water for one minute. Specimens with critical tension were destroyed in the process. By means of other tested tubes, which were taken at random during the melting process at the automatic pump, it was possible to determine whether the melting and pumping would proceed correctly or whether corrections had to be made before the first fully formed tubes went to the test field. If the failure of the sample series was more than 5%, the entire production lot was rejected.
VI. CheckGen            
The measurements and examinations carried out on the tubes were classified into two groups:
CheckG the Gruppe A:
This included tests that had to be carried out on all tubes (comparable to a static test field).
CheckGen the groupppe B:
This included tests from random samples that were used for quality control. If the test series was rejected by more than 5%, this meant rejection of the entire production batch.
1.   CheckGen the groupppe A:
These tests began with the subjective observation of the manufactured structure and the control after melting and vacuum pumps. Then all tubes went through the activation process. About 250 volts were placed between each electrode - related to all other grounded electrodes - and thus lint was removed and short circuits were detected. In order to sort out possible burner errors, a high voltage was applied to the burner of each tube for a short time [for about 10 seconds]. Then the tube was measured for its electrical characteristics, followed by a short-circuit and shake test of at least one minute at 50 Hz and an amplitude of 0.5 mm. The tubes were operated under A-class amplifier conditions and the interference voltage at the output of the anode circuit was measured at the same time. A 10-hour test was then carried out for all tubes under A-amplifier operating conditions. This was followed by a test of the cathode activation with underheating with subsequent repetition of the electrical test. For a sample amount of 12 tubes, measurement curves were recorded before and after a ten-hour test operation, which were compared with one another after completion. By comparing the results, it was possible to see whether and what changes the tubes had undergone.
Of course, depending on the type of tube, its functional values ​​such as slope, gain, power, freedom from noise and other parameters were always measured 100%.
In order to obtain a stable reference point and at the same time to compensate for any deviation of the service life curve from the desired course at the beginning of the service life,allTubes aged for 50 hours under normal operating conditions.
The final test of the tubes began with the fact that they were in an A amplifier circuit in continuous operation on a vibrating table with an acceleration of 2.5 g* or briefly subjected to 5g. This test revealed all loose contacts or noise disturbances, mainly caused by dust inclusions. If too many tubes failed, there was an urgent need to intervene in production. In addition, the tubes were subjected to an impact test, which was also carried out with voltage applied. The tube followed the movement of a hammer, which was raised several times every second and then dropped again.*Acceleration due to gravity