Break in period

Drubin notes

the system was on, playing in the background. At one point during the evening, I began to notice extraordinary detail and elegance from the system. My brother noticed it as well. The amp sounded great!

Well, yes. The circuit will stabilise under normal operating conditions, as you yourself noted earlier

it's that individual circuit components (resistors, capacitors, wire perhaps) undergo some change during the initial hours of their use

However, you go on to note

This may not show in measurements but I think most of us agree that measurements alone do not the whole story tell

Not the whole story -- but some of it. Measuring the complete circuit during operation (DON'T really do this, for goodness sake :)-- I mention this only in discussion context) you may discover things... One easy component to track down is a phono equaliser. The components' tolerances can affect the de-emphasis; many of the ("good")resistors used are +/- 5% i.e. 10% total rated shift...

I am not sure what Krytrons have to do with the price of wheat in Alaska, or with audiophile-grade audio or even Radio Shack-grade audio electronics, for all it matters. Krytrons, made by EG&G or Perkin Elmer are high precision / high voltage switching tube-devices best known for their use in parallel honeycomb configurations when firing the chemical detonators that implode Plutonium 239 triggers of fission/fusion or fission/fusion/fission warheads. And trivially speaking, it is perfectly true that Krytrons are expected to fire precisely on spec the very first time they are activated. . . which is often their last time as well. . . no break-in planned, or allowed. So what? There seems to be a logical fallacy here somewhere. . . It is being contended that if for all electronic circuits, there exist at least one electronic circuit which can be built to fire completely on spec within 1 picosecond on first activation, then the same property must be inherently extensible to all electronic circuits. This is sheer nonsense. What we are discussing is whether for all electronic devices, there exists a subset of such electronic devices, digital, solid state, or vacuum tubed, which a significant portion of this audience extensionally deem to be high end audio gear, which tend to exhibit inherent and audible break-in patterns no matter how 'well' they have been designed/constructed/tested.

Guidocorona,

A kryton is an example of a high voltage application of a tube that is all; an application where a tube is a more reliable device than a transistor. I was trying to qualify the meaning of my previous comment of "low voltage applications" where transistors are often the cheap and preferred choice for designers.

Sorry if I offended you. Perhaps you missed the thread, which might explain the side track comments. I fully agree krytons are not related directly to audio applications. Although I suspect you will admit that reliabilty and response drift over time/heat/use is of concern in many electronics applications and not just audio. An undesirable response drift being analagous to extremly lengthy audible break-in time.

The components' tolerances can affect the de-emphasis; many of the ("good")resistors used are +/- 5% i.e. 10% total rated shift...

Just a clarification for others reading the thread (not a correction);

These are manufacturing tolerances between individual lumped elements such as a resistor. These figures do not represent the amount of thermal drift, drift with aging of the component. These values are not necessarily indicative of the accuracy of an amplifier.

However, if you were to simply change a resistor without finding a resistor with a very close match to the existing resistor then you could expect a sudden change in circuit response of this kind of magnitude.

Shadorne, if I get your comments right then I agree with them 100%. I feel very much that all components have their place and application.

Component stability (I would think) is as important in audio as in any other application. To that end one of the things that we really tried hard to do was to make our amps and preamps stable enough that even bias settings were only a very occasional adjustment. For tubes thats a big deal, and considering we make OTLs- well, the first thing we had to overcome was the idea that OTLs were unreliable.

At any rate, we got through all that but for whatever reason, we experience break-in phenomena all the time. What we have found about it is that it has *absolutely nothing!* to do with component or circuit drift. Nor is it some sort of illusion that is created in the mind- its very real. I think I mentioned earlier that some of the break-in effects are measurable too (and what some of them were).

Oddly enough, non of the effects seem to have anything to do with the fact that we use tubes. We have clearly seen what happens as capacitors 'form up' in the power supplies, what we've not been able to determine is exactly what is happening elsewhere although the evidence is pretty good that a lot has to do with wire. We have very little evidence that points to resistors, but we don't use non-precision parts either.

So bottom line is if I had to state what causes break-in, it would have to be filter capacitors and wiring.