Carl_Huff said:
My english is far (far far away
) from perfection but also the italian text was a bit confusing.I'll try to summarize concepts so it would be easier to translate (for me):
- Power supply noise reduction is carried by C4 and C7 placed between V+ and V-
- In a lot of audio designs you'll find bypass caps between rails and ground
- These bypasses improves the "virtual ground"
- These bypasses are useless for PS noise reduction
- OpAmps are only sensitive to noise present in only one rail
- Caps between PS rails do a better job filtering PS noise.
I'm not an expert but I think that what Mauro means is:
- if the same noise signal is present in both rails the virtual ground and rails are floating following noise so, in a relative meaning, it doesn't exists
- If noise is present only in one rail the cap between rails filters it
- In this context bypasses to ground are useless.
If I remember correctly in his posts Mauro said that oscillation is largerly prevented by the PCB design and resistors tight tollerance.
But apart all that remains the fact that the original circuit, which is stable and doesn't oscillate, doesn't have that bypasses.
Hi,
opamps and chipamps operate in ClassAB.
The currents in the rails circulate from either supply rail to power ground.
The decoupling at the chip power pins are there to reduce the impedance of those current circulation routes, to power ground.
Some opamps need decoupling on one supply rail only, some need decoupling from rail to rail and some need decoupling from both rails to ground.
Many opamps require a combination of these decoupling strategies.
Start by consulting the data sheets for these two chips.
opamps and chipamps operate in ClassAB.
The currents in the rails circulate from either supply rail to power ground.
The decoupling at the chip power pins are there to reduce the impedance of those current circulation routes, to power ground.
Some opamps need decoupling on one supply rail only, some need decoupling from rail to rail and some need decoupling from both rails to ground.
Many opamps require a combination of these decoupling strategies.
Start by consulting the data sheets for these two chips.
Let's not get into arguments about the technical aspects of this circuit design. That's a big reason why the previous MyRef thread became so lengthy and convoluted: many posters wanted to debate Mauro about why his circuit wouldn't perform as well as it does, and they all had suggestions about how to "improve" it. Mauro became frustrated with these "experts" and took great pains to describe his circuit topology and design philosophy, but it didn't always come clearly through translation. Those small bypass caps were added to the original version of the circuit, so we know it performed adequately without them. End of debate. Some felt their addition improved the performance, others weren't so sure. Try for yourself and decide, but don't discourage others from experimenting because you think you know it won't work.
One bit of caution: when you modify one monoblock and then listen in stereo with the other, unmodified monoblock, you might not achieve satisfactory performance. Obviously, any change to the signal of one side of a stereo pair will introduce distortions between channels and sound odd, especially with a characteristic as sensitive to mismatch as soundstage and imaging. You might not like what you hear, but that doesn't mean you didn't make an improvement to the single channel. Does that make sense?
Dario, you say the sound improved when you removed all small bypass caps. Is this a case where your sockets and extra wire might be more of a factor than the caps themselves? As the values of a component get smaller, any additional factors become larger in proportion. With these small value caps and their location in the circuit, perhaps the extra impedance of sockets and wires is as much a factor in performance as the components themselves. By removing the caps completely, you are in essence removing the sockets and extra wire from the circuit as well.
Most builders believed the sound improved with the addition of those small bypass caps. The change was so significant that the circuit was modified to include them. How can they all have been wrong? I'm not saying you didn't hear an improvement; I would never doubt someone else's ability to hear. I'm saying your test rig might not be valid for these small value components in a sensitive area of the circuit.
Peace,
Tom E
One bit of caution: when you modify one monoblock and then listen in stereo with the other, unmodified monoblock, you might not achieve satisfactory performance. Obviously, any change to the signal of one side of a stereo pair will introduce distortions between channels and sound odd, especially with a characteristic as sensitive to mismatch as soundstage and imaging. You might not like what you hear, but that doesn't mean you didn't make an improvement to the single channel. Does that make sense?
Dario, you say the sound improved when you removed all small bypass caps. Is this a case where your sockets and extra wire might be more of a factor than the caps themselves? As the values of a component get smaller, any additional factors become larger in proportion. With these small value caps and their location in the circuit, perhaps the extra impedance of sockets and wires is as much a factor in performance as the components themselves. By removing the caps completely, you are in essence removing the sockets and extra wire from the circuit as well.
Most builders believed the sound improved with the addition of those small bypass caps. The change was so significant that the circuit was modified to include them. How can they all have been wrong? I'm not saying you didn't hear an improvement; I would never doubt someone else's ability to hear. I'm saying your test rig might not be valid for these small value components in a sensitive area of the circuit.
Peace,
Tom E
madisonears said:
Some notes:
Automatic translation of Mauro on PSU
Mauro on PS bypassing
Well Tom only sockets was used, no wires here.
I know that sockets could modify bypasses behaviour but I'm quite confident that effect is very small.
For a final judgement is necessary that someone who has soldered those caps removes them.
BTW I've had same results with my Gainclone's bypass caps (that were soldered): once removed soundstage opened-up and dinamic increased.
Link to my thread on caps
Thanks Udaley, i really dont want to worst the quality of the sound, so i will try to build an full pre, i really dont want to spend a lot of money on an HQ pot.
Btw, now i switch my Sound Card from 44.1 to 48K and the HUM Gone :S
my card allow 48, 96, 192 using any of these value will get better sound? i really think that if the source have low bitrate, so the output will be crap too, what do you think?
Btw, now i switch my Sound Card from 44.1 to 48K and the HUM Gone :S
my card allow 48, 96, 192 using any of these value will get better sound? i really think that if the source have low bitrate, so the output will be crap too, what do you think?
Pacificblue, you stated cermaic caps are a better choice than film types for decoupling. Did you mean to write ceramics are a better choice for bypass?
Can you give us any examples of your experience with film vs ceramic in either case, or are you simply stating this as a fact?
I have installed the Mundorf SIO coupling caps into both amps. First impression is they are horrible! Now, after 12 hours play, they are getting better, but I'm still waiting for the Mundorf magic to appear. I have read these caps take 50 hours or more to settle in, so I'll be patient. Everyone raves about these caps, so they must be better than what I've heard so far.
Peace,
Tom E
Can you give us any examples of your experience with film vs ceramic in either case, or are you simply stating this as a fact?
I have installed the Mundorf SIO coupling caps into both amps. First impression is they are horrible! Now, after 12 hours play, they are getting better, but I'm still waiting for the Mundorf magic to appear. I have read these caps take 50 hours or more to settle in, so I'll be patient. Everyone raves about these caps, so they must be better than what I've heard so far.
Peace,
Tom E
parmetal
thanks for clarifying pacificblue!
for parmetal chassis. I've used one before and had it in clear alodine. I understand this is good for emi and the chassis metal is conductive.
I would like a black color but that only comes in anodized, which is non conductive. for this my ref amp, there is really no need to have the chassis conductive, everything is isolated?
only thing I would put is the IEC Earth ground and scrape the chassis where the bolt would go.
any reasons not to go with anodized?
thx
thanks for clarifying pacificblue!
for parmetal chassis. I've used one before and had it in clear alodine. I understand this is good for emi and the chassis metal is conductive.
I would like a black color but that only comes in anodized, which is non conductive. for this my ref amp, there is really no need to have the chassis conductive, everything is isolated?
only thing I would put is the IEC Earth ground and scrape the chassis where the bolt would go.
any reasons not to go with anodized?
thx
ClaveFremen said:
I've tryed again my AVX SkyCap X7R 0,1 uF ceramics, AVX BQ and MKS2.
I confirm my impressions on AVXs and MKS2s as PS bypasses they restrict soundstage and dynamic.
Ceramics are a lot better as PS bypasses
but they have the same effect of film caps but a lot smaller. If someone really wants to use PS bypasses ceramics are quite good and substantially better than film ones but for sure in my setup I'll get rid of these caps.
Obiouvsly these are my results using sockets, soldered caps could behave differently.