My "audiophile" LM3886 approach

bypass is different from decoupling.

The two terms are not interchangeable.
I tend to disagree. Many industry professionals use and understand both terms for the same thing, usually having a preference. I won't dispute the choice of word used by people like Paul Brokaw and Douglas Self (them being in more the "decoupling camp").
EDIT : Quote from Analog-Devices, AN202 "An IC Amplifier User’s Guide to Decoupling, Grounding, and Making Things Go Right for a Change", by Paul Brokaw :
"The key to understanding decoupling circuits is to note
where the actual load and signal currents will flow. The
key to optimizing the circuit is to bypass these currents
around ground and other signal paths"

:)
 
Last edited:
Hi Andrew,

I've thought that they were interchangeable but after a brief google search I understand that this is not so but the difference is not completely clear to me.

Can you define the exact difference,please?

Thanks in advance

Bypass is generally route specific and probably limiting in spectrum, decoupling is preventing any one from influencing the other(s) so each can operate independently undisturbed. Maybe a more appropriate way of saying it is that "bypass" is a method whereas "decoupling" is the objective.
 
Last edited:
Thanks again for you constructive contribution.................................................This is the n-th demonstration that your goal is not help nor instruct.

I really hope that you can solve your problems...
Can you define the exact difference,please?
no means no, I cannot do what you ask.

no, certainly does not mean that worthless diatribe you posted in your reply.
 
i know i've been told multiple times on this matter(and i always reply back in same manner as i'm doing now) but i just can't ignore my ears when it comes to final judgements.

there's no detrimental aspect whatsoever with large axial caps in c7 as far as my ears are concerned.
everything's improved when a higher quality cap is put in place. musicality, harmonics, texture, you name it, everything.
in fact the effect of this cap is so significant that i couldn't mind spending the most of my budget on it, because i believe that it's the key location for improvement once the blackgate is in c9.

how crude the small through hole caps sound will only be realized after c7 is swapped out with russian teflon, pio, mundorf series ect- i've tried them all and they're all worth their price of admission.
if someone's willing, i urge you to give it a try. (clave?) you might not be sorry that you did ;)
 
Last edited:
i know i've been told multiple times on this matter(and i always reply back in same manner as i'm doing now) but i just can't ignore my ears when it comes to final judgements.
...
if someone's willing, i urge you to give it a try. (clave?) you might not be sorry that you did ;)

As I've said I trust that you hear an improvement, no doubts about it.

But I can't ignore nor recommend that, to be effective as a bypass cap (and that bypass is needed), C7 should be tiny.

I've 10nF K40 russian PIO, I'll try them on C7 but I'm very skeptical they can sound better than FKP2... I've already tried them on the DCB1 and, while musical, they were overly colored and lacked detail.

When they arrive (I've ordered them for my crossovers) I'll try also Mundor SGO.

If you want to try a cap that really shines on C7 and it's effective too buy a NOS ERO KP1834 (polystirene film/foil) from 10nF and up, you can find them on eBay.
 
hi clave,

thanks for considering my recommendation. i'm humbled...

i can tell you that you'll likely find the same result with the russian pio as you did on your dcb1. your impression of the cap certainly doesn't differ from mine- i also found them to be very musical yet severely veiled.

it was when i tried the russian teflon that i realized there was another upgrade path. this was incidentally my first stab at abominating c7, and the result floored me. going from 100nf mkp4 (i had reverted back from 22nf fkp2 because the dynamic & tone balance was lacking) the extension opened up to no end with transparency that was unmatched.

the teflon has a truncated top end tho, that mundorf supreme cured.

i've also tried rifa styroflex, mundorf silver/oil, silver/gold and even reverted back to mkp just to see if i had deviated too far. nope, mkp is still no match. there's definitely some magic to be tapped out of c7 yet. ;)

please don't make your judgement until you try a higher end cap than the russian pio. thanks :D
 
As I've said I trust that you hear an improvement, no doubts about it.

But I can't ignore nor recommend that, to be effective as a bypass cap (and that bypass is needed), C7 should be tiny.

I've 10nF K40 russian PIO, I'll try them on C7 but I'm very skeptical they can sound better than FKP2... I've already tried them on the DCB1 and, while musical, they were overly colored and lacked detail.

When they arrive (I've ordered them for my crossovers) I'll try also Mundor SGO.

If you want to try a cap that really shines on C7 and it's effective too buy a NOS ERO KP1834 (polystirene film/foil) from 10nF and up, you can find them on eBay.

I wasnt aware that we can use 10nf on c7. Im currently using a 0.1uf BG NXhiQ on that position. Maybe ill try a FKP2 instead. Today i swapped the 330pf FKC for a 300pf MIAL polystyrene in c32. The sound is more neutral now, the FKC had a slight euphoric presentation.
 
How about NO input cap at all? Assuming that the typical source has AC-coupled output anyway, and even if not -- unless there's 100mV of output offset (very very unlikely) -- I see no true need for an input cap. It is just prone to pickup noise, the bigger the more so...

EDIT: Ah, it's about the LM318 rail-to-rail decoupling cap. Now, I see some impedance fighting games there... hard to track down without proper test equipment, but when it does have that much significance to sound it can be measured, in my experience.
 
Last edited:
thanks for considering my recommendation. i'm humbled...

You're welcome but in my BOM, anyway, there will never be such a big cap.

It's just curiosity. ;)

mkp is still no match. there's definitely some magic to be tapped out of c7 yet. ;)

I know it well, it's more than 2 years I'm recommending a film/foil cap...

please don't make your judgement until you try a higher end cap than the russian pio. thanks :D

I've also those russian Teflons.

Please, do a favour to yourself, try the KP1834.

I wasnt aware that we can use 10nf on c7. Im currently using a 0.1uf BG NXhiQ on that position. Maybe ill try a FKP2 instead.

Please tell us which cap will you prefer.

The NX HiQ is already a great cap.

Today i swapped the 330pf FKC for a 300pf MIAL polystyrene in c32. The sound is more neutral now, the FKC had a slight euphoric presentation.

Yes, polistyrene caps are utterly transparent.

Not suitable for all positions and hard to find nowadays, though.

How about NO input cap at all?

While with no input cap it works perfectly and sound great the designer used it, along with other input nets, to polarize the LM318 in a particular way.

Also there's a designer choice and a strict relation between the input cap and the 220uf feedback one.

At least Penasa wrote me so.... ;)

With my limited knowledge I have only a vague idea on what goes on...

EDIT: Ah, it's about the LM318 rail-to-rail decoupling cap. Now, I see some impedance fighting games there... hard to track down without proper test equipment, but when it does have that much significance to sound it can be measured, in my experience.

Can you elaborate?

Thanks in advance.
 
I have 1uF for C4, but also removed C1/C2. In parallel with C4, I also have 1000uF 63V cap. What I did was take big steps to increase these till there was no significant change, the gradually get an optimum value. I think it's better to focus on the value first, than try different brands of capacitors.

The difficulty in using only the listening process for evaluation is how can one tell the change is improving the amp but revealing deficiency in other parts of the system.
 
Last edited:
The difficulty in using only the listening process for evaluation is how can one tell the change is improving the amp but revealing deficiency in other parts of the system.

that is a good question which made me ponder for a bit. i'm afraid i don't have a set answer for such a broad categorization as 'deficiency in sound'

however if we to delve into specifics, answers become easier.
let's consider distortion f.e. -a main criteria for measured fidelity that our ears are also well tuned for.

for that it's simple- if it sounds improved, it has improved. :)

because there can't be a case where the amp's distortion lowers while letting more distortion through. all distortions are added up at the final output of woofers, and subtraction in one area cannot carry over to the other component in a greater number.

same with transient response. if it's gotten 'faster' in one place, it will sound 'faster'.

as for tonality, coloration, i think those are tricky subjects that even the measuring devices are not sufficient enough to handle. and reason why even people here say this hobby is part alchemy, i believe ;)
 
Last edited: