You possibly linked to the wrong post. If not then trying to make a fool out of me is a good display of character deficit.
Last edited:
Remarks like this add nothing to the conversation.
You really think that a component having some wire legs and being pushed through some FR4 intrinsically improves the fidelity of low frequency (eg audio) signals in a circuit ?...
I'm assuming of course using eg MELF resistors not Thick Film and COG/NPO not X7R etc...
I have often used bi-pass capacitors in the past when tweaking consumer grade products. Maybe in ignorance but often with pleasing results.
Anyway it is well documented how a bi-passed electrolytic decoupling capacitor can have a poor impedance due to resonance and that the situation can often be improved with a small value resistor in series with the electrolytic.
How does this relate to a bi-passed capacitor when used for signal coupling?
Anyway it is well documented how a bi-passed electrolytic decoupling capacitor can have a poor impedance due to resonance and that the situation can often be improved with a small value resistor in series with the electrolytic.
How does this relate to a bi-passed capacitor when used for signal coupling?
BTW mentioned ceramic SMD caps; for hi-performance resistors, the recommendation is to use exclusively thin-film resistors not smaller than size 1206. They are not expensive down to 25ppm tempco; for best performance you want 5ppm ones, but that gets expensive. But such resistors, when limiting the voltage across them to 5V or so, are extremely linear and are pretty much perfect for practical purposes.
One good use: many power amps have a feedback network of say 22k series and 1k to gnd for 23x gain. In the very best amplifiers, the limit of linearity is actually the temperature- and voltage related changes in the feedback resistor (the 22k) as a function of Vout because that voltage across the 22k can be quite high.
Now if you use 1k thin film SMD, size 1206, 5ppm, with 22 in series to replace the single 22k R, almost all temco- and voltage-coefficient effects cancel almost completely. The resistors still show minute chnages as a result of temp- and voltage effects, but they all do it in unison and the ratio 22k to 1k nicely remains the same!
This single change can drastically inprove an already great amp!
Jan
One good use: many power amps have a feedback network of say 22k series and 1k to gnd for 23x gain. In the very best amplifiers, the limit of linearity is actually the temperature- and voltage related changes in the feedback resistor (the 22k) as a function of Vout because that voltage across the 22k can be quite high.
Now if you use 1k thin film SMD, size 1206, 5ppm, with 22 in series to replace the single 22k R, almost all temco- and voltage-coefficient effects cancel almost completely. The resistors still show minute chnages as a result of temp- and voltage effects, but they all do it in unison and the ratio 22k to 1k nicely remains the same!
This single change can drastically inprove an already great amp!
Jan
Ive noticed that the stacked film(box) type capacitors have a slightly lesser likelyhood of causing an audible resonance when used with say, a 100uf electrolytic.
One thing for certain is that capacitor(and battery) technology is moving ahead.
However, I haven't used anything like this for coupling, except for some tests just out of curiosity.
One thing for certain is that capacitor(and battery) technology is moving ahead.
However, I haven't used anything like this for coupling, except for some tests just out of curiosity.
Last edited:
no need for AP, make a RC low pass, then just amplify the V across the gnded C single/combo and use a decent usb or PC soundcard
to quash anti-resonant peaks you need the damping R in series with the hi-Q C == the small bypass cap
quality bi-polar Al electros for coupling will have ESR of <~ 0.2% Xc/120Hz
~ 20 mOhm for 100uF
to quash anti-resonant peaks you need the damping R in series with the hi-Q C == the small bypass cap
quality bi-polar Al electros for coupling will have ESR of <~ 0.2% Xc/120Hz
~ 20 mOhm for 100uF
The referrence for this is a great series of six (!) articles on combining capacitors and the pitfalls by Kendall Casor-Perry.
Here's Part 1:
https://linearaudio.nl/sites/linearaudio.net/files/bypass_capacitors_part1.pdf
The full set can be downloaded here:
https://linearaudio.nl/miscellaneous
(Scroll down; 3rd from the bottom).
Jan
Here's Part 1:
https://linearaudio.nl/sites/linearaudio.net/files/bypass_capacitors_part1.pdf
The full set can be downloaded here:
https://linearaudio.nl/miscellaneous
(Scroll down; 3rd from the bottom).
Jan
Ralf, the article give a lot of info about the effects on capacitance when you parallel several values and-or types. Was that not what we were discussing?
Try to look through a particular app and see the important info that´s there.
Jan
Good one. There are several cases though where the difference with the added caps were almost invisible. Its complex.
Jan
- Status
- Not open for further replies.