Little Dot MK8SE / MK6 Super Mods (All verified mods are on first page)
May 18, 2016 at 12:38 PM Post #1,546 of 4,154
Here it is http://www.diyaudio.com/forums/power-supplies/214869-single-vs-parallel-capacitor.html
It's a long thread and I'm too lazy at the moment to look for it but there was mention of making sure the impedance paths are aligned properly or some such.
 
If lends favor to MrCurwens argument if you are hearing a sonic gain via this method. Which means the grid drive mod may be something important to try. Although honestly I don't see how an active device is going to beat a capacitor in the end in terms of low distortion.
 
May 18, 2016 at 12:40 PM Post #1,547 of 4,154
Yes the WCF cap is directly affecting the tonality .
I also inadvertently had one disconnected on the left channel negative tube , and the result was about a 10% loss of some treble and focus.
That was only with one tube,  so not full signal. 
 
May 18, 2016 at 12:44 PM Post #1,548 of 4,154
You may want to try a bundle of $1 caps and see what happens. The more you parallel the closer to a perfect cap it becomes.
But you have to take into account that parallel caps add capacitance.
So you can buy like 10 100nf caps instead of 10 1uf caps.
You should also make sure that the increase in capacitance is not playing a part in better sonics.
 
May 18, 2016 at 12:53 PM Post #1,549 of 4,154
Completely off topic, sorry for that, but just to illustrate what I meant by not having much aesthetic skills:

http://kalleria.org/livejournal/Kuva1339.jpg

This is my flagship amp, not in my possession anymore. Unfortunately I didn't take a 'final' picture of it; here the buffer (sitting on top) is not yet operational, as you can see from the indicator light not being on. Also the PSU and signal chassis indicator lights are of very different brightness. I fixed this later, important mod!

Not very mailable, unfortunately. But I am thinking more and more about mailing a smaller amp for audition, first to one man in Denmark who I've been tutoring and then onwards to somebody else in Europe. Preferably somebody who has very expensive caps to try out. I don't know, I should make another thread for this, sorry for the off topic.

The amp looks clean,
and I bet if the case was black it would look sleek and very professional.

Plus it would need your signature in gold ink on the corner for finishing touch..
:)
 
May 18, 2016 at 12:57 PM Post #1,550 of 4,154
You should also make sure that the increase in capacitance is not playing a part in better sonics.

I have made sure I feel a .33uf sounds best and any more will lead to a thickening of sound in cheaper caps which I do not like.
Only The .47uf Mundorf was able to pull it off and sound very sweet and energetic,
But my caps choice is for natural trebles and I believe I have that with the VitaminQ in that position.
 
May 18, 2016 at 1:10 PM Post #1,551 of 4,154
I'll end up using Duelund interstage and Jupiter WCF caps but if parallel caps really do sound better then a single jupiter then there is some valuable information to be gained here.
Technically speaking multiple caps are better in pretty much every way but I never considered it would be as big of an improvement as you say unless I'm misreading.
 
May 18, 2016 at 1:20 PM Post #1,552 of 4,154
I find cap parallelling smears the finest details away. This is contingent on the rest of the amp being able to reproduce those finest of the fine details.

Well, out of necessity I parallel caps of same type. I prefer to use same production batch if it is possible (often not).

But two different types are definitely out.

The guitar amp DIY guys (which I also am) actually found that they can create exciting new sounds by doing dissimilar parallels with caps. So not for HIFI, at least the dissimilar parallel.


I just recalled I never responded to the message about russian caps. I've not used the ones in that post, except for K73-16 which is for assist duties. The different russian caps are as different as different American or European or Japanese caps.

SSG is best, KSG second best, KBG best in big numbers, MBGO good for all.

Some of the PIOs other than KBG have a lousy retro sound, which many like but I don't consider HIFI. Excellent for guitar amps, though.


But, in the end, if the circuit responds a lot to cap rolling, it's probably not a very good circuit. General rule, has served me well thus far. There are some exceptions.
 
May 18, 2016 at 1:20 PM Post #1,553 of 4,154
I like to mention of note last night listening session.
I was laying down and a usual track I played scared the crap out of me!

I was hearing treble/air movement of a person passing by,
that in past did not notice as the detail or ambient cues was not resolved enough for my brain to interpret it in the past.

It literally scare the crap out of me for a second!

The problem is that I do not know which mod is responsible:
1- the bias to powertubes,
2- the change in driver tubes (6sn7)
3- using EH & DR powertube pair
4- Bypassing my ygggy USB with the "F1" spdif converter, in this thread:

http://www.head-fi.org/t/803111/xmos-xu208-usb-bridges-the-latest-gen-has-arrived/0_20

:)
Oh, forgot new tube burn in!
 
May 18, 2016 at 1:23 PM Post #1,554 of 4,154
Plus it would need your signature in gold ink on the corner for finishing touch..
:)


Me and the owner of the amp I linked a picture of go to high end shows and shops to audition things.

After one of those outings we actually laughed around that maybe I should do that signature, since all the BIG LEAGUE amps seem to have it. =)

The speaker posts are gold though so technically it still is eligible in HIGH END, despite it's other sins.
 
May 18, 2016 at 1:24 PM Post #1,555 of 4,154
My topic on the cap paralleling was specific to the WCF position,

And it is good to know about possible smearing which I would not choose to do in a coupling position.
 
May 18, 2016 at 1:29 PM Post #1,556 of 4,154
I like to mention of note last night listening session.
I was laying down and a usual track I played scared the crap out of me!

I was hearing treble/air movement of a person passing by,
that in past did not notice as the detail or ambient cues was not resolved enough for my brain to interpret it in the past.

It literally scare the crap out of me for a second!

The problem is that I do not know which mod is responsible:
1- the bias to powertubes,
2- the change in driver tubes (6sn7)
3- using EH & DR powertube pair
4- Bypassing my ygggy USB with the "F1" spdif converter, in this thread:

http://www.head-fi.org/t/803111/xmos-xu208-usb-bridges-the-latest-gen-has-arrived/0_20

:)
Oh, forgot new tube burn in!
 
May 18, 2016 at 1:30 PM Post #1,557 of 4,154
I was hearing treble/air movement of a person passing by,
that in past did not notice as the detail or ambient cues was not resolved enough for my brain to interpret it in the past.

It literally scare the crap out of me for a second!

That happened to me a few times after upgrades. Spooky realism. Depends on the quality of the recording though.
 
May 18, 2016 at 1:30 PM Post #1,558 of 4,154
The WCF caps are 100% as much in the signal path as the coupling cap between input and output stages.

Of course, it is inside a local NFB loop, so it's not so critical. If the devices inside NFB loop had more gain and transconductance, the quality of the cap would become irrelevant.

Comparing to the gyrator cap. I use "assist duty quality" K73-16 there, because it is inside a very very strong full degeneration NFB loop (from FET output straight to FET input basically). I used to use very good quality caps there but they don't really make any difference inside the loop.

It is the local NFB loop inside the WCF that lowers the output impedance of the WCF and of the amp. If you can hear any difference between caps inside that loop, you know the performance of the output stage is not optimal or very impressive.

That doesnt' mean it's not GOOD. It just means the rabbit hole in fact is more deeper whether you wish to go there or not.
 
May 18, 2016 at 1:43 PM Post #1,559 of 4,154
Of course, it is inside a local NFB loop, so it's not so critical. If the devices inside NFB loop had more gain and transconductance, the quality of the cap would become irrelevant.

Hmm, not sure I understand this. The WCF cap is required for proper push pull operation, good signal integrity ensures that the bottom triode conducts oppositely to the top triode allowing proper current flow through the load.
How does more gain make the cap matter less? The WCF ensures less then unity gain anyway right?
 
May 18, 2016 at 1:48 PM Post #1,560 of 4,154
One of my test tracks (I have many for different purposes and qualities) for the creep factor is Violenza Domestica by Mr.Bungle:

https://www.youtube.com/watch?v=-qVjAFCHoaI

My girlfriend really enjoys the musical crazyness of Mike Patton. Towards the end of the track (maybe 4 mins in) there's a section where he talks directly to your terrified soul. How horrifying it sounds is a measurement of the amps fidelity.

Some decent transient tests in that one as well.


Then there's the position of singers head with regards to the microphone. There's many good tracks for this, but one to mention is Gun Street Girl by Tom Waits. I don't know if he's actually drunk while singing, but it sure seems so. =)

Also the physicality of the instruments is one way to describe realism. Every decent studio recording is much more of a live performance (if done well).

Good song though, much more pleasant than the Mr.Bungle song.
 

Users who are viewing this thread

Back
Top