|
|
Post by Casey Leedom on Apr 11, 2017 21:59:10 GMT -5
I know this is going to raise the ire of many people. So I'll skip the entire issue of people saying they hear differences in Solid State Equipment after that equipment has "broken in". And for the purposes of this discussion, I'm not talking about differences between cold and warmed up electronics. I'm talking about the idea that Solid State Electronics can change over longish periods of time to "break in".
My question is this: what proposed mechanisms do people hypothesize? And it's not good enough to say things like "the Capacitors are breaking in", etc. That just pushes the problem down one level. What are the specific physical mechanisms at work?
For instance, in the domain of "Speaker Break In" people will often say things like "the speaker cone suspension surround becomes less stiff (lower elastic constant) with use".
So, anyone want to throw down on this? :-)
Casey
|
|
|
|
Post by audiosyndrome on Apr 11, 2017 22:03:14 GMT -5
Dielectric absorption specifically in the capacitors.
Russ
|
|
|
|
Post by Casey Leedom on Apr 11, 2017 22:07:53 GMT -5
So that's for an Electrolytic Capacitor, right? This leads to the capacitance going up/down? Speed of charging/discharging going up/down?
Casey
|
|
|
|
Post by pknaz on Apr 11, 2017 22:09:46 GMT -5
Dielectric absorption specifically in the capacitors. Russ Wouldn't the charge in the dielectric be dissipated once the capacitor discharges? If you mean to say that the dielectric changes is capacity to adequately perform within specification, wouldn't that mean that the capacitor as a whole, over time, would change specs? That's kind of a bad thing, isn't it? |
|
|
|
Post by Casey Leedom on Apr 11, 2017 22:22:23 GMT -5
Oh, I thought that he was referring to the Dialectric Electrolyte absorbing into the Capacitor Plates. Literally "soaking" into the plates as if water into porous paper.
Again, I'm really not trying to start fights or anything. I'm just "pushing on a string" that doesn't make sense to me. But I've been confused about things in the past (usually multiple times per day) so I wanted to hear what the proposed mechanisms were for Solid State Electronics "break in".
Casey
|
|
|
|
Post by boomzilla on Apr 12, 2017 0:57:25 GMT -5
Physical components, such as capacitors and resistors, should not, over time, drift sufficiently in value to exceed their design tolerances. If they do, then either they're defective or at the wear-out phase of their service life.
That said, it is a common and (seemingly) widely-held consensus among those who work with audio every day, usually, high-end audio salesmen, that leaving components on 24/7 makes them sound better. Now this consensus has never (to my knowledge) been subjected to rigorous scientific testing. And there are those (here on the Lounge, and elsewhere) who say that this is an "audio myth" that has no actual validity.
I, personally, have no specific opinion in this debate. I normally leave my equipment off, but turn it on (when possible) 30 minutes in advance of critical listening. This not only protects the equipment from lightning strikes but also provides thermal stability prior to listening.
Boom
|
|
|
|
Post by Gary Cook on Apr 12, 2017 1:28:24 GMT -5
My view is simply based on personal experience where we used to hot swap amplifiers during performances. Take the failed one out of service and replace it with an brand spanking new one straight out of the box. I couldn't tell any difference and neither could the musicians and they are trained to pick even the smallest difference in the sound of their instrument. I've done similar with my pair of XPA-1L's, I had one a couple of weeks before the second one, so the first one had about 100 hours on it before the second arrived, brand new out of the box. I couldn't tell any difference even with mono input, sounded the same to me.
Electronically I just don't see exactly which components are going to change after what is a tiny part of their life cycle. For example if they burn in after, say, a hundred hours then that must logically represent a large % of their serviceable life. As for "warming up", other than when the ambient temperature is below the components' rated temperature (commonly 0 to 70 degrees celsius), I just don't see there being an issue. If it's within their operating temperature then they are certified by the component manufacturer to be in spec. I don't know about you guys but it doesn't get to 0 degrees celsius in my lounge room in Sydney.
Cheers
Gary
|
|
|
|
Post by vneal on Apr 12, 2017 5:09:37 GMT -5
I was told you have to let the snake oil warm up so the system can sound fluid
|
|
|
|
Post by audiosyndrome on Apr 12, 2017 9:39:13 GMT -5
So that's for an Electrolytic Capacitor, right? This leads to the capacitance going up/down? Speed of charging/discharging going up/down? Casey capacitors are made from conductors and insulators (dielectric). The dielectric is not a perfect insulator and "charges" up over time (the absorption). Once fully saturated all is good. Until then the dielectric continuously charges and discharges adding a small element of noise to the signal. This is what the various dielectric bias systems (DBS) attempt to eliminate. Russ |
|
|
|
Post by Casey Leedom on Apr 12, 2017 11:32:03 GMT -5
Ah, thanks for explaining your hypothesis Russ. I'm clear on what you mean now.
So you're postulating that the Dielectric Constant of the Dielectric (insulator) isn't in fact constant and changes over a portion of the initial in-use life of the Capacitor? Or would this postulated change affect the speed with with the Capacitor charges/discharges?
Hhmmm, rereading your note, it sounds like you're talking about "noise" in the Capacitor's "Discharge Signal"? I.e. instead of a simple exponential decline in charge/voltage across the Capacitor, it's a more "interesting" curve? That should show up in testing with an oscilloscope I would think ...
Casey
|
|
|
|
Post by leonski on Apr 12, 2017 22:31:51 GMT -5
Yep, Russ makes the most common arguement about capacitor 'break-in'. And I tend to agree. Though the DBS system is a little overkill, especially if you use the component on a regular basis. The fact that the batteries in such systems essentially last shelf-life, tells me Zero Current Flow. Electrostatic field propagation? I don't see that happening in what is essentially a good insulator. Diealetric Constant when measured very quickly and with high resolution equipment (a Gigahertz scope?) will indeed vary over very short time periods. And while the insulator part of a capacitor is not 'perfect', neither is the conductor. It has resistive and maybe capacitive or inductive characteristics, too. www.humblehomemadehifi.com/download/INFO_ClarityCap_Technical_Report.pdfGood article from Clarity Caps. NOTHING is a perfect ANYTHING. Semiconductors DO wear out. And the capacitors or capacitive measure (no PURE anything, resistor, capacitor or whatever) in such devices, especially LSI (Large Scale Integrated) which have anywhere up to in the Millions of Transistors ALL with some leakage, and other micro-measureables. Wearing Out of semiconductors is nutty, you say? (Go Ahead!) Well, the metalization layers exhibit a Phenom called 'Electromigration'. At places of thinnest metal, current density is HIGHEST and the metal tends to flow with the current. An Atom or so at a time. Over long times, the metal will get thin enough to FUSE (act like a) and the device fails. High power devices, like output transistors may actually be less prone to this problem, given the thick metalization layers involved and distributed gate structure. Stuff like grain boundaries and crystal structure of the metal layer play a huge role in this. semiengineering.com/how-to-deal-with-electromigration/Copper is very common these days while when I started, the vast majority of devices were fabricated with Aluminum/ 1% Silicon. Metal is applied to transistor usually thru a process of 'sputtering' which takes place in a vacuum chamber using a 'plasma' (excited argon atoms, in this case) to dislodge aluminum from a 'target' which than travels a few MM and 'sticks' to the substrate. Thickness can be controlled to very high standards. The metals internal structure is largely dependent on such measurable as depositon rate, temperature, partial pressure of 'other' stuff besides Argon, and MORE. Analysis is performed using measures of reflectance and resistivity of the depositied film all the way to SEM photography and other more sophisticated analysis. |
|
|
|
Post by pknaz on Apr 12, 2017 23:47:27 GMT -5
IMHO, any solid state electronics that seem to sound different (I won't say better) is completely attributed to psycho-acoustics. Just my $.02  |
|
|
|
Post by leonski on Apr 13, 2017 1:08:37 GMT -5
IMHO, any solid state electronics that seem to sound different (I won't say better) is completely attributed to psycho-acoustics. Just my $.02 amp differences can largely be accounted for by the fact that even 2 amps which measure the same are doing so into a resistor won't perform the same driving a real world speaker. And, NO speaker I know of is purely resistive. Many factors come into play, not the least of which is reactance of the load. So? A speaker will, at various frequencies look either like an inductor or capacitor to the partnering amp. Amps vary substantially in ability to drive such loads. Even those measuring the SAME into a resistor. It gets a little more complicated when you start dealing with phase angles (of the reactance). That is NOT to say that you don't get 'accustomed' to whatever sound you are hearing. That may be at least some of break-in. I know I occassionally hear a system at a show which sounds AWFUL and I leave as soon as polite, if not somewhat sooner! |
|
|
|
Post by pknaz on Apr 13, 2017 3:22:35 GMT -5
IMHO, any solid state electronics that seem to sound different (I won't say better) is completely attributed to psycho-acoustics. Just my $.02 amp differences can largely be accounted for by the fact that even 2 amps which measure the same are doing so into a resistor won't perform the same driving a real world speaker. And, NO speaker I know of is purely resistive. Many factors come into play, not the least of which is reactance of the load. So? A speaker will, at various frequencies look either like an inductor or capacitor to the partnering amp. Amps vary substantially in ability to drive such loads. Even those measuring the SAME into a resistor. It gets a little more complicated when you start dealing with phase angles (of the reactance). That is NOT to say that you don't get 'accustomed' to whatever sound you are hearing. That may be at least some of break-in. I know I occassionally hear a system at a show which sounds AWFUL and I leave as soon as polite, if not somewhat sooner! Sorry, my mind was thinking faster than my fingers were typing, what I meant to say was..... ...any solid state electronics that seem to sound different (I won't say better) after being turned on for a period of time, and allowed to warm up, is completely attributed to... |
|
bootman
Emo VIPs 
Typing useless posts on internet forums....
Posts: 9,358
|
Post by bootman on Apr 13, 2017 4:33:00 GMT -5
IMHO, any solid state electronics that seem to sound different (I won't say better) is completely attributed to psycho-acoustics. Just my $.02 amp differences can largely be accounted for by the fact that even 2 amps which measure the same are doing so into a resistor won't perform the same driving a real world speaker. And, NO speaker I know of is purely resistive. Many factors come into play, not the least of which is reactance of the load. So? A speaker will, at various frequencies look either like an inductor or capacitor to the partnering amp. Amps vary substantially in ability to drive such loads. This is why we need to stop testing across a resistor. ...and Maggie's are the closest you will find to being one.  |
|
|
|
Post by leonski on Apr 13, 2017 12:53:12 GMT -5
Maggies sound good with a variety of amps, both SS and Tube, as long as they can simply supply the Juice needed for YOU.
The amp in the lobby @whitebear Lake is a smallish tube amp and sounds fine at 'lobby' levels.
Alternatives exist to straight resistor testing from a variety of 'dummy loads' (simulated speakers) to the Power Cube system.
I don't know that ANY standard has been agreed to by a large number of amp or speaker makers.
|
|
|
|
Post by boomzilla on Apr 13, 2017 14:53:33 GMT -5
Want a hellish test for amps? The original Apogee ribbons had sub-one-ohm dips AND ugly phase angles. If your amp can drive one of THOSE, it could drive anything.  |
|
KeithL
Administrator 
Posts: 9,928
|
Post by KeithL on Apr 13, 2017 16:39:44 GMT -5
What you say is true - but only to a limited degree - and ONLY for capacitors with a DC bias on them to begin with. So it could conceivably be true for coupling capacitors on the inputs and outputs of amplifiers - especially tube amps because of the high voltages involved. Some electrolytic capacitors do in fact exhibit a sort of energy storage, which may cause them to act non-linearly, and that may vary over time. (Take an electrolytic capacitor, put a voltage on it, let it sit for several minutes, remove the voltage, and short the capacitor to drain it. Then remove the short. If you measure it several minutes later, you may find that a slight voltage has mysteriously appeared... the capacitor has acted a tiny bit like a battery. This happens more with certain types than others, and usually isn't significant enough to be audible, but it's at least real.) HOWEVER, it's NOT AT ALL TRUE for speaker and interconnect cables. Both speaker cables and interconnect cables carry an AC signal - with no DC component. This means that their normal resting voltage is 0 VDC, and their normal operating voltage is 0 VDC. Since both of those points are 0 VDC, interconnects and speaker cables are already at their optimum resting voltage if you just leave them alone, and they can't "charge up to it" over time. There are many electronic components that operate more linearly with a bias voltage on them. It's true for polar electrolytic capacitors, which are designed to operate with DC voltage, and not AC voltage, because the bias prevents them from operating at the wrong polarity. And it's also true, in a very different context, where a continuous "bias" maintains current flow in both of the transistors of a Class A/B amplifier stage, so there isn't a non-linearity at the 0 V point where one switches on and the other switches off - which is crossover notch distortion. It's also true for lots of mechanical things.... for example, the weight you set on your turntable provides a downward bias to keep the needle from jumping out of the groove. However, it's NOT true for most (non-electrolytic) capacitors, and it's NOT true for cables. If anything, they're hoping to eliminate some sort of distortion where the voltage "flips" from negative to positive and back again at the 0 VDC point. Oddly, though, nobody's ever been able to detect or measure any such distortion with cables. So that's for an Electrolytic Capacitor, right? This leads to the capacitance going up/down? Speed of charging/discharging going up/down? Casey capacitors are made from conductors and insulators (dielectric). The dielectric is not a perfect insulator and "charges" up over time (the absorption). Once fully saturated all is good. Until then the dielectric continuously charges and discharges adding a small element of noise to the signal. This is what the various dielectric bias systems (DBS) attempt to eliminate. Russ |
|
|
|
Post by leonski on Apr 14, 2017 0:38:31 GMT -5
Want a hellish test for amps? The original Apogee ribbons had sub-one-ohm dips AND ugly phase angles. If your amp can drive one of THOSE, it could drive anything. An ultrasonic 'chirp' followed by silence or a really AWFUL smell would be it for most amps. Krell built at least part of its reputation based on some of their amps being able to drive the Scintilla's awful load. The trick with amp testing is to find its limits in a non-destructive, measurable and repeatable manner. |
|
