2. amps-preamps
  3. 1315759817

Why 24/7 warm-up period on amps?

The 24/7 warm-up period on amps seems excessively unsupported. Yes, an amplifier (pre-amp or power amp) will change it's circuit factors as the init heats up since the resistive and capacitive values stabilize...but for months on end? Do we still have a "warm" heart for tubes, that do indeed need to get "hot" to work right?

A capacitor charges up based on it's RC time constant, which is in the SECONDS range, not days. OK, if you add the heat sink area so the heat going out is stabilized I can see maybe an hour or so. My DNA-225 gets HOT in thirty minutes, at which point it's steady state. That even assumes it doesn't have temperature correction circuits to make it more stable, and less subject to change over time.

Break-in periods are hard to judge what people think is happening. Circuit P/N junction temps get hot pretty fast. A mechanical device like a speaker or phono cartridge, sure, they will work-in just like a well used rubber band. But silicone? Factory burn-in is designed to find weak components that degrade outside of SOP ranges, not to "center" their attributes in a normal stable circuit. Did someone forget to add enough heat sink compound to a PNP or NPN transistor, for instance?

Assumming break-in is real, not to be confused with the warm-up period, once it's done it's done. After that it would be warm-up only time. And, warm-up is a simple thermodynamic process. It only takes so long to warm-up and it isn't "days" on end. Maybe hours...if even. Once things are to temp the circuit constants are set. What else is changing? A heat sink is designed to warm-up and hold a delta temperature where the measured performnace is flat. A small amp (pre-amp gain stage) has smaller heat sinks for this reason. Heat and resistance are related, so you have to pick a temp and hold it. You design to THAT attrubute on the component.

Wire conditioning in the amp? ( go here - http://www.angelfire.com/ab3/mjramp/golopid/grain.html) As well as several other sites and textbooks.

The DC path is just that, DC. The magic is the purity of the DC, not the wire moving it around. You either have the right voltage and current capability (wire size)or you don't. Once the amp is on, the wires capacitance hardly matters. PP, PE or Teflon dielectrics only ionizes-tree and fail at break down voltages around impurities, not below that. You do not want to ever ionize the insulation in normal practice.

AC is an interesting issue. The AC complex signal is ALTERNATING differently at each and every frequency point, so the magnetic and electric fileds keep switching with respect to frequency. So the dielectric can not have polarity, or current "direction". The dielectric will not "align" to anything.

Grain structure in copper does not change unless you melt it. It's set when the rod is made. Annealing just resets elongation by improving homogeneous grain alignment, not the grain boundary characteristics since wire is resitive annealed at well below the temp that would fully reform the grain boundary around impurities in the copper. Oh, all modern 9/16" rod copper is made in induction ovens and is essentially OFC grade. All wire is drawn from that rod. Modern copper is also "high conductivity". Again, these terms are throw backs to days gone by with coke furnaces and open air annealing to critical temps where impurities could be picked up, changing the grain boundaries around impurities.

I also notice the people seem to tout TEFLON over Polypropylene or polyethylene dielectrics. Teflon costs more, it is higher temperature capable to 150C-200C (like 80C on polyethylene isn't enough in electronics) but Teflon has a worse dissipation factor and loss tangent. Using Teflon has a more NEGATIVE influence on electricals than olefins. Teflon's velocity of propogation at RF frequencies way above 1MHz is 70% verses 66% for solid olefin dielectrics. But that is at RF. And, you can nitrogen foam either to negate that advantage of Teflon at RF, but NOT Teflon's high price, loss tangent or dissipation factor. Capacitance adjusted Teflon is a poor choice. So the important factors are capacitance, dissipation factor and loss tangent. We can easily fix the velocity of propagation. PE and PP is superior across the board and cheaper (that's probably the problem!).

Good circuits are good circuits. Could you even make a circuit that had electricals parameters that were undefined till it ran, "forever"? Nope, can't be done. Design would then be a game of chance. I don't think that it is. Stabilized junction temps are used to set electrical componenet attributes with respect to temperature. You can design heat sink characteristics to place "hot" components where thet need to be temp wise to meet a circuit requirement. A poorly designed amp that allows thermal run-away under load isn't appropriate and isn't made...for long. There is indeed a circuit junction temp that rather quickly defines the measurable performance of the circuit, and a STABLE delta attribute approximation(s) when a circuit is designed. You know going in what they will be in operation steady state.

So, I hear my speakers and phone stage "break-in. And they don't go backwards once thet are broken-in. They can, in fact, get worse and simply break-down! But my amp sounds fine in short order. The circuit reaches a thermodynamic steady state and we're off to the races. I just can't see a circuit that needs 24/7 "on" period to stabilize...unless it just isn't stable. To me that's a poor design, and one subject to possibly serious load induced instability when the circuit falls outside of the stable design region(s).

I'd sure like to see MEASURED attributes that support 24 /7 warm-ups on sound. I have yet to see any measured data to support this. Show me components used in amps that take MONTHS to reach stady values. I have read PLENTY to support first to third approximation(s) on amplifier circuits ambient thermal temperature stability points. Many circuits are designed to run "cold" and have inverse circuit systems to keep changes due to temp deltas away. This way, you have a more stable circuit at all times. The opposite designis technically UNSTABLE till it gets to temp. This also limits what you can do as it can't blow-up when it is cold BEFORE it gets hot and stable. So the circuit is a compromise.

So just what are the resistive, inductive and capacitive break-in periods on quality components used in a circuit? In God we trust, all else bring data.- unknown
rower30

Showing 8 responses by rower30

rower30 OP
211 posts
 
Well, I see I lot of, "I thinks..." and no real data past typical steady state warm-up (maybe an hour). SS circuits are pretty immune to on/off transients so paying more in electricity than an amp cost (A DNA-225 is over $250.00 to $300.00 a year to run at 135 watts at IDLE!) over it's life seems wasteful.

It seems funny to me everything in audio is better "hot", and no one seems to like it "cold". Circuits like to be COLD! But, you can't keep them COLD! So designers work around that and design to "hot".

Then we get to a PC and enthusiasts we want to cool the heck out of it! Well, an amp is designed to reach a known temp delta and be linear in that range. It shoudn't take "forever" to get there. New KRELL amps run at virtually no power draw at sleep, so they can't be "on" all the time. Not anywhere near RUNNING steady state, anyway. But we all know how good KRELL amps are, right? Or any amp made to sleep.

So past "brreak-in" a concept that shouldn't take too long (find the stuff that won't stabilize at the operating delta when cycled) the forever-on situation seems dubious at best. Yes, I know, we can pretend this is religion and just "believe" it.

I'd bet a bunch of bucks that you can't hear the difference in an amp that's reached steady state temps and one that's been there a year.

I'd like to do a blind test on power cords, too. Power goes all over the county on high voltage lines, step it down at ana xformer, goes through the house in 12-14 AWG ROMEX, through a gazillion dollar power cord that magically realigns the power (how it ever "knows" what it "was" when it left the plant I'll never know) and THEN goes into a CRAPPY IEC outlet, and THEN into the simple wires from the IEC outlet inside the amp to the power DC supply circuit block. Any decent DC block will turn all the AC to pure ripple free DC. If it doesn't, no power cord is going to change that. DC is like "white", there isn't anything there but potential to do work. No phase, no time shift, no frequency variation, just potential energy. Your best lead-in is the ROMEX from the wall! Run that right into your amp. Why make the AC get screwed-up through yet another passive lead? Active leads? Your power supply is agnostic to active leads as it spits out pure DC.

Speaker leads are different, the amp is trying to pump a signal down a load that is trying to force voltage BACK into the amp! Hook a VOM to a speaker and move it and see what you get, a voltage, one that is apposing what's going in the speaker! Speaker leads help, or hurt the amps ability to deal with that back EMF characteristic. It isn't magic, the complex sinusoidal addition of the back EMF can distort the forward EMF and cause cancellations. SHORT leads make this less a complex circuit. So yes, speaker leads are the the MOST "hearable" leads in your system.

High impedance interconnect leads? A precision 25 AWG mini coaxial cable with low capacitance is what you need. Nothing fancy, just good quality. There is (should be) virtually no real current in these leads. The lead is a first order filter (resistor and cap to ground), so the longer it is, the worse the roll off at high frequencies. Keep them short. Special "powered" leads? These isn't enough of an E field to scare a house cat at audio let alone "ionize" the dielectric. And, you better darn not have much of a magnetic field (current induced) or you input impedance is too low and it will swamp out you output amps. Remember, what's on the opposite side of your RCA is UGLY. But we forget what's inside the "box" don't we.

Pre amp to power amp leads need low cap. too. Big ass copper does ZERO at the lengths we're talking about. The attenuation is NOT the problem in leads with no current (E=IR). There is no real voltage drop across the leads. What there is, is an increase in capacitance as the leads get longer (say bye bye to higher frquencies at 3 dB per octave at the filter fo frequency which drops as the leads get longer). So all the gold and silver on earth won't do anthing except take your money. What WILL help, is short LENGTH and low capacitance.

Balance leads have HALF the capacitance as single ended RCA leads. The two wires are about TWICE as far apart, leading to LOW capacitance (about 8 Pf ft verses 17 PF/ ft). So the roll off is half the single ended RCA's roll-off. And, you get NOISE rejection added in for a bonus.

I want to believe all sorts of stuff. But I don't. I can't seem to think I know stuff when I don't. I can hear every thing I said and 'show' you. A steady state amp is a happy amp, be it hot or cold. It just as to be steady state and designed around that temperature.

Do you guys and gals even know how sloppy a typical PNP or NPN "junction" is is inside a transistor? And our beloved SOUND goes through all kinds of these devices? We can't readilly see it, so we pretend all the pretty stuff matters.

So send me data on transistors that explains their phase, amplitude and gain changes with forever-on after steady state and I'll "listen" to that. I won't even get into resistors, which are terribly temperature dependant.

So all the, I just know people can just know. That doesn't help anyone. The one's who really know, and have data they can point to, chim in.
rower30 OP
211 posts
 
Well, I hit a hot button for discussion it seems. Do mean this to be a "learning" experience, not arguments. Sorry for any flames going any direction on this, even if I'm just a catalyst to it all!

First order filter should be 6dB / octave not 3dB. It is a VOLTAGE amplitude drop. So it goes 6, 12, 18, 24 dB ETC as you gang filters in series. So, you have to be careful to use 20LOG and 10LOG in the right places. I was thinking optical cable half POWER bandwidth point, which is measured at -3dB. Wow, sorry about that!

"These effects can be quite measurable!! For example, I have seen a 3-volt drop across a 6 foot power cord cost a tube amp of about 35% of its total output power. If you want a reason to look for, that one is pretty basic!"

No, a power supply cord is in SERIES with the ENTIRE ROMEX lead and all the way back to the power station. An insignificant voltage drop and even less so if the lead is equal to the ROMEX in AC resistance. So if you use a 12AWG cord you're NOT going to see appreciable voltage drop across the cord. Do a voltage divider rule to the ENTIRE DC circuit! The AC circuit could care less about what we just paid for the cord. What you WILL find, is CRAPPY wall and IEC sockets that are independent to the cord used. THAT is what you measured. The wall outlet has a limit of 20 amps, no getting around it. You can only get so much continuous DC from that. LARGE filter caps can leverage power gained over time, to the SAME net power over a shorter time. Some call this dynamic power or "high current". But the TOTAL power over time has to be the same.

All things being the SAME (dielectric and conductor size and type - stranded or solid), a 150-ohm twinaxial (two coaxial in parallel) will ALWAYS have half the capacitance per foot. The shield is taken into account when you set the insulation thickness or, you would NOT have a 150-ohms cable would you? Of course not. Yes, if you buy a 120-ohm twinaxial cable the capacitance is not half, but it isn't apples to apples on the "root" insulated conductor. And, the conductor spacing will be twice the spacing or MORE with the same root single ended conductor impedance.

RCA leads do NOT "shield" magnetic "hum" at all. This type of interference is diffusion coupled clear through the shield and non-magnetic (foil or copper) shields. So, to get rid of magnetic coupling requires a TWISTED pair and differential mode transmission with common mode rejection - sometimes called CMMR. The "shields" on audio cables are RF only, and really do very little at that since most inputs use a RF filter cap to ground to strip off RF. It's more show than go at audio. That, and it is CHEAP to do over balanced lines. Power cord have shields to mostly BLOCK crap going INTO the A/C grid FROM your equipment (PC monitors ETC). Collapsing electric fields across a resistive connection throws off RF (lightening and AM radio). Magnetic lines cancel at ninety degrees, but it's impractical to be always ninety degrees and go "forward" to the amp! Sooner or later, you have to turn the cable.

The only other method to block AC hum, is to use ferrous or "magnetic" type materials that intercept and reroute the flux lines around what's inside the shield.

Poor audio circuits can be swamped out with RF, which is terribly inefficient to amplify. So terrible sound can be rooted in RF leakage into the gain stages, clipping the circuits and throwing the audible band into chaos. Super wide band amplifier stages can simply amplify the RF, and NOT clip the audio levels so much. It doesn't matter where clipping happens, it goes all the way through at the audio level after the stage.

Still, I'm NOT a taker on amplifiers ideally needing to be "warm" to sound good (The P/N junction are actually HOT as heck by nature, and run better when COLDER. The new Macintosh amps are set-up to be COOLER running to IMPROVE reliability AND sound. Sure, old school amps are biased to be hot since they didn't take the time to design around a COOLER thermal equilibrium, and the amp was ONLY at design attributes when it was warm. But, THAT point is static after about twenty minutes or so. Transistors like to be cooler to work well, and last a long time. COLD is good, it's just that the DESIGN can not keep the circuit cooler that FORCES a higher thermal stability inflection point. Designers don't WANT that. There isn't a SINGLE attribute on a transistor, inductor, resistor or capacitor that likes HEAT to be better.

You guys and gals go ga-ga over your low loss leads, but never give to the notion how inefficient a HOT transistor is? HEAT is HIGHER resistance everyone! That means you need MORE gain to offset the HEAT, which leads to more circuit paths, and LESS purity of the signal. Sooner or later, the HEAT offsets the GAIN, and the amp design collapses in on itself.

So what you "hear" is an amp that need twenty minutes to be at the DESIGN values of all the components using contemporary (older?) circuit designs.

As for 24/7? Sorry, I don't buy it, and my ears don't buy it and I don't want to buy the electricity to buy it! After twenty minutes my DNA-225 is SILLY hot. I'll turn the radio on for twenty minutes and catch the news, than listen to music to forget all about the news!

Oh, I still contend that speakers, cable / amps are heavily interrelated due to back EMF and how the amplifier's output stages are effected. The amps damping factor with respect to frequency, complex amplitude and phase cancellation does indeed alter what you hear. Each amp likes a different cable to alter the negative effects of the speaker, and itself. But, you can't change your amp or speaker, so we change cables. Some argue (DynAudio) that speaker leads should be CONSTANT impedance at audio. But, some amp and speaker combinations may not like that "ideal" world since the amp and speaker aren't ideal. No ideal LOAD (the speaker) throws a signal BACKWARDS to the amp! So much for ideal circuits.

The pre amp leads, not so much. HIGH impedance terminations and LOW current don't interact as badly. The do have "issues", (there is current) just not so much.

So of all the discussion, I 100% agree on speaker leads contributing to sound (I can hear it). Power leads and pre-leads, no, not so much. I haven't heard it. And, on pretty good stuff; Benz Micro Ruby 3 with Vandersteen Quatro SIG wood series II speakers.

Good stuff everyone.
rower30 OP
211 posts
 
Yep, and the IEC plug is rated at just TEN amps. Count them; 1,2,3,4,5,6,7,8,9,10 AMPS! That's all you get. A far cry from what people realize. The conductors in the cord are NOT dropping the voltage you are measuring. Can't be true. You are measuring the terminations, like it or not. This supports bringing the ROMEX right into you amp's transformer leads (many do this). Two terrible connections GONE!

A power cord cannot reconstruct what should be there, it has no idea what "should" be even is. It can only "change" it into something else yet again. Not a bad reason to convert it to DC is it? Screw the A/C other than connections, make sure the DC is done right. Most people use power conditioners to offset crappy DC power supplies. There is no other reason. True DC has no master other than the potential to do work.
rower30 OP
211 posts
 
We had a customer actually bury 9/16" SOLID copper rod in his wood floor to his speaker terminals from the power amp. And, he ran the ROMEX into his power supply. Way cheaper than the silly cord. If you have a power conditioner, the power cords make even less sense. Cords have no active way to condition the power at all. AC is AC regardless of the phase angle when it is rectified to DC. Noise on the line? Sure, a big old bank of capacitors can eat that up.

I haven't seen a correlation to what's on the transistors DC V+ bias voltage showing in the audio path. Not to say the data isn't trere, I haven't seen it shgown in an output signal test. Power supply ripple is pretty darn low anymore.
rower30 OP
211 posts
 
ROMEX is 100% soft annealed copper, it is simply the AWG size that limits flex. You can run equivalent AWG stranded copper cords straight from the box (hard-wire, no connectors)to your device. AC is 60 cycle, it only knows resistance (AWG size).

You can ALSO run ROMEX in adhered ridgid EMT or unadhered flexible EMT conduit to your device and, with the suitable conduit interconnects. So it can be done, safely, with ROMEX in conduit. Done all day long in industrial plants to REMOVE crappy connections anywhere possible, THOSE are fire hazards!
rower30 OP
211 posts
 
Plants run THHN wire in flexible and solid conduit all day and night. It's SAFER than ROMEX run bare inside your walls! Yes, flexible conduit is UGLY in your house, but the NEC, which the fire marshal follows, allows this circuit.

People confuse flex life with flexibility. They are not the same. SOLID THHN wire in flexible conduit will last just fine based on the conduit's self limiting min bend radius. Of course, SOLID WALL conduit has to be FIXED service because it will kink when it is bent and pinch the wires.

Some circuits REQUIRE SOLID wire so it is either "good" or "bad" (fire alarm wire). When solid wire is "bad" it's an open or very high resistant "touch" circuit that can be tested for. Stranded wire can lose strands over it's life and burn off when full rated current is drawn after the wire ages and strands break. Stranded wire is responsible for the majority of fires for this reason.

I just bought a used PASS LABS XP-10 pre amp, and it does have a "stand-bye" mode, which powers off a good part of the circuit. But, a 10 watt draw is pretty light. Circuits that draw VERY small levels of power would be designed accordingly, and to the typical junction temps of those circuits. There is no magic in circuit temperature stability.

I'm not talking about circuit life, but "sound quality" after thermal equilibrium. Shoot, most of you change stuff before a layer of dust gets on it! I buy half my stuff used and have no fear at all with SS equipment. My Apt 1 amp is 30 YEARS OLD and still fine driving 1979 B&W 801's! The APT 1 amp has no power switch, you plug it into the wall. No, it isn't for 24/7 "sounds better" it's because a SWITCH is GARBAGE and over rides every piece of wire in the power circuit, so he took it out.

I'm still amazed at how little the IEC plugs, wall outlet plugs and power switches are really looked into. They are the predominat part of the circuit. But, we may silly money for "wire" and ignore the critical plugs and switches.
rower30 OP
211 posts
 
Let's see...we see a big voltage drop across the power cord, but ignore that TINY little device called a FUSE that EVERYTHING power related goes through!

Wow,I guess a $200.00 Rotel integrated amp would be simply crazy good if we bypass the fuse. Why spend more and risk blowing something nice up with no fuse! The change would be so dramatic that spending more would seem silly.

Seriously, it's hard to look at 10 AWG wire and ignore the worst link in the chain. The fuse IS THE CIRCUIT for all practical purposes unless someone knows something about a series circuit I don't. The voltage drop will pick on the worst component in the chain.
rower30 OP
211 posts
 
"However you can also consider the fuse to be a common denominator in the AC chain, in that you might use the same fuse while auditioning different power cords, and still be able to hear differences between"

I would just LOVE to do a real blind test on this!! I sold stereo for seven years, and we did blind tests all the time after work. Oh what fun, and NO, I was almost always "wrong" on what was what on wire, cords, stands, rock weights, and if my pet rock really loved me.
More to discover