Burn-in truths

Graham Slee wrote: Quote from Analog Devices "Op Amp Applications" Handbook; Chapter 7: Hardware and Housekeeping Techniques; Section 7-1: Passive Components by James Bryant, Walt Jung and Walt Kester... "All resistors tend to change slightly in value with age. Manufacturers specify long-term stability in terms of change - ppm/year. Values of 50 or 75 ppm/year are not uncommon among metal film resistors. For critical applications, metal-film devices should be burned-in for at least one week at rated power. During burn-in, resistance values can shift by up to 100 or 200 ppm. Metal film resistors may need 4-5000 operational hours for full stabilization, especially if deprived of a burn-in period." Metal film resistors are the best "all-rounders" among resistors only excelled by "bulk metal or metal foil" types which have the disadvantages of being low power and very expensive. Virtually all our resistors are metal film. I investigated burning-in on a production scale and had to abandon the thought: the cost was absolutely prohibitive - it would have pushed prices up far more than double. On a one by one DIY basis I'm sure it could be worthwhile. Most resistors are run well within and often at a small fraction of their power ratings simply because that is how a design pans out. Therefore the user should be prepared for up to 6 months total burn-in. A sobering thought for us all, never mind those who have difficulties in separating hi-fi myth from hi-fi fact.

I would hazard a guess that the earlier part of the 6 months is probably the steeper part of the burn-in curve. And true Mike, we don't have time to run the full 6 months on every permutation, but nowadays we're not rushing to market so quickly (in fact I personally need crutches but that's another story...), and I reckon 90% (ish) of the components will be fully burnt-in by the last tweak before signing off.

From a DC point of view, op-amp input current vs resistance matching goes a long way to reduce the effect of ouput drift with burn-in value changes. I've noticed it's beneficial to the sound too.

As you both commented, the sound is the final arbiter.

True Mr Sun - I've been harping-on about capacitor burn-in for quite some time. From the humble electrolytic to the sublime polypropylene, they all seem to take a few days, even weeks. But resistors? You don't see much about resistor burn-in, and I thought that was interesting. Silicon too, as you note, takes time to "develop". And so do interconnects.

I guess they've all got one thing in common - insulation. And when we find there is no such thing as an insulator - just very poor conductors, I think that is the giveaway that answers why burn-in takes so long. You can "whirl" an interconnect to loosen the grip of a poor conductor (the plastic insulations) from a "super-conductor" (the copper wires), but how do you loosen the grip of the coating on top of a metal film resistor element? Burning-in at full rated power for one week will definitely cause some thermal expansion - perhaps the tight grip of insulation is the real reason behind burn-in? Why things don't sound right straight away.

After all, a capacitor is layers of insulation and conductors in intimate contact. The better sounding dielectrics (dielectric means insulator) are those less prone to adhesion - and polypropylene will not readily accept anything stuck to its surface. And we do find that Teflon has great sound properties - Teflon is "non-stick". So, it could be safe to assume that electron flow at the point of contact between conductor and "insulator" is the mechanism that breaks down, or loosens the grip? And if so, then the lowest current part of a circuit is going to take the longest.

All this could be complete speculation on my part, but there is some very good supporting evidence: for those able to grasp how overhead electricity cables in the power distribution infrastructure work - the cables are uninsulated not for reasons of economy because they're "up there", but because insulated alternating current cables get hot due to "drag" or resistance (a "Biker's" analogy being wind resistance...) of that insulation - and being uninsulated is why they put them up there. Now that argument may not sound tangible - more like I made it up on the spot? No! Every copy of the IEEE xth edition on-site guide used by electrical installers, and part of the qualification process for electricians, bears this out - just look up the different current ratings for the same type and rating of wire: the more insulation - the less efficient it gets. Yes, it is about heat disipation, but how is that heat caused in the first place? And you'd have to be a "Boy Scout" or "Girl Guide" (being politically correct...) to really understand that!