Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[polypals]

Opamps often have more open loop gain than is needed.
Reducing 100 dB olg with80 dB nf is not sound design practise, no matter how fast the opamp  is. It leads to high amounts of transient intermodulation distortion. TID is bad for your ears.

[Kai]

Opamps often have more open loop gain than is needed.
Reducing 100 dB olg with80 dB nf is not sound design practise, no matter how fast the opamp  is.

Sorry, but this is not the case for usual cheap audio purposed OP-amps.
TL 07x eg. has an open loop bandwidth of ca. [email protected], not much reserve for NFB in an gain or filter stage.
OP-amps generally have very high open loop gain (up to around 120dB), but only at DC or very low f.
They cannot be used without high NFB, or to say it better, high NFB is a working principle on OP-amps

nf is not sound design practise, no matter how fast the opamp  is. It leads to high amounts of transient intermodulation distortion.

TID doesn't come from NFB, but from circuits too slow to follow the signal, so that NFB cannot cure the distortion.
If TID appears an amp can be considered as faulty design, or used for the wrong purpose.

Higher NFB always gives cleaner signals as long as the whole circuit works within it's limitations.
This means with not too high frequencies or to much output current or voltage.

The question is, if high linearity is to be considered the goal or if certain types of nonlinearities are welcome to enhance the signal.
This is certainly the case with lot's of tube devices.

Regards
Kai

[Jim Williams]

072 type opamps with 40 db open loop gain at 20k hz will give you 25 of loop gain correction via negative feedback. That isn't much but they still show .001% or less THD at unity gain.

5532 and modern 'audio' opamps have 60 db OLG at 20k hz giving an additional 20 db of loop correction at 20k hz.

Some do better than that, the ADA4898-1 has about 80 db OLG at 20k hz, much better. The National LME series have about 75 db OLG at 20k hz. Combined with decent slew rates, CCIF IMD is at residual levels on the Audio Precison analyzer. That is .00015% or 1.5 ppm. I suspect it's even lower as it will not increase using noise gain techniques to expand the analyzer's measurement range.

Bob Pease of National did some tests using a network analyzer and some noise gain techniques on the LME49720 opamp. He determined actual THD was at -154 dbu once the noise was factored out of the measurements. In this situation, negative feedback is a beneficial effect as I know of no other active devices with errors that low.

In other words, a modern opamp is not a problem.  Nor do they need a break in period. Some film capacitors do and there may be some discussion on whether a plastic diaphram capsule may also need it, it is after all a film capacitor too.

[klaus]

The last two posts are WAAAY too esoteric for most participants of this forum. Could both of you please paraphrase, condense, and simplify their message?

Thanks!
KH

[Uwe]

Neither of these Op-amp design topics relates to the original rather strange question of "Breaking In" microphones.

In a 50+ year career of intimate involvement with vacuum tube and solid state electronics, acoustics and transducers, I have yet to experience this phantom phenomenon so passionately described without solid scientific explanation, physical evidence or credible measurements. The need for tube warm-up is a different topic, as is the long term aging of some components. Operation of devices outside their design limits, for example exposure to excessive SPL may cause breaking them, not to be confused with "Breaking In".
"Breaking In" electronic devices is also advocated by various Hi-Fi enthusiasts who are convinced that running loudspeakers and headphones for several days with low level white noise conditions them and their connecting wires...

In short, "Breaking In" your microphones (or other transducers) has no verifiable benefits, but, aside from potentially shortening their life expectancy, does no harm.

Uwe Sattler,
Technical Director at Sennheiser Electronic Corp. and Neumann|USA (retired)

[klaus]

The breaking in of components and its possible effect on sound is a controversial subject, because the phenomenon- if it exists- is hard if not impossible to "verify" as Uwe mentioned, certainly with conventional electronic testing apparatuses.

Further discussion depends largely on whether we are willing to respect subjective impressions by those who claim they hear the difference. Which, in turn, pretty much eliminates skeptics from the discussion.

But skeptics do not like to be eliminated from discussing a phenomena which they do not regard as real; so they want to discuss the subject further, but with opponents whose opinions they fundamentally do not respect or accept. Likewise, believers in the break-in phenomenon do not accept the basic premise of the skeptics who deny its existence unless believers can produce scientific proof.

Two ways to deal with this issue remain, as I see it. Either you:

1. categorically deny the existence of any phenomenon which you cannot measure, or whose existence you cannot explain. If that is the case, you should stop arguing, let it be, and move on.

Or you:

2. experienced the phenomenon as real, and would like to learn more about its existence empirically, even in the absence of scientific proof rendered by current measuring devices. You wish to deepen your understanding of this phenomenon by sharing under what conditions, in what circuits, what voltage environments, with what components, etc. you have witnessed component break-in, and what effect the breaking-in had on the sound of the device.

I have experienced some, but no all, of the various types of breaking in a component discussed and reported in HiFi magazines. For example, I usually do not send out mics in which I have performed major circuit upgrades/restorations/repairs right away, but 'break them in'. In this case that means leaving the mics on under realistic conditions for two to three days, before I send them out.

The only difference I consistently hear is in broken-in capacitors when they are in prominent audio circuit positions (i.e. coupling caps). Breaking them in removes a bit of midrange harshness, I have found. How do I know that for sure? I have on occasion switched out a high quality coupling capacitor installed in a test mic, which has been in use on and off for ten years. When I install that capacitor in a similar mic which also uses the same exact model and vintage of capacitor- only new/unused up to that point- I hear immediately that the new one is harder and less musical compared to the broken-in part.

The only other component where I can hear a difference and where I  have made the same tests of switching-out the same brand, model, and vintage is the mic cable: I usually connect FET/solid state mics I work through a 25 foot Gotham GAC 3 cable to the mic pre and phantom source. Occasionally I switch out that mic cable with a brand-new, unused one from the same batch as the cable I have been using for the last decade or longer. It's terminated with the same high quality connectors and has the exact same length.
Here, too, I recognize a smoother midrange in the broken-in cable (the rest of the frequency range and the dynamic behavior seems unchanged). The difference in midrange resolution is not as stark as with capacitors, but it improves in the same way: better connection to the music, with less artifice in the way.

I cannot extrapolate my experience with breaking in these two types of components to others, where I have not heard differences. And I do not expect or need others who read this to be able to duplicate my experiences. But I will continue to follow through with breaking in components whenever I think they make a positive audible difference in outcome.

P.S.: To finally address the original question: I have not heard an improvement in the sound of condenser capsule membranes with time or usage.

[hasbeen]

The only other component where I can hear a difference and where I  have made the same tests of switching-out the same brand, model, and vintage is the mic cable

I cannot extrapolate my experience with breaking in these two types of components to others

I am trying to wrap my mind around the mic cable phenomenon.

When you say "I cannot extrapolate" does that mean you have no explanation as to why a 'broken in' mic cable sounds different from a new one?

[klaus]

Sorry, I was obviously using language that was too sophisticated for my own good, i.e. muddled.

I meant to say: what I have experienced with break ins of capacitors and mic cable, I would not automatically assume to be valid for other types of equipment which I have not tested the break-in hypothesis on.

[hasbeen]

Sorry, I was obviously using language that was too sophisticated for my own good, i.e. muddled.

I meant to say: what I have experienced with break ins of capacitors and mic cable, I would not automatically assume to be valid for other types of equipment which I have not tested the break-in hypothesis on.

Still, I am interested to hear your thoughts as to why a broken in mic cable might sound different than a new one?

[boz6906]

I suggest 'Dielectric Fatigue' makes the cable's velocity of propagation non-linear over a varying voltage...

DF would also affect the capacitance of the cable as the signal changed, dynamically changing the cable's HF attenuation.

Maybe...

http://ieeexplore.ieee.org/xpl/login.jsp?tp=&arnumber=5310594&url=http%3A%2F%2Fieeexplore.ieee.org%2Fxpls%2Fabs_all.jsp%3Farnumber%3D5310594

[Jim Williams]

Propagation delay in cables are measured as a percentage of the speed of light.  70% is a good spec if it's linear vs frequency. Most audio grade cables will show drop-offs at higher frequencies.

Besides high end Kimber silver I use Belden 9182 LAN network cable here for mic connections. The reason is propagation delay at 70% up to multi-megahertz, 8 pf per foot stray capacitance, heavy 22 awg FEP teflon insulated wires and 150 ohm impedance. It also sounds very good too.

In 1994 I brought some rolls down to Paramount Picture's foley building for the AE's to try out. They ended up wiring the entire building with it even thought they just repaired the building after the 1994 quake. The AE's demanded it so it was installed.

I brought some rolls of it to Ray Kimber's factory in Ogden, Utah. Ray has a great lab with nice test gear like HP network analyzers and such. We compared it to his pure silver AGSS 19 awg stranded 3 braid. Tone was revealing on both but the silver had that little extra detail. The largest difference was background noise, the AGSS is silent even though it's a 3 braid, not a traditional coax design.

That background noise I hear in every other audio screeened cable, including the Belden 9182 LAN cable. You have to audition Ray's stuff to understand that we have all accepted cable noise our entire lives. Only when it goes away do you realize we have pushed that noise out of our minds while working. When it goes away it gets your attention.

[klaus]

Still, I am interested to hear your thoughts as to why a broken-in mic cable might sound different than a new one?

I really do not have any thoughts worth posting. This field is new, and controversial, as far as scientific explorations into it are concerned. Remember, we are still in a time period where a reputable audio engineer (and the then-importer of a well-respected recording device) put on a seminar at the University of Berkeley, to prove that lamp cord sounds exactly the same as a high quality speaker cable.

I am confident, that, as measuring technologies advance, we will find out what molecular or other changes occur in a cable over time.

[polypals]

A lot of differences in audio cannot be explained with measurement methods well known and accepted.
Compare a solid state power amp with a tube amp both with similar frequency response and distortion specs. I am pretty sure the tube amp wins.
The conclusion tube electronics are better is not correct. We simply do not have the right parameters to compare.

There is only one valid means to use, that is our hearing. That does not make the job easier.
On the other side most people who listen to recordings give way to their brain to make up for any deficiences. It is just a bunch of professionals that are bothered. These are the people who do not accept shortcomings and keep trying to improve the process.
They often have to stand up to engineers and designers that do not understand the findings of  professionals gifted with good hearing.

[Uwe]

The process of hearing is highly subjective and involves unprovable biases and psycho-acoustics. I still say that a repeatable measurements are worth a thousand 'pseudo expert' opinions. The assumed changes over time (not to be confused with breaking-in) of dielectric properties in any cable conducting electrical signals would be far too small to have any effect at any property for audio frequencies (frequency response or phase), period!

[polypals]

Human hearing is the most valuable if not the only instrument to judge quality of equipment.
To degrade human perception as subjective and influenced by bias and psycho acoustics is nonsens.

Blind tests based on scientifically and statistacally respected methods give reliable conclusions.

These are not to be confused with a line up of twelve loudspeaker pairs that can de selected at random without even compensating for efficiency. 

Please give me one measurement that predicts the perception of what the human hearing registers.
Leave out transducers, that field is even more complicated. Lets narrow the case down to analogue amplifiers.