|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
68.190.170.143
In Reply to: RE: Can resistors actually "sound crappy"? posted by airtime on September 09, 2014 at 06:09:06
Congratulations on passing your critical listening test. Every component in the signal path, and many in shunt positions or the power supply, can affect the sound.
I have experimented with a dozen or more types of resistors in my own power amps and active filters. The resistor that affects the sound the least is called a Z-foil naked Vishay. It is crazy expensive for something as simple as a resistor, but it is the ultimate in purity and accuracy, if that's what you really want. You cannot buy them at Mouser, but only from parts specialists such as Parts Connexion or the manufacturer, Texas Components.
As a general rule, carbon film resistors tend to smooth out details while presenting a more musical sound. Metal film resistors are more accurate, but can contribute to stridency, especially if they have magnetic leads or end caps, as most cheap parts do. Of the mass market stuff, I would stick with Vishay/Dale, which are non-magnetic. They cost only a few cents more than the cheapest metal films, and are well worth it. A bit more exotic, but still reasonably priced with very good sound, is PRP, made in USA, which are used by some premier equipment manufacturers. Not sold at Mouser, but at Parts Connexion and Sonic Craft. Yet another step up would be Caddock MK series. There are plenty of others, but these are the basic, best value parts. The "best" resistor does not always sound the best at every place in a circuit. It is a lot of fun to buy a few different types of the same resistance and experiment. Resistors take only a few minutes to break in, so your testing can proceed rapidly. You will be surprised by how much one tiny part can affect your system.
Peace,
Tom E
Follow Ups:
Words like "accuracy" and "neutral" used to describe hifi parts/equipment seem to be just a matter of the ear of the "behearer" to me, unless you're referring to some kind of electronic test results - and I dunno what that'd be.I'm a technical ignoramus, so correct me if I'm wrong, but......
If a resistor is required at some point in a circuit and you can't hear, say, preamp X without that resistor, when you put various resistors in that position how can you judge that resistor Z has no effect on the sound, or is more "accurate" than resistor Y?
I'm not claiming there's no difference in effect on sound, but I'd think its just a matter of preference rather than being able to determine one is more "accurate" or "neutral" than another.
Edits: 09/11/14
Which part sounds more like real music? That is the ultimate reference. All my experiments are done by listening, with no technical testing involved. Of course, that means my results are purely subjective. I believe that can still yield decent conclusions if comparisons are made impartially, which is not easy. If a component colors the sound, it is not accurate. The naked Vishays have a very solid reputation and specifications superior to other resistors, and I have determined that, in many positions in a circuit, they are indeed the best (very likely the most accurate) parts available. Of course, that doesn't mean everyone will like them, but most people do.
Peace,
Tom E
In the end its a subjective issue, and one man's "neutral" is another man's "unnaturally threadbare", one man's "naturally rich/warm" is another man's "muddy". For example, the tech I use (Ben Jacoby, pretty good rep) thinks Vishays sound terrible. IIRC he much prefers tantalum resistors, and maybe some types of carbon too.
As I said, so far I'm too ignorant to make my own judgment, but hope to rectify that in the near future. Thanks for the reply.
Carbon films should be almost as good as metal films, maybe a tad more noise, but it shouldn't be audible. I never bother with carbon film because good metal films are so cheap. It's so difficult to make a comparison between resistors that I discount almost all reports of audible differences between types.The metal foil resistors aren't all they're cracked up to be- go read what the Audio Precision people have to say after testing them recently.
Carbon composition is unstable so some might be pretty good, but they can be very noisy and the values drift. They have no place in modern design. Some say they're lower inductance, but if you have any lead length at all, that benefit quickly disappears. Put a standard 1/4 W metal film on a vector impedance analyzer and you'll see it does fine at 50-100 MHz. Metal oxide can give you a very small resistor (that can run burning hot) but their temperature coefficient is lousy. I avoid 'em.
In SMT, thick film resistors are a lot noisier than thin film (metal film).
There are no sonic properties of resistors that aren't apparent from measurements of voltage coefficient, noise measurements and maybe temperature coefficient. It's also highly debatable if magnetic end caps or leads have any audible effect, though material types might give you some small thermocouple effects. Again, the tests to prove it are far too hard for people to set up. You'd have to build up truly identical units to compare and few are willing to go to that trouble, then you still have all the issues of doing an unbiased comparison between two devices.
It turns out that you get a huge bang for your buck with common 1/4 W metal films from KOA or whoever.
(It's OK, I do have several pairs of flameproof underwear at the ready.)
Edits: 09/11/14
Some of your post makes sense, some of it is mere techno babble. One cannot measure what a component, even a resistor, sounds like. TC and noise are important, but they do not determine what a part sounds like. Comparable values do not always produce comparable sound. End of story, for me. I think most people post in this forum because they are interested in what improves sound, not what has better measurements. Too bad you discount reports of audible differences before even trying for yourself. Try swapping resistors in the feedback loop of an opamp. Then put away your vector impedance analyzer and listen.
Yes, carbon comp sucks for audio for the reasons you cited. Yes, the difference between carbon film and metal film is minor. I have HEARD very good and very bad examples of each. I really don't care how they measure, other than accuracy of value, which can be selected if 1% spec isn't sufficient.
No, it's not easy to test parts. I have spent A LOT of time doing it, and I have heard differences.
KOA resistors have magnetic end caps, and don't sound very good to me in unbiased testing. I did not even know they were magnetic until after I listened to them, but I knew something wasn't right.
I do not claim to have golden ears, but I am not afraid to use the average ones I've got to draw conclusions.
Peace,
Tom E
I used vishay bulk foils/z foils for many years. But some time ago I came to the belief that current production doesn't sound as good as older production (say more than 10 years ago). I noticed this originally when building a DAC and using such resistors as IV components.
Searching for an alternative I came across Neohm ultra precision wirewound (UPW), similar price to Vishays. Good specs - 3ppm temp coeff, very low noise, Imho, these sound superior to current vishays and even older vishays. Downside is that they are not available in the complete range. But recommended for critical locations. Just my $0.02
Are those readily available?
TE
Of course you can't measure what anything sounds like, but you can measure a bunch of other things and make a pretty good prediction. Those "other things" are very well established and understood, and can be measured to far higher resolution than any ear can discern. People are always looking for some "factor x" that hasn't been discovered, but affects how things sound. So far there's no evidence that it exists.Identical signals sound identical. Build two circuits, one with your favorite resistor and one with resistors you consider junk. Feed 'em the same signal, the music of your choice, and do a differential measurement of the outputs. If you see something, now you have something to investigate. If not, regardless of subjective opinion, there is no audible difference between the parts. It's really that simple.
Edits: 09/12/14
Yes, identical signals sound identical.
Signals that are not identical do not sound identical.
What is your point, other than that you believe everything that can be heard can be measured? The X factor of sound has not been found to be measured yet, so it does not exist. All SS amplifiers measure the same so they all sound the same. You're really in the wrong place here.
Peace,
Tom E
Well, yes, you're probably right on most points, other than solid state amps. They certainly don't all measure the same, nor do they sound the same. Do you have any reasons that different resistors would sound different? Anything testable or provable beyond a reasonable doubt?
I was being facetious about SS amps. Whether or not they measure differently, many do sound different.
I wish I had even a clue as to why such little resistive bits can make a difference. Someday, maybe an enterprising researcher will uncover a new way to measure the aesthetic qualities. Until then, I'll take existing specifications into consideration but always let my ears finally decide.
Some people insist that resistors are directional, but that's where I draw the line. I simply cannot conceive how the current flow can be different in one direction versus the other. I think they do hear something, but I believe it has more to do with the part "settling" somehow and will change back after a short time. I have heard it with speaker wire and IC's. After a few minutes of conducting, both directions sound the same to me unless there is an actual physical difference between the two directions.
Peace,
Tom E
Do you have any reasons why various resistors should sound the same? After all they all have different values for their measurable parameters such as Temp Coeff, Voltage and Frequency dependent changes in resistance, parasitic inductance and capacitance, modulation noise, thermal noise, etc.
For instance : Given that some resistors have TCr of 1ppm and others 1000ppm, and that this parameter directly affects one of the chief non-linearities (ie distortion mechanisms) of resistors, thermal modulation, please explain why you think that such different resistors would sound the same?
Also people also have tried and found audible differences between various types of resistor, and it's no coincidence that the top rated resistors usually have a very good TCr (eg zfoil 0.02ppm). So the all-resistors-sound-the-same argument is looking shaky imho.
I do not have a reason. I'm sure TC and other measurable specifications are a factor, but there seems to be something beyond that. Similar types, say MF with identical specs, do sound different. Magnetic properties may play an important role.
I wouldn't expect a resistor with 1000ppm TC to have a place in modern audio equipment.
By the way, have you read through the entire thread? You seem to have the wrong impression of my statements.
Peace,
Tom E
I should also mention that it's horses-for-courses when it comes to resistors. (As with all components). Sometimes low noise is paramount. Sometimes low distortion due to TCr. And sometimes what you need is really low inductance (eg grid stoppers, gate stoppers), in which case, carbon films can be the best. Blindly replacing all your resistors with zfoils is not necessarily going to lead to audio nirvana.
I wasn't really suggesting that blindly replacing all resistors in any circuit with only Z-foil was the path to nirvana, but it could be a path to the poorhouse. I use about six different types in my little chip amps, and I've tried many more. I can't pretend that what I have is the optimum configuration, as there are simply too many possible combinations. But what I use now was determined to be the "best" after MUCH experimentation.That was my initial suggestion: buy a few types and experiment. The OP determined by accident that resistors sound different. I agreed, and suggested he try others.
I think you meant carbon comp, not carbon film for low inductance. Even then, there are probably better choices.
Peace,
Tom E
Edits: 09/12/14
Long ago I designed an amplifier (non-audio) that ran quite high voltages. It had linearity issues that I eventually traced to some MOX resistors. I'm not sure if it was tempco or voltage coefficient, but I've been wary of them every since, though they're probably fine in supply, loading or non-signal related locations.
In a different application some thin film SMT resistors were replaced with cheaper thick film. The noise increase made the circuits useless. You guessed it, I've been wary of them ever since. Again, they're appropriate in many spots, but not in my audio circuits.
Having replaced bunches of carbon comps that drifted high in value or became crazy noisy over the years, I can't see using them except as a special effect. Once they drift, I haven't seen baking bring them back. IMO, desiccant storage is probably a very good idea.
Those are all real measurable effects. What I haven't seen is any artifact resulting from the use of decent metal films (or foils) that would lead me to believe an improvement could be had. I could fault (or praise) them in a DC metrology application for ppm level drifts or changes in value from soldering or the generation of thermocouple voltages, but those things just don't show up in (my) audio applications.
In a solid state amp the two most important resistors are probably the feedback resistors, as any imperfection there becomes a fundamental part of the amp performance. Those, and the input stage, are where the right parts probably make the most difference. Carbon comps aren't suitable, wire wounds aren't suitable, thick film isn't suitable, MOX isn't suitable, so what does that leave us? Metal film and foil. Then it's just a question of what brand and specifications.
why aren't wire wounds suitable?
Wire wounds are extremely quiet stable resistors, but being wound they do have some inductance. Even "non-inductive" winding patterns can't cancel it completely. Now, at audio frequencies this inductance is rarely an issue. The problem is that our circuits and devices often have far greater bandwidth than audio. As an example, the best audio opamps often have gain bandwidth products of tens of megahertz. Power amps may have response at hundreds of kilohertz. Keeping those circuits stable and well behaved requires thinking about component characteristics far outside the usual audio bandwidth. I spite of that, wire wounds can be used to good advantage in many locations, just not where a bit of inductance could cause instability.
I've read that it can act like a coil and cause high frequency problems?????? At least thats the poop in the speaker world.
charles
Well yeah a wire wound on a coil forms an inductor. That being said, metal and carbon film resistors have their conductive element flashed on a ceramic core and the element is laser etched in the form of coil too. The construction is similar in that wayOf course there are non inductive wirewounds and non inductive film types also.
Edits: 09/13/14
Especially as Emitter followers.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: