Sorry if this has already been posted, but has anybody seen this app note from AKM? http://www.akm.com/AppsNotes/layout2.pdf. Says to use carbon film resistors for best sound quality, talks about blind tests too. Certainly surprised me.
Carbon resistors, especially larger carbon resistors have less capacitance and much more hf losses.
Given the HF losses of big carbon, compare what happens if you use a metal film resistor for 120k (feeback) and a 1/2 carbon film resistor for 3.3k (towards ground) as the feedback voltage divider in an amplifier (sets the gain). Yes, it could possibly drop out the HF noise by the factor of the gain.
If overdone, treble may decrease slightly; however, the most probable thing to happen is that the amplifier becomes resistant to amplifying radio frequencies. There are audio applications for metal film too, such as making a remarkably clean sounding baxandall (bass-n-treble).
recently i have removed the vishap resistors and used mcap's mresist resistors on in my speakers...these resistors are real nice...the sound quality is much better than before
Hmm, its quite odd to have a "much better" effect like that at large signal. Perhaps the previous resistor (that was replaced) has much fiction in its advertisment/datasheet?
The usual approach tends to work as well--This is take two or three of cheap 5w resistor and parallel them (speaker resistors). This makes for half the inductance or less. And its not reliant on truth in advertising.
Some people get really brainy and parallel several of the really inexpensive 3w resistors because those have generally higher quality anyway. I haven't explored that technique.
As far has "hearing" effects this small, I don't. However, like many people I do feel a pressure/mild headache sensation when something is wrong, and a relief when its right. This sensation is oddly similar to "hearing" a very high pitch signal (at or higher than 10k), so perhaps it is harmonics of some sort.
A couple of years ago near the beginning of this thread someone said they could hear no difference in resistors. I concur providing a reasonable quality MF or MO is used. I am wondering exactly how testing was carried out by those who perceive a difference. In my experience of these matters most people don't a-b test properly. They make a change and then go back to their system. That is unreliable as mood changes, hunger, weather etc can make a bigger difference to perception than any small component. Only double blind testing is acceptable. Maybe James Randi will offer a reward to anyone who can prove a difference between a standard 1% MF and any exotic resistor.
I find it absolutely bizzarre that anyone would spend $20 on a single bulk foil resistor in any position. It also strikes me as strange that so little time is invested in exploring the properties of semiconductors. As the late Les Sage once said in an article Resistors and Capacitors are the goody goodies of components compared to transistors which are proven to distort.
I find it absolutely bizzarre that anyone would spend $20 on a single bulk foil resistor in any position. It also strikes me as strange that so little time is invested in exploring the properties of semiconductors. As the late Les Sage once said in an article Resistors and Capacitors are the goody goodies of components compared to transistors which are proven to distort.
I also find it odd to spend that much on resistors.
If you want isolation, decreased capacitance, and hf failure mode, a simple 1/2w carbon film will do, and in areas where there's gain this may pass a greater proportion of audio band to hf noise (or it may do the opposite, depending on design).
If you want decreased inductance, paralleling cheap resistors will do. . . exactly what inductors in parallel will do, and this can make less inductance than if you spent too much on an advertisement.
As for exploring the properties of semiconductors, transistors especially, it might be really awesome if you intend to build only one amplifier. However, I've been quite discouraged with discontinuance versus availability issues with transistors. Getting what you want is somewhere in-between impossible and unlikely, although it is doable, given enough time and resources.
It seems that the chip amplifier's reproducible results, plus a huge fight with availability of the support circuit components, in order to get exactly what you want, is an easier battle to win, although I don't want to say that its easy at all. In some ways, this is harder, because you still can't have exactly what you want, being that the chip maker has locked away many of the options inside the chip, thus getting the performance makes for a more intense job with the support circuits (if one expects hi-fi) being that the support circuits become more important than usual. And, so I wonder if success is more or less likely to be affected by availability.
Last edited:
Do non-inductive resistors make a huge difference or it depends on where it's used? or should I say, could I hear the difference?
for CRC type power supply, does it matter if it's non-inductive or not?
would you use cheap metal oxides or spend some more on non-inductive resistors? I'm assuming for psu, it would not matter.
thanks for your input.
for CRC type power supply, does it matter if it's non-inductive or not?
would you use cheap metal oxides or spend some more on non-inductive resistors? I'm assuming for psu, it would not matter.
thanks for your input.
The RC filter is sized specific to each individual transformer (and diodes) used, and although I'm not good at power supply design on paper, success in practice, with cleaner power, has the obvious effect of decreasing the temperature at the amplifier's heatsink--so, a temperature probe is a simple way to find out.
At audio frequencies, the self-inductance of almost any kind of film resistor is trivial. Temperature coefficient is a more significant measure of quality and carbon resistors are considerably the worst kinds in that regard, especially carbon composition.
Wirewounds may have enough inductance at higher frequencies, but non-inductive windings minimize this and lower temperature coefficient elements are available.
How audible is any of this? I don't know. Where I use film resistors I like to use the lowest Tc part, 100 ppm metal film is the minimum and Vishay bulk foil is by far the best. Tantalum resistors remain unspecified in any objective measure, but I assume they're the equal of metal film. I may have to buy a few and see what I can measure. Obviously you don't want to use a part where it will overheat, so a wirewound is necessary in some applications.
How much do you want to pay? That's up to you. In the land of audiophile excess, thousands of dollars are spent on speaker wire and interconnects. Is $20 so much to pay for a Vishay Z-foil 'naked' resistor? Well, it is for me; I'm a poor man and that's no joke. If I can't find it in the surplus market, or don't want to pay $4 or so at Michael Percy for a Vishay VSH conformal-coated foil resistor, I'll use a regular metal film and call it good enough. Percy sells those too. I think the PRPs and Holcos are pretty good, and regular Vishay Dales are pretty good too, at a much more reasonable price. Heck, a carbon film or even composition in non-audio circuits usually work just as well.
My first objective in selecting a 'best' part IS being objective; do the specifications support the contention that it is 'best'? I'll consider reputation, with some reservations and then the impact on my slight budget. Thus my reluctance to buy 'boutique' parts; much of the time I just can't.
What's maddening about specifying parts is the subjective business of how a particular part might "sound" in a particular application. At best there are only general guidelines and personal experience, and how deep your pockets are.
Last edited:
Hi Daniel,
Everyone,
Okay, the very first thing you need to do is to determine the magnitude of the parameter you want to discuss. In this case, inductance in a wire-wound resistor. Also, understand that the series resistance of the wire (chosen for it's high resistance) is in parallel with the inductance. What that means is that the impedance is damped due to a parallel resistance. Therefore, the inductive effect is diminished.
With me so far?
For those brave and curious souls who want to find out more, go for it. Either measure some resistors or read the data sheet. I have found inductive metal film resistors, but the effect of the inductance is only significant at a few hundred kilohertz and on up into the megahertz. Not so much in the audio band, but plays heck in a CD player or tuner section. Note: there is a clue in there somewhere that I have talked about before. In fact, not an issue at all, unless you are using a device that may happily oscillate up in the MHz when the circuit itself poops out at 350 KHz.
So Bengali, no. You will not hear the resistor unless it causes something to go wrong at high frequencies (MHz!).
So what should you be using in a power supply? Think about it a bit. A power supply will probably run parts at a higher temperature and there will probably be current spikes and surges. Considering this, you want a part that withstands surge currents and heat while remaining reliable. A wire-wound part here sounds like the right answer to me. The only reason to use many in parallel is to spread the heat over a larger area of the PCB, and to use a less expensive part in quantity. If they had high inductance, this would be good. You want to maximize this, not design to reduce it. This is valid for a normal, linear power supply. For switching power supplies, the rules change because you are operating at 50 KHz or more (easily 500 KHz or even a couple MHz these days).
Other good choices for our power supply are Metal Oxide. For current path areas, Metal film and Carbon film are both poor choices. Neither will stand up to current surges for long. Naked foil whatever is just very stupid here. In fact, the only places I would use these are in extremely low noise amplifiers or oscillators, and in low noise voltage references and error amplifiers. They are wasted everywhere else.
BTW, almost all resistors have a distortion mechanism called voltage dependence. The actual value of the part may change in response to a voltage change across it's terminals. Fear not though! No problem as far as the better quality parts go, because variable voltage effects are far greater in most semiconductors and many capacitors. BTW, by "better quality parts", I don't mean anything to do with price, I am referring to the suitability of the part to the circuit it's in.
Now would be a good time to mention that Carbon Composition resistors (the ones I grew up with) have by far the highest noise and voltage dependence of any other resistor type I know of. They aren't even stable over time or temperature, but they are completely non-inductive (except for their leads). That is why they are used in high frequency circuits or as gate damping resistors. See, whether a part is good or not depends entirely on the application. For audio use, Carbon Composition resistors are terrible parts. That is unless you are restoring something with all it's shortcomings intact.
-Chris
You really know how to make my head spin! In a word - NO!
Everyone,
Okay, the very first thing you need to do is to determine the magnitude of the parameter you want to discuss. In this case, inductance in a wire-wound resistor. Also, understand that the series resistance of the wire (chosen for it's high resistance) is in parallel with the inductance. What that means is that the impedance is damped due to a parallel resistance. Therefore, the inductive effect is diminished.
With me so far?
For those brave and curious souls who want to find out more, go for it. Either measure some resistors or read the data sheet. I have found inductive metal film resistors, but the effect of the inductance is only significant at a few hundred kilohertz and on up into the megahertz. Not so much in the audio band, but plays heck in a CD player or tuner section. Note: there is a clue in there somewhere that I have talked about before. In fact, not an issue at all, unless you are using a device that may happily oscillate up in the MHz when the circuit itself poops out at 350 KHz.
So Bengali, no. You will not hear the resistor unless it causes something to go wrong at high frequencies (MHz!).
So what should you be using in a power supply? Think about it a bit. A power supply will probably run parts at a higher temperature and there will probably be current spikes and surges. Considering this, you want a part that withstands surge currents and heat while remaining reliable. A wire-wound part here sounds like the right answer to me. The only reason to use many in parallel is to spread the heat over a larger area of the PCB, and to use a less expensive part in quantity. If they had high inductance, this would be good. You want to maximize this, not design to reduce it. This is valid for a normal, linear power supply. For switching power supplies, the rules change because you are operating at 50 KHz or more (easily 500 KHz or even a couple MHz these days).
Other good choices for our power supply are Metal Oxide. For current path areas, Metal film and Carbon film are both poor choices. Neither will stand up to current surges for long. Naked foil whatever is just very stupid here. In fact, the only places I would use these are in extremely low noise amplifiers or oscillators, and in low noise voltage references and error amplifiers. They are wasted everywhere else.
BTW, almost all resistors have a distortion mechanism called voltage dependence. The actual value of the part may change in response to a voltage change across it's terminals. Fear not though! No problem as far as the better quality parts go, because variable voltage effects are far greater in most semiconductors and many capacitors. BTW, by "better quality parts", I don't mean anything to do with price, I am referring to the suitability of the part to the circuit it's in.
Now would be a good time to mention that Carbon Composition resistors (the ones I grew up with) have by far the highest noise and voltage dependence of any other resistor type I know of. They aren't even stable over time or temperature, but they are completely non-inductive (except for their leads). That is why they are used in high frequency circuits or as gate damping resistors. See, whether a part is good or not depends entirely on the application. For audio use, Carbon Composition resistors are terrible parts. That is unless you are restoring something with all it's shortcomings intact.
-Chris
Hi Damon,
Huh, I guess I was typing while you were doing same.
In truth, you put this subject in personal terms that make sense, something I did not address at all. For me, most things are simply technical. You were able to relate in terms of impact to a DIY person who's read far too much in the way of partial truths.
Your post is a darn good one. I'm glad you chimed in.
-Chris
Huh, I guess I was typing while you were doing same.
In truth, you put this subject in personal terms that make sense, something I did not address at all. For me, most things are simply technical. You were able to relate in terms of impact to a DIY person who's read far too much in the way of partial truths.
Your post is a darn good one. I'm glad you chimed in.
-Chris
Well, I'm just hoping for nice clean DC from my power supplies. Any RC filter that I put in the power supply would be specific to zero out noise that comes from a specific transformer and rectifier arrangement. Megahertz? No, I'd like 0hz please. Thanks.
EDIT:
But, at least this thread did wander enough to mention that the power supply is very important to audio quality, and probably much more important than differences between similar resistors, brands of resistors, etc. . .
EDIT:
But, at least this thread did wander enough to mention that the power supply is very important to audio quality, and probably much more important than differences between similar resistors, brands of resistors, etc. . .
Last edited:
I've done a fair amount of equipment "mods" in my time and for most signal path positions have generally used Holco and Resista and more recently PRP metal film resistors. To be honest, I've never been able to decide which I liked best in terms of sound quality. However, recently I've built a few shunt-mode passive pots/preamps using NOS supplies of some discontinued resistors as well as currently available resistors and this has enabled me to clearly hear the differences in the various brands. The advantage of this simple circuit of course is that it lets you change only one resistor and listen to the effects of the change.
At 66 my ears aren't what they used to be, so I wouldn't presume to say which is the most accurate or preferable resistor (even though I have my favorites) but I can say that there is a unique sonic quality to each resistor I have tried.
At 66 my ears aren't what they used to be, so I wouldn't presume to say which is the most accurate or preferable resistor (even though I have my favorites) but I can say that there is a unique sonic quality to each resistor I have tried.
Perhaps you would enjoy the additional variety of adding a lightspeed attenuator to the collection?
There's also the discrete buffer/preamp projects at Decibel Dungeon, one of which has a multipass inverting mode arrangement on an amplifier that is a downscaled 1950's amplifier style modeled into a transistor preamp. The manufacturing variances in resistors and other components could be noticeable because your power amp will apply gain to signal from this preamp.
Thanks for the suggestions. Yes...I've read about the Lightspeed and would definitely like to try it at some point. However I'm not sure I'm up to matching the LDR's (right now). Also, I think what I have is pretty darn good (and for very minimal time and $). Good enough that it has replaced my excellent active preamp for line level sources. With the right combination of components (pots, wiring, connectors and, of course, resistor) it can sound exemplary.
Joachim
did you notice any change of sound after a certain period (break in)?
When yes please let me know.
Thanks
In opposite to Vishay Bulk foil wich need 5 hours of burn in to get the bass tight I do;nt think Charcroft does.
Erlend,
Just saying the PRP's is the worst sounding resistor does not help us move this forward. Please give us some details, especially the equipment you tried them in. I for instance have used them in a Hagtech Clarinet and in my ARC clone preamps. I like the way they sound. They do tend to be somewhat thin, especially before they break in. I find they take at least 100 hours to open up. They are not for people who want an old fashioned big tube sound. They do work great with the 6H30 stuff I work on.
John
Just saying the PRP's is the worst sounding resistor does not help us move this forward. Please give us some details, especially the equipment you tried them in. I for instance have used them in a Hagtech Clarinet and in my ARC clone preamps. I like the way they sound. They do tend to be somewhat thin, especially before they break in. I find they take at least 100 hours to open up. They are not for people who want an old fashioned big tube sound. They do work great with the 6H30 stuff I work on.
John
- Status
- Not open for further replies.