Hi,
You are right of course. I'm supposed to know that (rootsumsquarerule) but keep forgetting. I must be getting old or just need a break or both.
As it stands ever which way, the capacitor in the circuit determines the noise (and mostly linearity) though.
Ciao T
You are right of course. I'm supposed to know that (rootsumsquarerule) but keep forgetting. I must be getting old or just need a break or both.
As it stands ever which way, the capacitor in the circuit determines the noise (and mostly linearity) though.
Ciao T
OK, now for some sorting out. First, it is very important to have a LOW NOISE power supply voltage for Vendetta like circuits. This is because there is little if any POWER SUPPLY REJECTION in this circuit. It is one of its drawbacks, so you have to think of the final regulator as part of the circuit, itself. It is that important.
We could use either passive or active regulators, BUT we can't just pick out an IC regulator and use it directly. It is just too noisy! This is mostly because the REFERENCE inside the chip is NOT bypassed with a large cap. It is interesting that the IC voltage regulators designed by National Semi in the '60's could be QUIETER than most IC regulators today. They actually allowed effectively bypassing the internal reference with a relatively large cap. I developed the Vendetta circuits using these early regulators. Since, this approach is impractical, today, another approach is to use a relatively noisy, but stable IC regulator, then a large passive filter, or a very low noise follower, that can be either a cap multiplier, or a synthetic zener reference acting like an equivalent to a zener in practice, but its components and action are different. This is: a current selected, low noise, low Gm jfet (yes, they exist) driving a resistance that is bypassed with a large cap. This gives regulation to DC, good input isolation, and essentially low noise. This is called a 'Norton Equivalent' of a zener circuit.
If you look at the Vendetta power supply linked here, you will see an example.
We could use either passive or active regulators, BUT we can't just pick out an IC regulator and use it directly. It is just too noisy! This is mostly because the REFERENCE inside the chip is NOT bypassed with a large cap. It is interesting that the IC voltage regulators designed by National Semi in the '60's could be QUIETER than most IC regulators today. They actually allowed effectively bypassing the internal reference with a relatively large cap. I developed the Vendetta circuits using these early regulators. Since, this approach is impractical, today, another approach is to use a relatively noisy, but stable IC regulator, then a large passive filter, or a very low noise follower, that can be either a cap multiplier, or a synthetic zener reference acting like an equivalent to a zener in practice, but its components and action are different. This is: a current selected, low noise, low Gm jfet (yes, they exist) driving a resistance that is bypassed with a large cap. This gives regulation to DC, good input isolation, and essentially low noise. This is called a 'Norton Equivalent' of a zener circuit.
If you look at the Vendetta power supply linked here, you will see an example.
That is what I use when the voltages are high enough for not using a trillion LEDs or when I want it variable with a trimmer, only problem is drift since the Norton will directly translate I to V. Using the proper goss Jfet at its cross point Tc id/Vgs curves (usually Vgs(0Tc) @ Vp-0.65V) with some Rs, goes well enough though. Cascode Norton Vref is even better.
Actually capacitors create noise! There is a base level of vibration around .005 G (gravity) or more just about everywhere. When a capacitor has a dc voltage applied any movement between the plates will create a noise voltage. The stiffer the construction of the capacitor the lower the noise. The types that are rolled round and then squished flat have the lowest vibration pickup when mounted with the narrowest dimension subjected to the source of the vibration. Fluid filled capacitors such as electrolytics do show more internal damping.
So any design of a low noise power supply must take into account all sources of noise.
Noooooooo regulators at all. LOL
No academic talk here!
BS, I am using a totally passive RC filtered supply.......
No academic talk here!
Last edited:
Simon
"The types that are rolled round and then squished flat have the lowest vibration pickup when mounted with the narrowest dimension subjected to the source of the vibration"
eeh where did that come from...??? that statement is plain wrong..as you can't retain tension in flat foils...only round shapes can do that..and there the king of mechanical stability is the polypropylene caps as they are rolled from stretched foil.. that then is baked so the foil crimps and forms a mechanically stable unit..
But for sure here we talk about El-caps for the R-C arrays where versions like the Elna Silk.. has good mechanical properties and should also sound good...
"The types that are rolled round and then squished flat have the lowest vibration pickup when mounted with the narrowest dimension subjected to the source of the vibration"
eeh where did that come from...??? that statement is plain wrong..as you can't retain tension in flat foils...only round shapes can do that..and there the king of mechanical stability is the polypropylene caps as they are rolled from stretched foil.. that then is baked so the foil crimps and forms a mechanically stable unit..
But for sure here we talk about El-caps for the R-C arrays where versions like the Elna Silk.. has good mechanical properties and should also sound good...
Last edited:
Yeah, gravity is a b/i/t/ch. I had to reorient all my caps when the wife moved the amp to the other cupboard.
jan
That came from measurements. My article with some of the curves ran in the US and other editions of Elektor.
Actually the best for vibration resistance is tin foil with polystyrene, heat treated and then cyro treated. But I should charge a consulting fee to tell you that.
But one ting is evident...mechanics play a vital role in electronics... especially stuff that deals with higher power where forces are substantial...
Tinfoil is all I use for my crossovers...albeit not with polystyrene but more common types with stretched and baked polypropylene..Think its the weight of the tinfoil that aids in the calm and clear treble they are capable off....I know the Mr Gerhard aus Deutschland has similar findings....
Tinfoil is all I use for my crossovers...albeit not with polystyrene but more common types with stretched and baked polypropylene..Think its the weight of the tinfoil that aids in the calm and clear treble they are capable off....I know the Mr Gerhard aus Deutschland has similar findings....
Hi,
Yup, hence I am specifying capacitors with rather much better than average vibration properties.
I may not feel like fully explaining each and every design recommendation (in part because I cannot see why I should give hard-won design tricks away, in part because it always get me into silly argument with fecking ejits that demand proof and will not STFU over it, in part because I lack the time to explain everything and in part because being a little cryptic makes average people take you more serious)...
Yup. Encased ones potted with resin often also helps. While I suspect that many here laugh at that Yamamoto Dude in Japan who pots quite generic film caps in wooden sleeves and will laugh what he does off as voudoun (voodo), his apparent madness has serious methode.
Not true, stacked types or better silver mica will, if correctly constructed, SIGNIFICANTLY better the squished racetrack types.
Also, foil & film types are less sensitive than metalised film and the heavier foil (compared to default Alu Foil) types have even lower sensitivity (Tin, Copper, Silver).
Yes, plus some makers line the inside of the Can with vibration absorbing materials (often Bitumen) and use special grades of paper to damp resonances/vibrations further. The difference between such "audio grade" types and generic 'lytics is not easy to measure, but quite audible. Opening up some "audio grade" types and comparing them to generica can be illuminating, if one knows what to look for.
Sure. In this case we also need to account for the mechanical resonances passing AC current through the capacitors creates.
Many plastic film types makes lovely "tweeters" if you sweep them with a high level, high frequency sinewave...
Ciao T
Yup, hence I am specifying capacitors with rather much better than average vibration properties.
I may not feel like fully explaining each and every design recommendation (in part because I cannot see why I should give hard-won design tricks away, in part because it always get me into silly argument with fecking ejits that demand proof and will not STFU over it, in part because I lack the time to explain everything and in part because being a little cryptic makes average people take you more serious)...
Yup. Encased ones potted with resin often also helps. While I suspect that many here laugh at that Yamamoto Dude in Japan who pots quite generic film caps in wooden sleeves and will laugh what he does off as voudoun (voodo), his apparent madness has serious methode.
Not true, stacked types or better silver mica will, if correctly constructed, SIGNIFICANTLY better the squished racetrack types.
Also, foil & film types are less sensitive than metalised film and the heavier foil (compared to default Alu Foil) types have even lower sensitivity (Tin, Copper, Silver).
Yes, plus some makers line the inside of the Can with vibration absorbing materials (often Bitumen) and use special grades of paper to damp resonances/vibrations further. The difference between such "audio grade" types and generic 'lytics is not easy to measure, but quite audible. Opening up some "audio grade" types and comparing them to generica can be illuminating, if one knows what to look for.
Sure. In this case we also need to account for the mechanical resonances passing AC current through the capacitors creates.
Many plastic film types makes lovely "tweeters" if you sweep them with a high level, high frequency sinewave...
Ciao T
John,
Don't worry, no-one will take this "tinfoil" stuff serious, unless it has some stupid brandname written on it and is presented as SupercalafajalistickespeealadojusSonicCap or some such.
I have been preaching using these in preference to other options for over decade, doesn't look like anyone is listening.
Just look about people gushing about the silver metalised plastic caps from some marketer, while giving their vastly superior sounding tinfoils short shrift.
Our secrets are perfectly safe in the public domain...
Don't worry, no-one will take this "tinfoil" stuff serious, unless it has some stupid brandname written on it and is presented as SupercalafajalistickespeealadojusSonicCap
or some such.I have been preaching using these in preference to other options for over decade, doesn't look like anyone is listening.
Just look about people gushing about the silver metalised plastic caps from some marketer, while giving their vastly superior sounding tinfoils short shrift.
Our secrets are perfectly safe in the public domain...
A british cap manufacturer tested this with a laser interferometer. They measured the vibration of the cap material as a result of the electrostatic and electromechanical forces due to signal.
Worst case, the vibration amplitude was a bit less than the diameter of a molecule. Totally inaudible.
jan didden
Hi,
Well, that is all swell.
Please take a sufficiently capable Amplifier (try around 50V RMS) that does not mind driving a big capacitive load. Place the capacitor under test across the output, (maybe as part of the Zobel?), then sweep the amp with a test signal...
Please explain the differing levels in sound herd from different caps and the different frequencies, if the vibration is only one molecule I think we are talking about humongous large, long muthfocka's of molecules...
Honestly, it's easy to try.
What I find disturbing is just how noisy (as in mechanically producting the test-signals as transducers) most power amplifiers (Tube and Solid State) are, when tested in quiet rooms... I have heard noisy caps, noisy transistors, noisy power IC's, noisy resistors, even noisy wiring.
Ciao T
Well, that is all swell.
Please take a sufficiently capable Amplifier (try around 50V RMS) that does not mind driving a big capacitive load. Place the capacitor under test across the output, (maybe as part of the Zobel?), then sweep the amp with a test signal...
Please explain the differing levels in sound herd from different caps and the different frequencies, if the vibration is only one molecule I think we are talking about humongous large, long muthfocka's of molecules...
Honestly, it's easy to try.
What I find disturbing is just how noisy (as in mechanically producting the test-signals as transducers) most power amplifiers (Tube and Solid State) are, when tested in quiet rooms... I have heard noisy caps, noisy transistors, noisy power IC's, noisy resistors, even noisy wiring.
Ciao T
Last edited:
Hi,
Certain BJT's as well, I forgot which ones, they where popular in Pro-Amp'd in the 1980's.
I have heard power tubes sing like canaries as well.
OPT's often are extremely noisy.
In fact, many picups are mechanically extremely noisy as well...
Only truely monolithic components seem immune.
Ciao T
The older turtle can metal lats were especially vocal on high power sine waves. The OPTs in tube power amps many times also. Live kind of amps.
Certain BJT's as well, I forgot which ones, they where popular in Pro-Amp'd in the 1980's.
I have heard power tubes sing like canaries as well.
OPT's often are extremely noisy.
In fact, many picups are mechanically extremely noisy as well...
Only truely monolithic components seem immune.
Ciao T
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.