My cables are only 2.5 meters long.
You might try to post a wanted ad, as many people took part in the Mills group buy and I don't think all of them are using the resistors
, especially when the fame went on that Aleph X sounds best with higher PS voltage
Those were Mills wirewounds, 0.22ohm, 5W, perfect choice for a GC. I bought 100 of them, but I have a feeling I will use all of them
You might try to post a wanted ad, as many people took part in the Mills group buy and I don't think all of them are using the resistors
, especially when the fame went on that Aleph X sounds best with higher PS voltageThose were Mills wirewounds, 0.22ohm, 5W, perfect choice for a GC. I bought 100 of them, but I have a feeling I will use all of them
Will use 100 resistor!!!What a production!
Just to aproach the issue: how much do I expect to pay for a resistor similar to those?
cheers
Ric
Peter Daniel said:Those were Mills wirewounds, 0.22ohm, 5W, perfect choice for a GC. I bought 100 of them, but I have a feeling I will use all of them
Just to aproach the issue: how much do I expect to pay for a resistor similar to those?
cheers
Ric
Dear Peter:
>You certainly agree, that they (Rikens, Caddocks or generic type) provide different sound in some locations.<
Rikens and Caddocks don't even measure alike. But it does seem that there are sonic differences between nominally similar-structure resistors - Vishays vs. Alpha being one such case.
OTOH, the properties desired of a resistor for one location may be rather different from those in another location. Besides, some locations may simply not be very sensitive.
>I'm not trying to promote one brand over the other, but simply indicate, that experimentation is worth the trouble.<
Resistor experimentation is certainly easy enough to do. But it can be pricey, as you have undoubtedly noticed. And if you decide on a resistor that happens to be fairly expensive, you will pay a significant price penalty for each and every amplifier that you build,
>Since there is not much opportunity to improve the schematic design.<
If the amplification mode is non-inverting, you could bootstrap the entire power opamp. The bootstrapping could either track the output, or it could be compensated back to the input, similar to the methods outlined in Analog Device's AN232 by Walter Jung. Other interesting ideas would be to use digital power amps as the bootstrapping devices. Run a virtual ground. Maybe regulate everything - which could be linear or switching. Or perhaps low-loss synchronous rectification. No need to be shy
>and PCB is not required here<
PCB or no PCB, some type of physical structure will be required, and here it definitely pays to think, calculate and experiment. I am currently designing a partially PCB, partially P2P structure for one of my products, and I am up to revision #24. Based on my experience over the years, experimenting with the physical structures (including pcb) has much more potential to benefit the sound than swapping resistors (assuming that each location already has a resistor that has suitable properties for how it will be used). Experimentation with board and structure design requires time, effort, and money, but once the design has been finalized, you don't have to pay a price penalty on each and every amplifier that you produce. Unlike premium-grade resistors.
Mind you, I am not against using high-quality components. But what I see is that many DIY'ers jump to expensive components without doing the basics - which is to put as much thought, study and care as possible into their schematics and structures. As I have said before, high-quality components should be regarded as the icing on the cake of thorough and careful design - they should not be considered a substitute for incomplete or inept design.
>parts become comparatively higher priority.<
If you have exhausted all of your other possibilities, maybe.
hth, jonathan carr
>You certainly agree, that they (Rikens, Caddocks or generic type) provide different sound in some locations.<
Rikens and Caddocks don't even measure alike. But it does seem that there are sonic differences between nominally similar-structure resistors - Vishays vs. Alpha being one such case.
OTOH, the properties desired of a resistor for one location may be rather different from those in another location. Besides, some locations may simply not be very sensitive.
>I'm not trying to promote one brand over the other, but simply indicate, that experimentation is worth the trouble.<
Resistor experimentation is certainly easy enough to do. But it can be pricey, as you have undoubtedly noticed. And if you decide on a resistor that happens to be fairly expensive, you will pay a significant price penalty for each and every amplifier that you build,
>Since there is not much opportunity to improve the schematic design.<
If the amplification mode is non-inverting, you could bootstrap the entire power opamp. The bootstrapping could either track the output, or it could be compensated back to the input, similar to the methods outlined in Analog Device's AN232 by Walter Jung. Other interesting ideas would be to use digital power amps as the bootstrapping devices. Run a virtual ground. Maybe regulate everything - which could be linear or switching. Or perhaps low-loss synchronous rectification. No need to be shy
>and PCB is not required here<
PCB or no PCB, some type of physical structure will be required, and here it definitely pays to think, calculate and experiment. I am currently designing a partially PCB, partially P2P structure for one of my products, and I am up to revision #24. Based on my experience over the years, experimenting with the physical structures (including pcb) has much more potential to benefit the sound than swapping resistors (assuming that each location already has a resistor that has suitable properties for how it will be used). Experimentation with board and structure design requires time, effort, and money, but once the design has been finalized, you don't have to pay a price penalty on each and every amplifier that you produce. Unlike premium-grade resistors.
Mind you, I am not against using high-quality components. But what I see is that many DIY'ers jump to expensive components without doing the basics - which is to put as much thought, study and care as possible into their schematics and structures. As I have said before, high-quality components should be regarded as the icing on the cake of thorough and careful design - they should not be considered a substitute for incomplete or inept design.
>parts become comparatively higher priority.<
If you have exhausted all of your other possibilities, maybe.
hth, jonathan carr
Dear Jonathan
Do you have any views w.r.t. the ferrous / non-ferrous state of components / leads.
I've used for years the basic Beyschlag MRS25 metal films - they are cheap, low noise and sound acceptable.
The Vishay / Dale RN60's I've been kindly sent by a partner are a lot better sounding though, and this is in a PSU regulator application, not a 'direct' signal path app.
There are a number of obvious differences: -
a) The RN60's ar more conservatively rated w.r.t. Pd ratings.
b) They are physically larger.
c) They are of a non-magnetic construction.
I'm utterly convinced of the sonic benefits, but cannot easily pin the effect to a single cause - I've found copper-leaded caps to sound better too, but how much of this is due to the fact that these devices are aimed at audio, generally, and have other characterisitics that are better?
The Welwyn RC55's are an interesting example I should try - these have steel endcaps but are reported to sound good.
I for one would be vry grateful for your input or ideas, primarily because the options open to me at the moment are all expensive, and I'm a cheapskate
Finally I agree that these are 'icing on the cake' and that other issues are often more important. Most would be better placing their efforts elsewhere for big improvements.
That icing tastes nice to me at present
Regards,
Andy.
Do you have any views w.r.t. the ferrous / non-ferrous state of components / leads.
I've used for years the basic Beyschlag MRS25 metal films - they are cheap, low noise and sound acceptable.
The Vishay / Dale RN60's I've been kindly sent by a partner are a lot better sounding though, and this is in a PSU regulator application, not a 'direct' signal path app.
There are a number of obvious differences: -
a) The RN60's ar more conservatively rated w.r.t. Pd ratings.
b) They are physically larger.
c) They are of a non-magnetic construction.
I'm utterly convinced of the sonic benefits, but cannot easily pin the effect to a single cause - I've found copper-leaded caps to sound better too, but how much of this is due to the fact that these devices are aimed at audio, generally, and have other characterisitics that are better?
The Welwyn RC55's are an interesting example I should try - these have steel endcaps but are reported to sound good.
I for one would be vry grateful for your input or ideas, primarily because the options open to me at the moment are all expensive, and I'm a cheapskate
Finally I agree that these are 'icing on the cake' and that other issues are often more important. Most would be better placing their efforts elsewhere for big improvements.
That icing tastes nice to me at present
Regards,
Andy.
Jonathan,
I was specifically talking about GC amp and the parts selection for inverting mode of operation. Simplicity was a key to the expected performance and I was not really tempted to complicate matters by using bootstraping or regulated supplies. Although I didn't try regulators, my feeling is they would not bring any improvement.
This circuit is very sensitive to any changes and because it uses only 2 extra parts (beside IC), any differences caused by the resistors are immeditely obvious.
Although the circuit is very simple, I spend countless hours on experimenting and tweaking until the performance became satisfactory. For instance, I noticed that paralleling PS caps is not good and single cap sounds better. Adding bypass caps (so popular with both DIY and established companies) is not good either, as although it may seem nicer sounding in the beginning, takes away natural integrity from the sound.
I was initially quite satisfied with Riken in feedback loop (which was already improvement over Holco), but initial response from people testing amps, was lack of depth in the soundstage. I bought Caddock resistors long time ago, but didn't really have desire to try them out (if it works, why bother with different parts, was my partner's suggestion). Working on next project, the monoblocks, I put the Caddocks in a feedback location and to my suprise, they sounded just right, more liquid and dimentional than Rikens. Swapping Kimber umbilical cord to Cardas wire improved things even further.
So as you see, I'm not that crazy about exotic components (I'm not even taking plastic jackets off the caps) and using them as a cure for everything. But, if I see that a certain part makes a big difference, I just can't dismiss it. As I said earlier, my amp is using only 2 resistors. Comparing that to an average circuit, containing approx. 20 resistors, I think I can allow myself the luxury to pay $6 for my 2 pieces
We finally managed to get a review of the amp http://www.6moons.com/audioreviews/audiooasis/audiozone.html
Maybe you could tell me, is there anything here, that can be further improved structurally?
I'm a firm believer that structure design of any audio unit has big influence on final performance, sometimes even bigger than the resistors, indeed
After all, my actual background is in aircraft structures, not electronics, so it should be understandable that I prefer to play with parts and not schematics.
I was specifically talking about GC amp and the parts selection for inverting mode of operation. Simplicity was a key to the expected performance and I was not really tempted to complicate matters by using bootstraping or regulated supplies. Although I didn't try regulators, my feeling is they would not bring any improvement.
This circuit is very sensitive to any changes and because it uses only 2 extra parts (beside IC), any differences caused by the resistors are immeditely obvious.
Although the circuit is very simple, I spend countless hours on experimenting and tweaking until the performance became satisfactory. For instance, I noticed that paralleling PS caps is not good and single cap sounds better. Adding bypass caps (so popular with both DIY and established companies) is not good either, as although it may seem nicer sounding in the beginning, takes away natural integrity from the sound.
I was initially quite satisfied with Riken in feedback loop (which was already improvement over Holco), but initial response from people testing amps, was lack of depth in the soundstage. I bought Caddock resistors long time ago, but didn't really have desire to try them out (if it works, why bother with different parts, was my partner's suggestion). Working on next project, the monoblocks, I put the Caddocks in a feedback location and to my suprise, they sounded just right, more liquid and dimentional than Rikens. Swapping Kimber umbilical cord to Cardas wire improved things even further.
So as you see, I'm not that crazy about exotic components (I'm not even taking plastic jackets off the caps) and using them as a cure for everything. But, if I see that a certain part makes a big difference, I just can't dismiss it. As I said earlier, my amp is using only 2 resistors. Comparing that to an average circuit, containing approx. 20 resistors, I think I can allow myself the luxury to pay $6 for my 2 pieces
We finally managed to get a review of the amp http://www.6moons.com/audioreviews/audiooasis/audiozone.html
jcarr said:
If you have exhausted all of your other possibilities, maybe.
Maybe you could tell me, is there anything here, that can be further improved structurally?
I'm a firm believer that structure design of any audio unit has big influence on final performance, sometimes even bigger than the resistors, indeed
After all, my actual background is in aircraft structures, not electronics, so it should be understandable that I prefer to play with parts and not schematics.
Attachments
Peter: If your approach works for you, and if you are happy and confident with the results, there is no need for you to be concerned with my opinion, or anyone else's. It's your design, right?
>Maybe you could tell me, is there anything here, that can be further improved structurally?<
With your background, I am sure that you are just as capable as I am at designing physical structures - possibly more so. FWIW, I do a fair amount of pcb layout and P2P design work using a mechanical CAD program (VectorWorks) in addition to a PCB program (Protel 98).
>it should be understandable that I prefer to play with parts and not schematics.<
Ah, but studying new things and gaining new abilities is always a joy. If you get into it, electrical schematics and structures are just as much fun as mechanical designs and structures.
regards, jonathan carr
>Maybe you could tell me, is there anything here, that can be further improved structurally?<
With your background, I am sure that you are just as capable as I am at designing physical structures - possibly more so. FWIW, I do a fair amount of pcb layout and P2P design work using a mechanical CAD program (VectorWorks) in addition to a PCB program (Protel 98).
>it should be understandable that I prefer to play with parts and not schematics.<
Ah, but studying new things and gaining new abilities is always a joy. If you get into it, electrical schematics and structures are just as much fun as mechanical designs and structures.
regards, jonathan carr
Dear Andy:
>Do you have any views w.r.t. the ferrous / non-ferrous state of components / leads.<
It appears that to find specific numbers on resistors, I will have to dig very deep. OTOH, I did locate some notes on the distortion added by speaker terminals equipped with gold-plated steel O-lugs, as opposed to speaker terminals that had no such ferromagnetic subcomponent.
The base measurements were conducted at a level of 10V rms into an 8-ohm load, and the power amp that served as a test bed had a residual distortion of 0.00014% under these conditions.
With the gold-plated steel lugs at the power amp output, the distortion rose to 0.0018%, and the added distortion products appeared to consist primarily of 3rd-order harmonics, but higher-order distortion products could also be observed. When the load impedance was reduced to 4 ohms, the distortion increased to 0.0098%.
OTOH, we should not forget that Nichrome - the resistive material used for most metal-film resistors (including Vishays) - is a ferromagnetic material.
FWIW, jonathan carr
>Do you have any views w.r.t. the ferrous / non-ferrous state of components / leads.<
It appears that to find specific numbers on resistors, I will have to dig very deep. OTOH, I did locate some notes on the distortion added by speaker terminals equipped with gold-plated steel O-lugs, as opposed to speaker terminals that had no such ferromagnetic subcomponent.
The base measurements were conducted at a level of 10V rms into an 8-ohm load, and the power amp that served as a test bed had a residual distortion of 0.00014% under these conditions.
With the gold-plated steel lugs at the power amp output, the distortion rose to 0.0018%, and the added distortion products appeared to consist primarily of 3rd-order harmonics, but higher-order distortion products could also be observed. When the load impedance was reduced to 4 ohms, the distortion increased to 0.0098%.
OTOH, we should not forget that Nichrome - the resistive material used for most metal-film resistors (including Vishays) - is a ferromagnetic material.
FWIW, jonathan carr
Hi,
Jonathan,
While this is certainly true, I somehow doubt you would encourage people to use resistors with ferrous endcaps and leadout wires, would you?
If we'd agree (hypothetically) that ferrous content in components, cable, etc. is best avoided then I can start to understand why some people have a preference for oldfashioned carbon composition resistors...even if it means sacrificing a little on noise and reliability in the process.
This seems to be fashionable amongst SE/Tube lovers lately and in the context of digital harshness I can certainly understand why, even if it deviates from neutral.
Are there alternative components you can recommend?
Cheers,
Jonathan,
While this is certainly true, I somehow doubt you would encourage people to use resistors with ferrous endcaps and leadout wires, would you?
If we'd agree (hypothetically) that ferrous content in components, cable, etc. is best avoided then I can start to understand why some people have a preference for oldfashioned carbon composition resistors...even if it means sacrificing a little on noise and reliability in the process.
This seems to be fashionable amongst SE/Tube lovers lately and in the context of digital harshness I can certainly understand why, even if it deviates from neutral.
Are there alternative components you can recommend?
Cheers,
Jonathan,
Thanks once again for a very enlightening and thought-provoking reply.
I feel I should apply a little more science to this process myself, to see if I can actually resolve in measurement the things I hear with my ears - I'm not sure if my limited test apparatus is up to it though...
Cheers,
Andy.
Thanks once again for a very enlightening and thought-provoking reply.
I feel I should apply a little more science to this process myself, to see if I can actually resolve in measurement the things I hear with my ears - I'm not sure if my limited test apparatus is up to it though...
Cheers,
Andy.
Dear Frank:
>I somehow doubt you would encourage people to use resistors with ferrous endcaps and leadout wires, would you?<
Depends on whether resistors with non-ferrous endcaps and leads are readily available in the values and sizes required. It also depends on other important parameters like value accuracy, noise, inductance, hygroscopics, tempco, and so on. Even when I observe something concrete, like the distortion imparted by ferrous construction, I try to keep everything in perspective, and maintain some idea as to the relative ranking of the various evils. And if this requires me to sometimes act as my own Devil's Advocate, so be it.
BTW, I should note that the DUT (gold-plated steel lug) was added at the power amplifier output, and so was not included in any sort of NFB loop. If it were placed inside a global NFB loop, I would expect the measured distortion to drop substantially. OTOH, normally, 3rd-order distortion products tend to increase by the square of the voltage or current applied (current in this case, going from an 8-ohm load to a 4-ohm load), But what was observed was an increase in distortion from 0.0018% to 0.0098%, which exceeded the expected four-fold increase.
>If we'd agree that ferrous content in components, cable, etc. is best avoided then I can start to understand why some people have a preference for oldfashioned carbon composition resistors.<
I will add that carbon comp resistors usually have no spiral trimming, with relatively low inductance per unit length as a result. Of course, the lack of trimming is part of the reason why their values are all over the map,
>even if it means sacrificing a little on noise and reliability in the process.<
And the fact that carbon comp resistors are usually rather hygroscopic, meaning that they can change value depending on a combination of the atmospheric humidity and operating temperature.
>This seems to be fashionable amongst SE/Tube lovers lately and in the context of digital harshness I can certainly understand why, even if it deviates from neutral.<
Based on how a lot of solid-state gear and also some tube amplification seems to sound, I can also understand why some audiophiles would feel this way, even though I personally don't share their views.
>Are there alternative components you can recommend?<
Admittedly there are certain components that I dislike less than most, but that is about as far as my enthusiasm is willing to extend.
hth, jonathan carr
>I somehow doubt you would encourage people to use resistors with ferrous endcaps and leadout wires, would you?<
Depends on whether resistors with non-ferrous endcaps and leads are readily available in the values and sizes required. It also depends on other important parameters like value accuracy, noise, inductance, hygroscopics, tempco, and so on. Even when I observe something concrete, like the distortion imparted by ferrous construction, I try to keep everything in perspective, and maintain some idea as to the relative ranking of the various evils. And if this requires me to sometimes act as my own Devil's Advocate, so be it.
BTW, I should note that the DUT (gold-plated steel lug) was added at the power amplifier output, and so was not included in any sort of NFB loop. If it were placed inside a global NFB loop, I would expect the measured distortion to drop substantially. OTOH, normally, 3rd-order distortion products tend to increase by the square of the voltage or current applied (current in this case, going from an 8-ohm load to a 4-ohm load), But what was observed was an increase in distortion from 0.0018% to 0.0098%, which exceeded the expected four-fold increase.
>If we'd agree that ferrous content in components, cable, etc. is best avoided then I can start to understand why some people have a preference for oldfashioned carbon composition resistors.<
I will add that carbon comp resistors usually have no spiral trimming, with relatively low inductance per unit length as a result. Of course, the lack of trimming is part of the reason why their values are all over the map,
>even if it means sacrificing a little on noise and reliability in the process.<
And the fact that carbon comp resistors are usually rather hygroscopic, meaning that they can change value depending on a combination of the atmospheric humidity and operating temperature.
>This seems to be fashionable amongst SE/Tube lovers lately and in the context of digital harshness I can certainly understand why, even if it deviates from neutral.<
Based on how a lot of solid-state gear and also some tube amplification seems to sound, I can also understand why some audiophiles would feel this way, even though I personally don't share their views.
>Are there alternative components you can recommend?<
Admittedly there are certain components that I dislike less than most, but that is about as far as my enthusiasm is willing to extend.
hth, jonathan carr
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.