Yes, leakage is higher with organic polymers compared to wet aluminum (100+ uA vs. 10+ uA), but we're still talking about microamps. If your circuit has a problem with delivering a few hundred microamps, then you need to become a better designer.
Low ESR is never a problem for decoupling unless the design is not very good. Low ESR is extremely valuable for power supply and noise decoupling, for example. ESR is stable with temperature also. So, they can handle much higher ripple currents than wet types. You can also get away with using a much smaller value to achieve the same results.
Organic polymers perform far better at high frequency than wet aluminum types. They have very low ESR and ESL compared to wet aluminum types.
There is no perfect capacitor. Know what you're using and why. Here's an article about using various kinds of capacitors in LDO regulators: LDO Bypass Capacitors: Analog Dialogue: Analog Devices
Typical LDO reference bypass pin impedance is >100k. Watch your leakage... in this position a piezoelectric sensor (ie, X7R cap) can produce interesting effects...
BTW good quality electrolytics that you'd use for DC-blocking leak much less than 1µA... a few days ago I got something like 5nA leakage on a 470uF cap, measured over several days.
😀
I measured some caps, for 2.5mm pin spacing all wet electrolytics scored around 4nH ; OSCON of same pin spacing scored about 3.5nH. Huge difference 😀
Facedown terminals solid polymer tantalums have low ESL. One of those scored 0.5nH. Bit expensive though.
Polymer caps are amazing in switching supplies though, really low losses.
Written by a diyaudio member !
I like ADP151. Nice LDO.
Last edited:
Always liked Chemicon's...Panasonic/Rubycon etc....Not too much experience/usage with Elna or Nichicon stuff...though hear about being pretty nice...
Used a lot of CDE and Illinois though, that's kinda more run of the mill/one dimensional sounding/imparting though imo.
Used a lot of CDE and Illinois though, that's kinda more run of the mill/one dimensional sounding/imparting though imo.
Last edited:
I'be used the Mundorf AG, however, would likely go with Panasonic or Epcos, now, nice caps though...
Everything I see so far are 500V units; no low voltage parts.
Guess I'll stick with Nichicon. Lots more affordable.
--Damon
I would like to ask you guys for an "Engineering test" -> testing a cheap capacitor against an expensive (audiophile) cap.
The circuit is measured and you need to find a capacitor with the right specs for optimal decoupling.
The audiophile cap is selected "by ear" for best sound quality possible.
I would like to know how to measure the circuit and find the right capacitor?
Then use the experience for my Squeezebox supplied by a Super regulator.
Anybody want to join?
The circuit is measured and you need to find a capacitor with the right specs for optimal decoupling.
The audiophile cap is selected "by ear" for best sound quality possible.
I would like to know how to measure the circuit and find the right capacitor?
Then use the experience for my Squeezebox supplied by a Super regulator.
Anybody want to join?
Not sure it will give you a valid test : you have some capacitors into a SB (touch or Duet ?) before the internal LDO regs : at least : one cap at the DC plug + some smd tantalum or polymer-tantalum after ! And a PS is not only one cap ?!
But if you try a Black Gate N to bypass directly the Xilinq FGPA... your SB will enter in another dimension if you use the digital output of your SB (giving you a proof than the good cap at the right place is important and not always a magic !).
Find off all those noise around the Silmic 2 : but in serie in a - ++ - or +--+ for DC blocking at he output of a buffer if needed I find this cap in decoupling not so good most of the time : low end is ridiculous, high end also...maybe worth it if you like only listen mediums ! 😛. Was it not an ask of Marantz to ELNA for their HDAM module in the 80's ?
But if you try a Black Gate N to bypass directly the Xilinq FGPA... your SB will enter in another dimension if you use the digital output of your SB (giving you a proof than the good cap at the right place is important and not always a magic !).
Find off all those noise around the Silmic 2 : but in serie in a - ++ - or +--+ for DC blocking at he output of a buffer if needed I find this cap in decoupling not so good most of the time : low end is ridiculous, high end also...maybe worth it if you like only listen mediums ! 😛. Was it not an ask of Marantz to ELNA for their HDAM module in the 80's ?
I have an squeezebox Classic with separate psu for the display and input to the Internal 3.3v regulator for digital circuits and and connected to my ES9018 DAC sounds great now with a 470uF oscon cap for psu and input on squeebox. The bass is a bit to much and Sound generally to Clinical
That is something I'd like to do also, have the same circuit, only difference being caps, one version sounding great, the other sounding bad, and correlate that with a measurable difference...
I don't, but I'm willing to learn, so if you are willing to impart your
knowledge I'm sure I'll learn and appreciate it.
I've had some laying around here in a bin somewhere so would appreciate
knowing how to use them.
I'm not really into parts swapping and listening tests for every little tweak... I model, measure the stuff (network analyzer, soundcard, transient) and if it looks good, then I dont worry about it. So far I can't complain about the sound 😀
That sounds great🙂
Can you tell what capacitor are you using now?
Could you measure the "bad sound" capacitor was not good for your circuit?
For power supplies, I haven't done capacitor listening tests, because it's not easy to do ABX on a capacitor... you'd need to build two identical circuits and change just the cap, probably. But then some other components may also be different... I test impedance and transient to make sure there are no resonances or time constants in the audio band, and that's it.
So, from a purely objectivist, measurements-based point of view, for example, as supply cap, Panasonic FM/FR is great. Very cheap, ultra reliable, vanishingly low leakage, zero microphony, near indestructible (I got several which were severly reverse biased and show no change in performance), ESR "just right", etc.
What I wanted to know is if someone happened to change a cap in a circuit (opamps, dac, whatever) and heard an large change in sound quality as a result. So, ideally, a simple circuit (that I could replicate) and a not subtle change (that I could also replicate). Then I'd like to measure both cases and see what has changed.
So, basically, if I have 2 identical circuits (measure the same, and indistinguishable in ABX), then I change a capacitor in one, and this is ABXable with a high score, then it would be very interesting to measure what actually changed inside, dont you think ? Is it purely impedance effects ? Interaction with regulator ? Some kind of distortion ? etc.
I would like if you could measure on your Super regulator + Squeezebox Classic setup🙂
I have ordered several capacitors - 47uF +100uF Elna Silmic 2 / Nichicon fine gold.
Do you have some of them to test with and some OScons capacitors?
It could also be interesting to know how much size matter 47 or 220uF regarding impedance stability issues?
I did some tweaking, sorry no measurement made.
I´m using Elna Silmic II 100uF / 6.3V in output (Sjostrom super reg) and in input U8 regulator inside Squeezebox SB3, on the output of the 3.3V U8 reg I have connected a small red BG N 10uF 50V
(standard SMD 10uF / 16V connected to U8 IN/out removed)
It sounds very nice, sound is balanced with all the details you wish with a bit "dark" high end sound.
Sound is still a little bit "tight" and middleway have a small "soft" sound colour.
If you use Oscon 470uF /10V the sound is "loudness" big bass, sharp and messed up treble.
I´m using Elna Silmic II 100uF / 6.3V in output (Sjostrom super reg) and in input U8 regulator inside Squeezebox SB3, on the output of the 3.3V U8 reg I have connected a small red BG N 10uF 50V
(standard SMD 10uF / 16V connected to U8 IN/out removed)
It sounds very nice, sound is balanced with all the details you wish with a bit "dark" high end sound.
Sound is still a little bit "tight" and middleway have a small "soft" sound colour.
If you use Oscon 470uF /10V the sound is "loudness" big bass, sharp and messed up treble.