Deleted

[maty]

Deleted

 

[trl]

Talking about tweaks, the last has involved an audiophile capacitor bypass in my modded KEF Q100.

https://www.audiocircle.com/index.php?topic=166472.msg1769563#msg1769563

Looks nice indeed, but a REW measurement done prior and after would explain the sound difference. Maybe the crossover frequency got changed a bit...or some group delay got increased.

 

[Julf]

The difference is such that the whole family has hallucinated by the change.

Ever heard of Clever Hans?

 

[JohnYang1997]

Just because something "doesn't exist in nature" doesn't mean it is audible.

We need to make a human hearing filter. After the filter all are the same.

 

[solderdude]

First one speaker and after the second to verify the change.

I have no logical explanation, as with the ferrite. 3.9 uF + 0.022 uF = 3.9022 uF. The resistence must be almost the same too.

The difference is such that the whole family has hallucinated by the change.

- End off topic -

Re-opening topic again..

At 20kHz the impedance of a 22nF = 360 Ohm.
At 20kHz the impedance of a 3.9uF = 2 Ohm.

0.55% of all current going through the tweeter is coming through the 22nF the other 99.45% still though the 3.9uF
The current through the 22nF is 45dB lower than the total current so 0.05dB is contributed by the 22nF.

I assume someone else already did the math for you as well.

Adding small capacitances in parallel to bigger ones is done in decoupling (the speaker filter is coupling) and is done for a reason.
The reason being keeping the internal resistance of the to be decoupled device as low as possible and is essential this is done as close as possible to the component. In this case we talk MHz to GHz range because larger capacitors start to rise in impedance again and smaller sizes take over.

 

[solderdude]

The ONLY way to test this would be to blind test with the 22nF via a switch operated by someone else and you don't know if it is in or out of the circuit. I am quite certain you, nor the whole family can hear any difference.
You can extend such an experiment over many months as well.
Pretty sure you will end up with a close to perfect 50% score.

Maybe it has something to do with harmonics, like resistors with very low inductance as Mills and others

No.

 

[solderdude]

Yes, I am incapable and have several bottlenecks and proud of it
Not a Diyaudio member.

 

[Veri]

It seems obvious that you would be incapable. If I say that the improvement is AMAZING, it is clear that anyone can hear it, even the elderly.

Sigh.

 

[trl]

In my second system, so cheap, I only have one now, presumably, the amplifier. The others bottlenecks have been eliminated or very minimized without damaging the sound.

It bothers me a lot to not know why the cap bypass and the ferrite improves the sound. No one offers a convincing answer, at least to me, but it is not obvious that the great sound improvement has occurred.

Why does it work? Crossover with RF ferrite (wires) and 10nF polystyrene (bypass) March 2018
https://www.diyaudio.com/forums/mul...rf-ferrite-wires-10nf-polystyrene-bypass.html

And since I am methodical, next week I will change the resistors. First only in one box to compare with the other.

It looks like a simple 2-way filter like this one here:

Source: http://hyperphysics.phy-astr.gsu.edu/hbase/Audio/cross.html
​

Maybe the added capacitor will lower tweeter's low-pass filter frequency with few hundred of hertz, but not sure without seeing the actual schematic and components values.

The ferrite bead might be causing some sort of oscillations (or post-ringing?) to the amplifier itself...but this is so easy to check with a scope connected to speaker wires or with REW and any microphone, by simply comparing before vs. after measurements.

Without couple of simple measurements I don't think anyone will be able to tell the whole truth here.

 

[maty]

I have deleted the comments. Better in other forums with threads opened about the question.

 

[Thomas savage]

@maty So we have another mess and you've wasted my time and members time who have gone to the trouble of replying to you.

Consider yourself on holiday.. 2 week suspension.