Great thread
The thread marries together some audiophile discussion with some truely interesting technical discussions. The level of civility is excellent despite these disparate views.
Yet, rock solid conclusions and advise are hard to decipher on Resistor island. Roger Modjeski's cap thread (Favorite Coupling Cap Brands and Types) is a little bit more millitant on engineering and he states that "A crossover is one place where caps can make a difference and I do believe anyone who says so. The problem I see is that audio enthusiasts have taken this to apply to all capacitor applications. Coupling caps in an amplifier are not doing the same thing. They have virtually no current going through them and no voltage drop across them. " Here is very clear advice on where to get bang for your buck on capacitor upgrades. (far from gospel especially for highly sensitive SET designs.) Here on this intellectual island such advice is not as often so plainly stated.
Roger does make some interesting comments on resistors... "I ask this question: If resistors made a big difference, and they do, who is going to make Boutique resistors and who is going to put them in? They are much harder to make than capacitors and require far more expensive equipment and difficult processes. Resistors do have some problems not discussed such as drift with temperature, drift with time and voltage coefficient (change in resistance with voltage causing distortion). In the ESL amps we use a $25 Caddock resistor in the feedback network because it has the lowest voltage coefficient available. There are 4 of these in the amp. They are beautiful with gold plated leads. They are non-inductive. This resistor swings 5000 volts of signal across it and that resistor alone is a major contributor to making the amplifier low distortion. Using a standard resistor would increase the distortion measurably. Of course I could blindly apply this discovery to every resistor in the amps and double the price of this amplifier. I could put them where their exceptional qualities would make no difference. There are some resistors that have no voltage across them and their value can change 100% and not make any difference in how the amp works. There are other resistors that have so little voltage across them that these qualities are vanishingly small. In the OTL-1 the DC considerations are larger than in most amps so I want resistors that won't drift over time so I use the mil-spec Vishays. They cost only ten cents vs 3 cents for common MF resistors. There are about 50 of them. The $25 Caddpcks would make no difference where they are swinging 0-15 volts vs 5,000 volts. They would make a $6,000 difference in the price of this amp. [/b][/b][/b]" And this is a great example of what this thread might have to offer which is where to spend your resistor dollars.
Well that is what makes this thread great. Yes, much like Laputa in Gulliver's Travels, it is a flying intellectual island with an Adamantine base. The Resistor Laputans are fond of mathematics, astronomy, music and technology, but fail to make practical use of their knowledge. I think that is the thread at its heart though most of the individual posters probably get a lot done outside of the thread.
The thread marries together some audiophile discussion with some truely interesting technical discussions. The level of civility is excellent despite these disparate views.
Yet, rock solid conclusions and advise are hard to decipher on Resistor island. Roger Modjeski's cap thread (Favorite Coupling Cap Brands and Types) is a little bit more millitant on engineering and he states that "A crossover is one place where caps can make a difference and I do believe anyone who says so. The problem I see is that audio enthusiasts have taken this to apply to all capacitor applications. Coupling caps in an amplifier are not doing the same thing. They have virtually no current going through them and no voltage drop across them. " Here is very clear advice on where to get bang for your buck on capacitor upgrades. (far from gospel especially for highly sensitive SET designs.) Here on this intellectual island such advice is not as often so plainly stated.
Roger does make some interesting comments on resistors... "I ask this question: If resistors made a big difference, and they do, who is going to make Boutique resistors and who is going to put them in? They are much harder to make than capacitors and require far more expensive equipment and difficult processes. Resistors do have some problems not discussed such as drift with temperature, drift with time and voltage coefficient (change in resistance with voltage causing distortion). In the ESL amps we use a $25 Caddock resistor in the feedback network because it has the lowest voltage coefficient available. There are 4 of these in the amp. They are beautiful with gold plated leads. They are non-inductive. This resistor swings 5000 volts of signal across it and that resistor alone is a major contributor to making the amplifier low distortion. Using a standard resistor would increase the distortion measurably. Of course I could blindly apply this discovery to every resistor in the amps and double the price of this amplifier. I could put them where their exceptional qualities would make no difference. There are some resistors that have no voltage across them and their value can change 100% and not make any difference in how the amp works. There are other resistors that have so little voltage across them that these qualities are vanishingly small. In the OTL-1 the DC considerations are larger than in most amps so I want resistors that won't drift over time so I use the mil-spec Vishays. They cost only ten cents vs 3 cents for common MF resistors. There are about 50 of them. The $25 Caddpcks would make no difference where they are swinging 0-15 volts vs 5,000 volts. They would make a $6,000 difference in the price of this amp. [/b][/b][/b]" And this is a great example of what this thread might have to offer which is where to spend your resistor dollars.
Well here is another post (quote reply I just made seems to have not shown up yet). Very related to the subject of resistors is the volume control and passive preamplifiers (most of which are resistor based). My engineering friend has punted on these Q&A comments from Placette (I just put in a Placette Remote Volume Control and it crushes my previous Daven volume control.) Here is what Placette says:
"Why do passive controls cause such a loss in dynamics?
As a general rule, it's the impedance limitations of the typical high-end volume pot that causes the poor reproduction of dynamics and high frequency loss. But because Vishay® S102 resistors provide a perfectly resistive load with virtually instant rise-time, our passive customers report that the dynamics are actually improved over what they were using before. Of course, the Placette's Active Line-stage also offers a resistive load (with an even higher input resistance) and it has the additional advantage of a very low output impedance to provide the ultimate in perfect transparency and dynamics. Keep in mind that even the slightest unevenness in frequency response or phase distortion will add a false sense of dynamics since the apparent volume of the playback will vary as the pitch of the music varies. These factors also add a muddiness to the sound. The goal of Placette users is to get as close as possible to the actual sound of live acoustic music by eliminating all distortions and colorations from the volume control.
Do all passive controls lack good Bass response?
Most passive controls do not transfer frequency extremes well. But because of the non-inductive load provided by Vishay® S102 resistors, Placette Passive Controls are able to accurately pass along a perfectly linear signal including the deepest bass. Keep in mind that it is very difficult to build a Tube Linestage or Transformer Volume Control without a slight bulge and some muddiness in the bass......."
On my system the increase in bass slam, sound staging and treble extension is amazing (I am not using the unit quite passively as it replaces volume control at the end of my line stage (Bendix 6900)). I've sought out by the book engineering calculations to explain this and have found nothing except that dropping from the 20k Daven to the 10k Placette would increase my treble extension more. I should have slightly less bass as my 3db down point with coupling cap is now a 8 hz. The Placette is hooked up in a way where it is almost internal to my line stage. I suppose one could say that the output impedance of the unit is lower then, but at the same time the 6900 is probably around 1700 ohms driving 10k instead of 20. The 10k load in combination with the 130k load of my amps is flatter in the resistance level presented to the output of the 6900. These engineering explanations just don't seem to account for the huge difference in sound.
I am forced to conclude that this is the sound of the resistors and would guess there is something to this; "Vishay® S102 resistors provide a perfectly resistive load with virtually instant rise-time, our passive customers report that the dynamics are actually improved over what they were using before."... "because of the non-inductive load provided by Vishay® S102 resistors, Placette Passive Controls are able to accurately pass along a perfectly linear signal including the deepest bass." I am hearing this big time and the difference is pretty jaw dropping. (I'll note the Placette as a pure passive with my high voltage line sources is not nearly as good in place of the 6900 and Placette with a loss of dynamics across the board.)
So my questions is why is this? Why do volume controls make so much of a difference? Might there be other resistor locations where these more Physics related material choices are as important.
"Why do passive controls cause such a loss in dynamics?
As a general rule, it's the impedance limitations of the typical high-end volume pot that causes the poor reproduction of dynamics and high frequency loss. But because Vishay® S102 resistors provide a perfectly resistive load with virtually instant rise-time, our passive customers report that the dynamics are actually improved over what they were using before. Of course, the Placette's Active Line-stage also offers a resistive load (with an even higher input resistance) and it has the additional advantage of a very low output impedance to provide the ultimate in perfect transparency and dynamics. Keep in mind that even the slightest unevenness in frequency response or phase distortion will add a false sense of dynamics since the apparent volume of the playback will vary as the pitch of the music varies. These factors also add a muddiness to the sound. The goal of Placette users is to get as close as possible to the actual sound of live acoustic music by eliminating all distortions and colorations from the volume control.
Do all passive controls lack good Bass response?
Most passive controls do not transfer frequency extremes well. But because of the non-inductive load provided by Vishay® S102 resistors, Placette Passive Controls are able to accurately pass along a perfectly linear signal including the deepest bass. Keep in mind that it is very difficult to build a Tube Linestage or Transformer Volume Control without a slight bulge and some muddiness in the bass......."
On my system the increase in bass slam, sound staging and treble extension is amazing (I am not using the unit quite passively as it replaces volume control at the end of my line stage (Bendix 6900)). I've sought out by the book engineering calculations to explain this and have found nothing except that dropping from the 20k Daven to the 10k Placette would increase my treble extension more. I should have slightly less bass as my 3db down point with coupling cap is now a 8 hz. The Placette is hooked up in a way where it is almost internal to my line stage. I suppose one could say that the output impedance of the unit is lower then, but at the same time the 6900 is probably around 1700 ohms driving 10k instead of 20. The 10k load in combination with the 130k load of my amps is flatter in the resistance level presented to the output of the 6900. These engineering explanations just don't seem to account for the huge difference in sound.
I am forced to conclude that this is the sound of the resistors and would guess there is something to this; "Vishay® S102 resistors provide a perfectly resistive load with virtually instant rise-time, our passive customers report that the dynamics are actually improved over what they were using before."... "because of the non-inductive load provided by Vishay® S102 resistors, Placette Passive Controls are able to accurately pass along a perfectly linear signal including the deepest bass." I am hearing this big time and the difference is pretty jaw dropping. (I'll note the Placette as a pure passive with my high voltage line sources is not nearly as good in place of the 6900 and Placette with a loss of dynamics across the board.)
So my questions is why is this? Why do volume controls make so much of a difference? Might there be other resistor locations where these more Physics related material choices are as important.
YepAnd this explains that a pair of resistors will never sound the same if serial or parallel connected for the same resultant value
No it dosnt. Can you tell me how? Explain how any of those effects are actually large enough to hear. There is the odd application ( usually extreme cases) where the resistance change from temperature or voltage may matter but rarely in audio. Can you show me the numbers? If not your speculating, and I'm not buying it. If you can hear the differance between different makes of metal film resistors in a voltage divider, prove it wit a Proper listening test. Your opinion means nothing.
Salesmanship and imagination. Very powerful things that shouldn't be overlooked.
Imagination
No doubt in my mind that their was a huge change. A local audiophile liked to claim the system had an antique sound and now he calls the system massive attack. I was expecting my system to sound sickly with the line stage tube loaded down so much, but i got a deal on the unit and local engineer/audiophile was a big advocate. I went for it, and it was the biggest upgrade I've heard in my system including component changes (have had a few failed upgrades that stand out I must admit).
I suspect that much of this may be an issue with the Daven which almost certainly uses "noninductive" wirewounds. These wirewounds do have some inductance especially in the lower values. At typical volume settings the lower valued resistors would be in series with each other as part of the load (parallel to the load resistors of the amps). Again, I am not sure that this adds up. But based on the salesmanship in the Placette Q&A, it could be that the Daven was kind of a worst case inductance scenario being replaced with the best case Placette.
I am no Engineer, but I do know my math and science. I bow down to the experts here, but it seems that something is going on that is defying conventional thinking.
Well my Phd engineer friend punted (which is pretty much the reaction so far here). The verbiage above I suspect is from the main guy at Placette (its from the Q&A section on their site).
No doubt in my mind that their was a huge change. A local audiophile liked to claim the system had an antique sound and now he calls the system massive attack. I was expecting my system to sound sickly with the line stage tube loaded down so much, but i got a deal on the unit and local engineer/audiophile was a big advocate. I went for it, and it was the biggest upgrade I've heard in my system including component changes (have had a few failed upgrades that stand out I must admit).
I suspect that much of this may be an issue with the Daven which almost certainly uses "noninductive" wirewounds. These wirewounds do have some inductance especially in the lower values. At typical volume settings the lower valued resistors would be in series with each other as part of the load (parallel to the load resistors of the amps). Again, I am not sure that this adds up. But based on the salesmanship in the Placette Q&A, it could be that the Daven was kind of a worst case inductance scenario being replaced with the best case Placette.
I am no Engineer, but I do know my math and science. I bow down to the experts here, but it seems that something is going on that is defying conventional thinking.
This document pretty much covers thin film technology, though not foil or wirewound tech even though they all commonly use nickel-chromium alloys.
I've never been certain whether voltage coefficient really has a significant effect, but it's good to minimize any possible source of non-linearity.
Power coefficient is another possible effect due to the (tiny) voltage generated across dissimilar metals in resistor terminations (end caps), which maybe could be an issue in the discrete resistors commonly used in most electronics. A ferrous endcap over a nickel-chromium thin film as opposed to copper or nickel endcaps might have a greater thermoelectric effect on linearity, rather than a supposed magnetic/inductive effect but this is speculation on my part.
But when you think about it, any electronic devices has >many< metal-to-metal connections, often of dissimilar metals. Some combinations generate more voltage than others, but I suppose it's best to keep everything as close to copper-copper as possible. Any thoughts?
I think bulk metal foil is designed to minimize ALL of these effects and several others, hence its reputation for a very neutral and detailed sound.
That's my theory, and I'm sticking to it. Just wish bulk foil resistors weren't so expensive.
--Damon
This document pretty much covers thin film technology, though not foil or wirewound tech even though they all commonly use nickel-chromium alloys.
I've never been certain whether voltage coefficient really has a significant effect, but it's good to minimize any possible source of non-linearity.
Power coefficient is another possible effect due to the (tiny) voltage generated across dissimilar metals in resistor terminations (end caps), which maybe could be an issue in the discrete resistors commonly used in most electronics. A ferrous endcap over a nickel-chromium thin film as opposed to copper or nickel endcaps might have a greater thermoelectric effect on linearity, rather than a supposed magnetic/inductive effect but this is speculation on my part.
But when you think about it, any electronic devices has >many< metal-to-metal connections, often of dissimilar metals. Some combinations generate more voltage than others, but I suppose it's best to keep everything as close to copper-copper as possible. Any thoughts?
I think bulk metal foil is designed to minimize ALL of these effects and several others, hence its reputation for a very neutral and detailed sound.
That's my theory, and I'm sticking to it. Just wish bulk foil resistors weren't so expensive.
--Damon
Noooo, do-it-yourself
do the A/B test and come back here after, there is an explanation .... 30 years to understand there is a 6dB difference ...
Last edited:
A little while back, I replaced the old carbon volume and balance controls in a friend's high-end tube preamp with a switched attenuator with a ladder of discrete Dale metal film resistors (probably 100 ppm tempco parts, good but not the best possible choice). I thought it sounded clearer and more detailed in the short hour I had to listen to it before returning it to the owner, who was _very_ pleased with the change. I have modified equipment before with better resistors without hearing obvious changes in the sound, but in this case seemed audibly superior.
My theory, for what it's worth, is that carbon composition is inherently non-linear to start with and has a sliding metal to carbon contact that probably contributes several more non-linear effects I'd much rather avoid altogether. I'd replace the controls in my Hafler SE100 preamp (I think Alps carbon potentiometers), if I could find parts that would fit even ungracefully.
At any rate, carbon potentiometers are surely the very worst resistive parts in any piece of audio gear as a source of non-linearity of multiple kinds, especially as they age and get mechanical wear. Conductive film seems to be an improvement, and stepped attenuators with metal-to-metal contacts and far better resistor technology is a great deal better still. The best kind of part might well be voltage-controlled FETs. But I'm sure that'd be a rich source of differing opinion as well.
My theory, for what it's worth, is that carbon composition is inherently non-linear to start with and has a sliding metal to carbon contact that probably contributes several more non-linear effects I'd much rather avoid altogether. I'd replace the controls in my Hafler SE100 preamp (I think Alps carbon potentiometers), if I could find parts that would fit even ungracefully.
At any rate, carbon potentiometers are surely the very worst resistive parts in any piece of audio gear as a source of non-linearity of multiple kinds, especially as they age and get mechanical wear. Conductive film seems to be an improvement, and stepped attenuators with metal-to-metal contacts and far better resistor technology is a great deal better still. The best kind of part might well be voltage-controlled FETs. But I'm sure that'd be a rich source of differing opinion as well.
Last edited:
yes for sure. Did you ever pinch a carbon comp / carbon layer resistor in your fingers while connected to a Digital Ohmmeter ? Wait one minute or two and look at the result...
Last edited:
Yes, that's the temperature coefficient, and says little of the high voltage coefficient. Composition resistors must have a tempco rating several orders of magnitude higher than a good metal film. But I can see a little drift even on bulk foil resistors with my 5 1/2 digit DVM.
I think tube audio aficionados simply like the kind of audio distortion that carbon produces, as they complain about the "sterile" sound of metal film. But of course that's just my theory.
Last edited:
That's one aspect of it, yes. If something makes a measurable change in parameters known to be audible (e.g., level, frequency response...), there's nothing remarkable about hearing differences. What you're describing is unlikely to fall in that category, so unless there's something really pathological in a simple potentiometer to cause those sorts of changes (and the physical stuff being peddled here is pure twaddle), your subjective impression is far more likely to be an artifact of your mind. Assuming that you're a human being, of course, we are ALL susceptible to that. Many industries count on that.
, we are ALL susceptible to that. Many industries count on that.
Well there is quite the cottage industry it seems in selling better volume contols and for quite a bit of money so it seems to be a prevalent condition of the mind at least. A few others have heard my system with the potentiometer change which is not subtle especially when partaking of explosions and gun shots on the typical Bluray action movie. And yet I am sure what I am hearing would not be measurable in frequency response largely (except the change in 3db down point from the switch to a 10k pot from a 20k pot). I agree with you that there is no explanation within the typical EE constructs and so we are left with twaddle it appears (the Physical stuff).so unless there's something really pathological in a simple potentiometer to cause those sorts of changes (and the physical stuff being peddled here is pure twaddle), your subjective impression is far more likely to be an artifact of your mind. Assuming that you're a human being, of course
The high end audio industry thrives on snake oil and twaddle. I am just looking for a more educated way to wade through it. With the advent of the super expensive components over $100,000 and much more it seems the market has become saturated with overpriced devices where the manufacturers are swinging for the fences. As one of lesser means, it is very intriguing to try to weed out the best snake oil and see what can be done in a DIY approach to bring home some of the sound.
Conventional flat earth engineering is very challenged to explain all that we hear in audio. This is not surprising since the music is conveyed by a delicate AC signal which generates a magnetic field. So, I say continue with the Physics twaddle. It appears to be the only functional framework for what our ears hear outside the doctrinaire of EE.
Relay
"43-way stepped attenuator is
based on a shunt design where the signal is fed
through a fixed series resistor and the shunt
resistor, is selected from 43 values ranging from
zero to infinity by means of a selector switch. A
simple ladder design has the disadvantage of
having a number of resistors in the signal path. A
switched ladder switches pairs of resistors, but this
required two sets of switch contacts in the signal
path, which is undesirable and a more complicated
switch is required. Although a shunt design
presents a variable input impedance to the signal
source, this does not matter in practice and the
benefits of a single resistor and switch in the signal
path far outweighs this issue. "
So is some of what I am hearing in the Placette not just the expensive Vishays? Could it be that the Placette has no mechanical switches and uses solid state relays instead. My tube loving heart quavers at the thought of the relays, but I wonder if this may be allowing this volume control to behave like two of these perfect vishay resistors without switches. Its a crack pot idea by any EE perspective, but does it twaddle in the realm of Physics?
As I read this it got me thinking about the mechanical switches in volume controls, etc. I recall Neville Roberts talking about this....
"43-way stepped attenuator is
based on a shunt design where the signal is fed
through a fixed series resistor and the shunt
resistor, is selected from 43 values ranging from
zero to infinity by means of a selector switch. A
simple ladder design has the disadvantage of
having a number of resistors in the signal path. A
switched ladder switches pairs of resistors, but this
required two sets of switch contacts in the signal
path, which is undesirable and a more complicated
switch is required. Although a shunt design
presents a variable input impedance to the signal
source, this does not matter in practice and the
benefits of a single resistor and switch in the signal
path far outweighs this issue. "
So is some of what I am hearing in the Placette not just the expensive Vishays? Could it be that the Placette has no mechanical switches and uses solid state relays instead. My tube loving heart quavers at the thought of the relays, but I wonder if this may be allowing this volume control to behave like two of these perfect vishay resistors without switches. Its a crack pot idea by any EE perspective, but does it twaddle in the realm of Physics?
- Status
- Not open for further replies.