开云体育

A couple of comments: The getter in receiving tubes (including TV sweep tubes) should be silvery, when it starts to be worn it becomes sort of white or cloudy. Heavily used tubes may show this around the edges of the getter on the inside of the envelope.
The resistors Drake used were carbon composition resistors, quite generally used at the time. The stability depends on exactly how they were made. The best of them were made by Allen-Bradley and sold under that name and also by Ohmite, exact same resistor. But, even those change with age. Typically a carbon composition resistor will drift up in value. The higher the original value the more it will drift. Probably any resistor of 100K or more will be found to have shifted a significant amount but even very low value resistors, like the 15 ohm cathode resistors in the T4XB will change, some actually drift low. Any composition resistor with visible seams or mold marks on the sides should be suspect. I don't know who made these but they seem to far less stable than the AB or Ohmite ones.
Modern resistors are carbon or metal film types which have much different construction. If made correctly these have excellent long term stability. They also do not have a voltage coefficient of resistance as do the composition type. While they are reputed to have higher reactance in fact they have LESS reactance than CC types, can be proved by measuring them. The reactance of the spiralled part is not significant and is very low Q. Carbon has a negative temperature coefficient, metal can have anything depending on the kind of resistance meterial used and the arrangement of the coatings (many resistors have more than one deposit of resistance material). I don't know about modern oxide composition resistors but they seem to be stable and can handle quite high power.
The instability of capacitors depends on the type of cap. Old type paper caps are not stable with time because of degradation of the paper dielectric. This varies with the paper used (it was special paper) and the effectiveness of the sealing of the element from external moisture. Heat in a capacitor can be generated by series resistance, that is, the dielectric has some resistance which can dissipate power. This is especially noticeable in electrolytic caps. A capacitor also has parallel resistance, usually known as leakage. Parallel resistance comes from anything that reduces the insulation of the cap. All electrolytic caps have some parallel resistance due to the nature of the dielectric. Modern ones use different chemistry than the old ones and have less leakage, but all have some. This passes some DC. Since one purpose of a capacitor is to block DC leakage makes it less effective for this function. Leakage, since it is resistance, also dissipates some power causing the cap to become warm. If there is a lot of leakage the cap can get quite hot and, since it originates with the dielectric the heat can cause the cap to expand and the dielectric, which is usually a paste, to boil out. These caps can explode if they get hot enough.
A paper cap with high leakage can also get warm but usually there is not enough to cause it to expand. However, caps used where blocking DC is important, such as grid coupling caps or cathode bypass caps, can cause malfunction of the circuit. In audio circuits leaky coupling caps can cause serious distortion since they change the operating point of the tube or transistor they are used with.
Old paper caps will generally need replacement due to changes in the value or the series resistance. Series resistance generally goes up.
Ideally, a capacitor of any sort should have infinite parallel resistance and zero series resistance. Some modern plastic film caps come close.
Mica and ceramic caps have a set of vices all their own. Mica was and is very widely used, especially for RF, because it has very good properties and is quite stable. The most common problem with mica is due to the type of construction. A great many caps are of the silver plated type. In these the mica dielectric is plated with a coaging of silver which forms the electrodes. These are inherently very stable and, since both the thickness of the mica and of the plating can be controlled very accurately they can be made with very close tolerances. The problem comes from the casing. If the casing allows moisture to enter it will cause the silver to oxidize forming crystals or whiskers. Eventually, this cause partial shorts changing the value of the capacitor and making it unstable in value. Same problem happens with plated crystals. This so called "silver mica disease" is common for capacitors made around the 1950s with Bakelite housings but does not seem to be a problem with later caps which are coated with epoxy by dipping. One should watch out for the symptoms of silver mice disease but not shotgun mica caps since most are perfectly stable.
The main problem with ceramic caps comes from a lack of understanding that they are not all the same. There are several types of ceramic dielectric; in general, the kinds with low dielectric contstant, called K in ceramic types, are very stable, rivaling mica but high K types have problems with ageing, temperature coefficient, voltage coefficient and sensitivity to mechanical deformation, i.e. they are microphonic. The caps used as temperature compensation are of the Low K veriety and are generally stable. The K value affects the size of the cap: low K can be quite large where High K can be very small for the a given value and voltage. The trade off is in stability, High K are NOT stable and should not be used where stability is a requirement. Other than that they are very long lived although their values may drift somewhat (I mean Hi K). In general, ceramic caps should not be shotgunned (with a few exceptions). Look out for mechanical damage around leads.
There is a great deal of tutorial information about capacitors on the web, mostly at manufacturer's sites. Worth doing some reading.
I seem to have written an essay. Didn't mean to. Sometimes my fingers just run away.

--
Richard Knoppow
Los Angeles
WB6KBL
SKCC 19998