Is there a "burn out" after the burn in period?

I’ve had my entire system not only turned off but unplugged from the AC for the last two days due to daily passing severe thunderstorms. Based on current predictions I don’t think I’ll have my system plugged back in and (source components at least) turned back on until tomorrow afternoon. Wonder how long it will take for everything to get back to where it was.

Ted Smith said But considering anecdotal evidence is often the first step in understanding something new.
Exactly.

But I keep my brain plugged in at the same time. There is a lot of nonsense out there as well. Sort of the observational equivalent of an XLR instead of just SE.

tony22 said I've had my entire system not only turned off but unplugged from the AC for the last two days due to daily passing severe thunderstorms. . . . Wonder how long it will take for everything to get back to where it was.
Let us know. I find the DS just needs enough time to feel gently warm to the touch. A couple of minutes it sounds great, just a bit longer it is at its best.
Elk said
tony22 said I've had my entire system not only turned off but unplugged from the AC for the last two days due to daily passing severe thunderstorms. . . . Wonder how long it will take for everything to get back to where it was.

Let us know. I find the DS just needs enough time to feel gently warm to the touch. A couple of minutes it sounds great, just a bit longer it is at its best.


Glad I did what I did. Had transients over the last day or so and when I got home today there was no power in the neighborhood. Doesn’t look like things will get turned back on until tomorrow.

So I had everything off for almost 3 days. My impression is that the system got to sounding pretty much like it should after about a day. Within 2 days I’m pretty sure it was back to normal.

That is a long warm-up period.

Well, it could be that I was just imagining it got a little bit better after the first day. But it was not good at all immediately after I got everything back up and running, and for some hours after that. I’d say that waiting a day is probably a good bet for a situation like this.

yes, my experience is that 1/2 hour from standby and overnight from a cold start have an obvious effect on overall SQ. I would express it more in similarity to some of the burn in variables rather than dramatic, but obvious none the less. Mind you, that example seems to be somewhat controversial as well.

Vive La Difference!

But this is for the entire system, right? So it doesn’t really say much about any single component. Did you happen to notice whether your power quality was bad (e.g., incoming THD higher) at first? After a significant outage it might take the utility a while to get everything back to normal, even after the lights come back on. Not saying that’s the reason here but just curious.

An interesting observation.

In my case [p10] plus I have noticed it when comparing DACs and we have learned NOT to power off a temporarily out of use machine since the rewarm wait affected our listening and comparison window.

I really have no idea how we could even test or accurately any of these differences other than objectively making subject suppositions. he he he

I do hear something though so it is easy to eliminate by standby and firing up the entire system an hour or so before a musical afternoon or evening.

stevem2 said

But this is for the entire system, right? So it doesn’t really say much about any single component. Did you happen to notice whether your power quality was bad (e.g., incoming THD higher) at first? After a significant outage it might take the utility a while to get everything back to normal, even after the lights come back on. Not saying that’s the reason here but just curious.

Yes Steve. I had everything powered down and unplugged. I can't answer as to power quality, but I didn't plug everything back in for a full day after the last power outage (which lasted about an hour). Frankly I think for a full system "recharge" a day is probably realistic. IME the thing that takes the longest (depending on how they're constructed) are the ICs. Seems some of them take forever to do whatever it is they do - dielectric charge settling? God knows what. When you have a whole system coming back to life I suspect some of the timeline is driven by component interaction.

I agree, Gordon; the change is not as dramatic as burn in of a new component.

As an additional point of perspective, consider this.

In racing, the engine gets hot due to the work being done. It ‘soaks’ into all of the metal but at differing rates due to mass and type of metal etc. of each part. This is called heat saturation, or heat soak.

One of the ‘theories’ is that until equilibrium is reached (the ∆ temp between the internal and external temperature has been minimized for the major mass items like the crank etc.) there is an increased ‘risk’ of deformation due to ‘stress’. This can lead to failure, or excess wear, or ‘less than optimal performance’.

I wonder how much this can affect our sensitive electronics.

Big cans and all the other non-major heat generating components (transformers, pots, resistors, solder connections etc.) all need to reach thermal equilibrium all the way to their core. To let the heat ‘soak’ into all of the components, to reach equilibrium.

And unlike a racing engine our audio equipment operates with a much narrower range of peak to ‘average’ temperatures, as well as lower temperature overall. And since the total heat output is much lower, that might be why it takes more time to ‘stabilize’ once turned on.

Just a thought or 2.

JJ

A reasonable thought, clever thinking.

When I hear heat soak, I immediately think of a heat soaked turbo/supercharger: the ECU maps for the temperature and power output is substantially decreased. I hope there is not an audio equivalent. :)

Heat soak testing is commonly used in Military standards (mil-spec.).

For safety critical electronics it is normal to cycle equipment for 200 hours (IEC61508).

I do not foresee such requirements introduced in the audio business.

If I am not much mistaken JR produce equipment in compliance with the ISO 9000 standard which is examplary.

Elk said A reasonable thought, clever thinking.

When I hear heat soak, I immediately think of a heat soaked turbo/supercharger: the ECU maps for the temperature and power output is substantially decreased. I hope there is not an audio equivalent. :)


Yeah heat soak is usually associated with peak temperatures during ‘racing’ conditions where ‘elevated’ temps are being dealt with.

There is, in our case, where the thermal contribution is more or less constant, or at least is no where near as dynamic as in an auto engine, and it’s this condition that seems to be of key importance to us audiophools.

In many cases reaching this thermal equilibrium can actually be heard (thus this thread) but as has been noted it can take days to reach full heat soak. As such it makes sense, at least to me anyway, why it can take so long to stabilize. Especially with a low thermal generating component such as pre-amps, DAC’s etc., which have a relatively large mass and/or plenty of air cooling holes and or fans etc. So it makes sense, that to fully stabilize why it takes much longer for some components than say an amp or other units that do create ‘plenty’ of heat.

And I figure that the medium and large electrolytic caps are the most susceptible to this because as the heat enters, the electrical characteristics will shift due to the spacing of the plates slightly shifting as well as conductivity etc., until there is a degree of thermal homogeneity.

JJ

Oh, it’s much worse that that. Taking just caps as an example: their values change significantly depending on temperature. The whole board and the air around it needs to reach equilibrium before they settle on a value. It’s worse when playing loud music heats things up and then then the system cools down during quiet passages…

(Edit: actually as I look at this chart, it’s not as easy to read as I’d like, first off NP0 is measured differently than everything else, but even so you can see that there are some caps that can be worse than electrolytics and you can see the kinds of variances out there (and why we stay away from Y5V !) )

https://en.wikipedia.org/wiki/Types_of_capacitor#Temperature_dependence :

Temperature coefficients of some common capacitors
Type of capacitor, dielectric material Temperature coefficient ΔC/C Application temperature range
Ceramic capacitor class 1 paraelectric NP0 ± 30 ppm/K (±0.5 %) −55 to +125 °C
Ceramic capacitor class 2 ferroelectric X7R ±15 % −55 to +125 °C
Ceramic capacitor class 2, ferroelectric Y5V +22 % / −82 % −30 to +85 °C
Film capacitor Polypropylene ( PP) ±2.5 % −55 to +85/105 °C
Film capacitor Polyethylen terephthalate, Polyester (PET) +5 % −55 to +125/150 °C
Film capacitor Polyphenylene sulfide (PPS) ±1.5 % −55 to +150 °C
Film capacitor Polyethylene naphthalate (PEN) ±5 % −40 to +125/150 °C
Film capacitor Polytetrafluoroethylene (PTFE) ? −40 to +130 °C
Metallized paper capacitor (impregnated) ±10 % −25 to +85 °C
Aluminum electrolytic capacitor Al2O3 ±20 % −40 to +85/105/125 °C
Tantalum electrolytic capacitor Ta2O5 ±20 % −40 to +125 °C

Yow Babeiz…

With a range of +22 to -82% which is over a 100% change, I can see why that class of cap would be ‘unacceptable’. 2guns_gif

And these sorts of thermal changes can be especially ‘critical’ in the final stage of power supply filtering in very low voltage sensitive components like dacs and MC cartridge amplifiers etc.

And when tubes are in the circuit, what with their usually higher operating voltages, small changes in capacitance have a larger influence on the amount of voltage shift, or lack of filtering etc. Not to mention their added thermal contribution.

And then there is the factor of the 3 different types of heat, how they each propagate, and dissipate differently etc. this whole area of influence can become complex rather rapidly…nerd_gif

JJ

Yow Babeiz is right!

This is completely nuts.

Thanks, Ted. Fascinating stuff you keep offering up.

Yes very interesting. It is amazing how much these variations can be huge!!

For me these variabilities explain why the sound is changing when you unplug the system, cool it down and switch it on. But is there some technical sheets giving a variability of the capacitors over the time which could explain the burn in phenomenon?

Burn in isn’t just one thing or even the same thing from one component to another. It has also been discussed to death. FWIW there are often specs about how much a component would have to change to cause it to be faulty in the manufacturer’s eyes. You can guess that these give hints as to what things do change over time.

Here’s a section from a Panasonic Surface Mount Type Aluminum Electrolytic Capacitors spec:Untitled.png

Depending on the circuit use different parameters might be the most important ones: value, leakage, dielectric absorption (tan δ), etc. A good design tries to minimize such effects, but nothing is perfect. Note that some of these parameters change by more than 100% over the lifetime.