Burn in question and evaluation before burn in

I think the question then becomes, is it even possible to
build an amp that is unaffected by component burn-in? My experience, and not
just with the Almarro, leads me to believe that it's difficult, maybe impossible

Yes - It is possible to minimize the problem of burn in to the point that most
engineers would be quite comfortable that changes would be inaudible. It
requires careful design but it is not rocket science. Removing capacitors from
the direct signal path is the first step as these components are indeed non-
linear. Tubes are also well known to change response over time - although
negative feedback can be used to minimize this issue.

Amstrod, simultaneously comparing two identical components with different usage hours would be revealing. I actually believe it would show burn-in as real, especially for mechanical transducers (speakers, cartridges, mics, etc.).

Shadorne,

I'm not an engineer, but your suggestions about burn-in resistant designs certainly sound plausible. Are your Anthem and Bryston designed around those principles? If not, are you searching for amplification that is? What have you found that meets your criteria?

Onhwy61,

Now you've opened another whole field of inquiry. I too believe mechanical components break in (not burn!) and that the effects are very audible. I own Zu Druids, and was discussing break in with Sean Casey, and he said the whole reason they started running in the big drivers at the factory was because they had too many returns; people would set them up and hate the sound of the new stiff drivers and want to send them back, not being patient enough to wait the 200-400 hours they need to loosen up.

I was thinking about your earlier post:

Are people seriously arguing that they can accurately compare the changes in the sound of high quality systems when separated by hundreds of hours of actual listening time? It can't be done. Memory is not that reliable. If a component takes 400 hours to fully settle in, that translates to 2 or 3 calendar months (assuming 3 or 4 hours/day of listening). There are too many variables involved for any reliable comparative conclusions to be drawn over such a time span.

If all that is true, how do we ever make judgments about whether one component is better than another? What time span is short enough to be able to accurate compare differences in what we hear? I use Albert's method of taking some notes, and I have my wife listen and give me her opinion - she hears completely different things than I do. I also use a few recordings I know VERY well as a reference.

If we truly have such lousy auditory memories, it certainly explains why we sometimes seem to be chasing our tails on the upgrade path.

David

Armstrod, the approach you and Albertporter describe could actually be as good an approach as exist for determining overall system satisfaction, but it is not the only way and it certainly isn't foolproof. Short-term A/B listening is the best way to determine whether there's a difference between two audio signals. Longer term listening is probably much better at determining whether those differences are important to music reproduction. That's just my opinion based upon my listening experiences. As with anything, there will be people who are far better at doing something than I can, so I must admit to the possibility that those people can hear things and draw meaningful conclusion from data that I can't.

...our suggestions about burn-in resistant designs
certainly sound plausible. What have you found that meets your criteria?

Most modern SS designs are pretty robust. The "radio shack" type
designs, which are so frowned upon.

Frankly, IMHO - there at least three things going on;

1) User acclimatization to a new sound (true for all gear) - it takes time -
especially if you are well accustomed to your music collection. This change
will be slow as it will require you to go through your entire music collection a
few times to re-train your sonic memory.
2) User trying to hear things about their new toy - the extra effort and
expectation generally means you do hear something (or think you did) almost
every time you listen (at least for the first six months of a new expensive toy)
3) Real audible changes (mostly in the first 30 hours) on certain purist
designs that use very little feedback and valves with coupling caps and
transformers, and certain speaker designs.