I must learn to read..
You seem to miss the point...If people want to design or build an "Excellent" design already out there (existing)...then find they can hear a difference with components???
This is not about ignoring engineering design..its about do components make a difference in a properly designed amp..Is engineering a cure all and you can use what you want and engineer out the use of cheap components?
Thats actually not correct because cost is not an indicator of a component.
Yes I agree minimise the effect..remove it completley?
The point is can they hear it if its correctly designed..then no they can't..the engineering tests prove it. However if this is the case then two systems built with different parts should always sound the same. SY you always say that the measurements of AUDIOfool parts are worse...however this is interesting because they should have no effect..no change in presentation of the sound.
Regards
M. Gregg
A capacitor (in the signal path) can have two separate duties.
a.) as a pass device where the AC signal across the component is near zero.
b.) as a filter device where the AC signal across the component is very far from near zero.
These two duties are very different.
Design the circuit and select the component to do one, or other duty.
A pass capacitor should NEVER have significant signal voltage across it.
A filter capacitor should always have a significant signal voltage across it.
Select the capacitor appropriately.
a.) as a pass device where the AC signal across the component is near zero.
b.) as a filter device where the AC signal across the component is very far from near zero.
These two duties are very different.
Design the circuit and select the component to do one, or other duty.
A pass capacitor should NEVER have significant signal voltage across it.
A filter capacitor should always have a significant signal voltage across it.
Select the capacitor appropriately.
I think he (post 82) is arguing with himself?
Many designs 'out there' (both DIY and commercial) are not 'excellent' so it is not surprising that components can sometimes change the sound, especially if they are expensive but flawed audiophile components.
Examples:
- poor matching of too low value cathode decoupling electrolytics causing channel phase imbalance at LF
- LF rolloff set too late in the chain (where signal levels are high) by a non-linear dielectric such as polyester
- physically huge polypropylene coupling cap adds lots of stray capacitance so aids instability or interference pickup
- poor grounding injects charging pulses whose amplitude may depend on reservoir ESR so cheaper cap injects less buzz (but some people like a little buzz?)
- 'handmade' cap is loosely wound so microphonic (like microphonic DIY cables some seem to prefer this sound)
So you see there are lots of ways that components can affect the sound, but they all show poor engineering.
Many designs 'out there' (both DIY and commercial) are not 'excellent' so it is not surprising that components can sometimes change the sound, especially if they are expensive but flawed audiophile components.
Examples:
- poor matching of too low value cathode decoupling electrolytics causing channel phase imbalance at LF
- LF rolloff set too late in the chain (where signal levels are high) by a non-linear dielectric such as polyester
- physically huge polypropylene coupling cap adds lots of stray capacitance so aids instability or interference pickup
- poor grounding injects charging pulses whose amplitude may depend on reservoir ESR so cheaper cap injects less buzz (but some people like a little buzz?)
- 'handmade' cap is loosely wound so microphonic (like microphonic DIY cables some seem to prefer this sound)
So you see there are lots of ways that components can affect the sound, but they all show poor engineering.
I think what this whole discussion misses is the reality of what most of us are doing here.
We are experimenters, trying out ideas, seeing how something sounds.
Most people start off with a component such as a DHT and see how best they can use it given the components and understanding they have and with the taste and ancillaries they have. I know for me its been a long slow evolution of a few core concepts, seeing how far I can push them, seeing what refinements can be added. Each stage slightly further on than the next.
Sometimes we just pick up a tube and no matter how suboptimal it might be we set out with the challenge of see if it can be made to work.
Quite frankly all this talk of "poorly designed" amplifiers misses the point, and is quite frankly just plain offensive to those people doing what they want with the things to hand. We can all learn good technique along the way - but its only part of a mix serving a bigger goal for most people.
Shoog
We are experimenters, trying out ideas, seeing how something sounds.
Most people start off with a component such as a DHT and see how best they can use it given the components and understanding they have and with the taste and ancillaries they have. I know for me its been a long slow evolution of a few core concepts, seeing how far I can push them, seeing what refinements can be added. Each stage slightly further on than the next.
Sometimes we just pick up a tube and no matter how suboptimal it might be we set out with the challenge of see if it can be made to work.
Quite frankly all this talk of "poorly designed" amplifiers misses the point, and is quite frankly just plain offensive to those people doing what they want with the things to hand. We can all learn good technique along the way - but its only part of a mix serving a bigger goal for most people.
Shoog
I am still unconvinced by the technical discussion. SY provided a relationship for AC voltage across the capacitor which is nothing other than the passband (LF) response. Although I agree that is good engineering practice to design for minimum attenuation and phase angle at the lowest frequency of interest, and this itself is responsible for the perception of "loose" bass, I do not see how this impacts midband and high frequency performance of the circuit, which is where I hear differences between capacitors. What I perceive are are differences in "presence", "detail", "presentation" (i.e. soundstaging), and deficiencies such as congestion, brightness, veiling, hardness, softness, etc. Now it begs a question about what is the scope of "poor" engineering raised above, i.e., extending it to components (such as capacitors). Now we are recognizing that these do contribute to the amplifier performance, and not all capacitors are equal, i.e., there are aspects of their construction that impact sound (use of non-linear dielectrics and winding tension being two examples cited above). This gets to my point that when dealing with real-world components, there are differences between them (i.e., none are perfect, ideal components), and these can influnce the sound. Finally, I would like to take issue with the concept that audiophile components are necessarily "flawed". Industrial components from the major manufacturers are engineered for mass production using processes that can be easily accomplished via automation, with low component variation (to the extend it is affordable to control) to ensure a consistant product meeting established performance specifications. However, they only manufacture parts that are commercially viable. There are experienced specialty manufacturers of capacitors for high performance applications (military, aerospace, etc.) where the components are subject to more stringent performance requirements. These are manufactured in limited quantities, or on a custom order basis. You will find that many of the "audiophile" capacitors are manufactured by these outfits, not the company whose name is on the label. I am not certain that you can claim that these products are neccessarily "flawed" (you need to have some broad support for that accusation, instead of cherry-picked examples), however they are expensive due to economy of scale, the materials used in production, and the more labor intensive manufacturing processes, and the marketing/distribution chain. And there are is wide body of people involved on both the amateur and commercial side who are convinced that these can and do make a difference that is worth paying the extra price. Of course, I am in complete agreement that it is far better to engineer an amplifier that minimizes use of these parts, and they should not be used to "tailor" the sound (like add a capacitor when not necessary just because you like the coloration it produces). I think you need to have a good, well engineered design first, and then experiment with more exotic components where appropriate to see if these can provide a more nuanced and revealing musical performance.
Just my thoughts,
I remember someone saying to me I need a circuit to do ...whatever...I said its easy got a piece of paper and drew a circuit..here this will do it...OHH no said the guy I don't feel competent to just take a guess...
..it will work ..look so I lashed it togeather on the bench (it worked)..Ohhh no I will wait until I can ask the head tech...so he did, he did the same circuit..LMAO
You know..if you build something and it sounds great..and if you change the caps and it sounds better...who cares...
If it beats the c..p in the shops, and it moves you how much is it worth to you? would you scrap it and go and buy an all in one system..
I know one thing Ive been doing DIY audio since I was about 10 years old and each thing I buit was worth its weight in gold to me..
Ive listened to so called engineers amps and was less than impressed with many..
How can that be? Going to many conventions..listening to so many systems that I lost count..I don't believe they were all bad designs..Not many had a profound effect on me...I was not happy with systems you could buy when I was young..and could not afford megga systems..and some were less than good..If you can change the sound of a quad /Naim/etc with caps then if your circuit changes thats good enough for me..
Regards
M. Gregg
I remember someone saying to me I need a circuit to do ...whatever...I said its easy got a piece of paper and drew a circuit..here this will do it...OHH no said the guy I don't feel competent to just take a guess...
..it will work ..look so I lashed it togeather on the bench (it worked)..Ohhh no I will wait until I can ask the head tech...so he did, he did the same circuit..LMAOYou know..if you build something and it sounds great..and if you change the caps and it sounds better...who cares...
If it beats the c..p in the shops, and it moves you how much is it worth to you? would you scrap it and go and buy an all in one system..
I know one thing Ive been doing DIY audio since I was about 10 years old and each thing I buit was worth its weight in gold to me..
Ive listened to so called engineers amps and was less than impressed with many..
How can that be? Going to many conventions..listening to so many systems that I lost count..I don't believe they were all bad designs..Not many had a profound effect on me...I was not happy with systems you could buy when I was young..and could not afford megga systems..and some were less than good..If you can change the sound of a quad /Naim/etc with caps then if your circuit changes thats good enough for me..
Regards
M. Gregg
Fine. Just don't pretend it is audio design based on circuit theory. You might not have claimed this, but people often do.Shoog said:
Sadly, some never learn good technique but continue to propagate misunderstandings or celebrate lack of understanding. If good technique is only part of the mix, what is the rest of the mix: poor technique?
In what ways would the voltage waveform presented at the speaker have to change in order to provide these subjective characteristics? (A few I can guess; others I have no idea).dmcgown said:
How do you know when things are 'more nuanced' i.e. that what you are hearing is formerly hidden detail rather than artificially added detail? As you couldn't hear it before, how do you know it was there?
When it provides a better sense that there are real musicians/instruments playing/singing within a 3 dimensional acoustical environment. This presumes you have a high resolution source (vinyl, for instance) capable of capturing and resolving the low level information, and know what real music sounds like.
Well I know what real music sounds like, although I don't get to many orchestral concerts these days. I'm sure there is a debate to be had about what constitutes a 'high resolution source' and whether vinyl qualifies; I don't have a strong view either way on that.
One thing I have noticed is that live acoustic music sometimes includes sounds which if heard via an amplifier might wrongly be regarded as distortion. Smoothing these away makes the sound more pleasant but less lifelike, hence the futility of tuning for 'best' sound rather than designing for accuracy.
One thing I have noticed is that live acoustic music sometimes includes sounds which if heard via an amplifier might wrongly be regarded as distortion. Smoothing these away makes the sound more pleasant but less lifelike, hence the futility of tuning for 'best' sound rather than designing for accuracy.
You are right, of course: you most often are! Most tubes have Gm ~ 1,000 to 5,000 µS ... which 1/Gm becomes 200 to 1000 ohms. And since most cathode resistors are on that same order-of-range, then the oft' quoted formula needs some a-fixing:
1,000,000 / (6.28 x F x ((Rc/Gm)/(Rc+1/Gm))
Where one then chooses "F" to be quite a bit lower (1/5th to 1/10th) than the lowest frequency that's audible, to minimize the phase shifting and chained-stage rolloff. 3 Hz.
Let's try Rc = 470, Gm = 3,000 µS, F = 3 Hz. Z complex = 195 ohm, so that comes out to 270 µF. JUST as you predicted. I'll remember this in the future, thanks.
GoatGuy
Actually, I suspect even my correction is wrong. There ought also to be a contribution to R from the mu-transformed load impedance, so you end up with something like R = Rk || ( 1/gm + RL/mu ). The last term can be ignored for a pentode. The mu might actually be mu+1 but for most valves it doesn't matter as mu is not known all that accurately anyway.
The wrong simple version with just Rk is regularly quoted on the web and in magazine articles so it is a persistent meme.
The wrong simple version with just Rk is regularly quoted on the web and in magazine articles so it is a persistent meme.
again... I think most of this becomes pretty academic though when all the parasitics are considered. It isn't clear to me that there is 1/mu coupling of the RL to the Rc (Rk, I guess is more standard, those darn Germans with their kathodes). I can believe the 1/Gm component though couples directly.
Wouldn't it be fair to say, "except for extraordinary combinations of tubes and components outside the 'nominal range', using 1 Hz as the -3dB design point, and ignoring the 1/Gm issue, largely specifies a satisfactory cathode resistor bypass capacitor."
C = 1,000,000 / (6.28 x R) = (160,000/R) in µF
The same example as above (470 ohm) delivers a 330 µF recommendation for Ck ... seems plenty big. Parasitics and all.
GoatGuy
Wouldn't it be fair to say, "except for extraordinary combinations of tubes and components outside the 'nominal range', using 1 Hz as the -3dB design point, and ignoring the 1/Gm issue, largely specifies a satisfactory cathode resistor bypass capacitor."
C = 1,000,000 / (6.28 x R) = (160,000/R) in µF
The same example as above (470 ohm) delivers a 330 µF recommendation for Ck ... seems plenty big. Parasitics and all.
GoatGuy
I remember in the old valve radios and TVs, had a horrible capacitors hated by all repairers, which burst and lost liquid.
Today, in a specialized magazine, it says: wow ! paper in oil capacitors "open the low end".
As my pocket does not allow me to buy expensive things, and my common sense sometimes works, I made a model and ran the simulation, surprise !, is a low-pass filter, does not open the bass, removes the treble !
As far as I could see, only Russian teflon capacitors, are careful shorting the windings at the ends to minimize inductance.
Ha, ha, Adio Grade Capacitors...
Maybe Gold foil-Snake oil capacitors must be better ?
OK, porosity of metal deposition in ordinary MKPs must not be very good, but at least I, don't hear it.
In view of the results that come up in this thread, now I know why.
Today, in a specialized magazine, it says: wow ! paper in oil capacitors "open the low end".
As my pocket does not allow me to buy expensive things, and my common sense sometimes works, I made a model and ran the simulation, surprise !, is a low-pass filter, does not open the bass, removes the treble !
As far as I could see, only Russian teflon capacitors, are careful shorting the windings at the ends to minimize inductance.
Ha, ha, Adio Grade Capacitors...
Maybe Gold foil-Snake oil capacitors must be better ?
OK, porosity of metal deposition in ordinary MKPs must not be very good, but at least I, don't hear it.
In view of the results that come up in this thread, now I know why.
Metal encapsulated, glass sealed paper in oil capacitors do not have a problem with drying out or leakage, there are plenty of vintage Vitamin-Q, West-Caps, and the like doing good service several decades after they were manufactured. Generally it is poor quality electrolytics and badly sealed paper caps that are the problem.
Steve Bench a number of years ago performed some interesting measurements of different capacitor types to see if there were any measureable basis for the sound of different capacitors. Basically, he was checking for linearity. Paper in oil caps did quite well in comparison with modern polymer dielectrics.
The "Sound" of Capacitors
I understand we are all constrained in our budgets, and exotica such as Dueland capacitors (just an extreme example), are beyond what we can justify in our builds. To tell you the truth, I cannot personally justify spending that kind of money on capacitors, even though I know from personal experience how good they can sound (or not sound, which is more to the point). Ridicule and distain for "audiophile" parts and those who use them seem to me to be a bit of a case of "capacitor envy"
Steve Bench a number of years ago performed some interesting measurements of different capacitor types to see if there were any measureable basis for the sound of different capacitors. Basically, he was checking for linearity. Paper in oil caps did quite well in comparison with modern polymer dielectrics.
The "Sound" of Capacitors
I understand we are all constrained in our budgets, and exotica such as Dueland capacitors (just an extreme example), are beyond what we can justify in our builds. To tell you the truth, I cannot personally justify spending that kind of money on capacitors, even though I know from personal experience how good they can sound (or not sound, which is more to the point). Ridicule and distain for "audiophile" parts and those who use them seem to me to be a bit of a case of "capacitor envy"
Last edited:
What Steve Bench calls hysteresis could just be harmless phase shift caused by series resistance or dielectric absorption; a much closer look at the curves would be needed to tell. The tantalums look bad (definitely non-linear), but the rest look OK - nothing surprising. It is quite easy to produce a graph which looks alarming to non-technical people; advertisers do it all the time. An oscilloscope is not a good tool for investigating all but the worst caps, such as high-K ceramics.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.