By the way, I usually use Panasonic FR where it makes sense to use smallish capacitance low ESR electrolytics, but since they just go up to 3300uF I wanted to try Nichicon UHW as well.
I just had the chance to use them (4700uF @25V) on a CD player that I wanted to tweak a bit.
Very nice results so far.
I just had the chance to use them (4700uF @25V) on a CD player that I wanted to tweak a bit.
Very nice results so far.
I actualy run the EPCOS Sikorel B41550 / 570 in my F6 and even at full load the caps stays at room temp , not even a single degree more ;-)
they are designed and build for much much more stress than audio device can give , and they do a great job in class A amp , tanks to the very low esr and very high ripple current .
.
they are designed and build for much much more stress than audio device can give , and they do a great job in class A amp , tanks to the very low esr and very high ripple current .
.
Last edited:
Thanks! The first two I knew. In this case I was wondering the difference between Epcos and Panasonic, they have the same capacitance and the Panasonic even has a higher voltage rating. Interesting that Epcos marks the plus-pin in that cap, never seen that before.Higher voltage = bigger case size.
Higher capacitance = bigger case size.
Higher dielectric thickness = bigger case size.
Higher Factor of Safety for the voltage rating = bigger case size.
And this Nichicon, which I know nothing about, that seems to have a shorter life.
LNT1J153MSE Nichicon | Capacitors | DigiKey
When i recapped the big Krell my shortlist of available caps included the same series Nichicon, Rifa PEH169 and Siemens B41456B9479M.
Surprisingly, the Nichicon sounded to me almost as good as the Siemens and much better than the Rifa. At the end used the Siemens and couldn't be happier. Panasonic caps are definitely not my taste.
This alleged poor performance may have had more to do with the location and provision (or non provision) of local supply rail decoupling.
It's the local supply rail decoupling that supplies the current to meet transient demands.
If this decoupling is inadequate, or in the wrong location, or omitted, then HF performance can be poorer, even getting as bad as oscillation.
The main smoothing capacitance has two main purposes:
a.) to smooth the ripple out of the rectifiers.
b.) to recharge the local supply rail decoupling whenever they become partially discharged.
One thing the main smoothing capacitor/s do not do, is to supply current for quickly changing HF transient demands.
It's the local supply rail decoupling that supplies the current to meet transient demands.
If this decoupling is inadequate, or in the wrong location, or omitted, then HF performance can be poorer, even getting as bad as oscillation.
The main smoothing capacitance has two main purposes:
a.) to smooth the ripple out of the rectifiers.
b.) to recharge the local supply rail decoupling whenever they become partially discharged.
One thing the main smoothing capacitor/s do not do, is to supply current for quickly changing HF transient demands.
Last edited:
No
Assuming a test using a design (questionable or not is irrelevant since it is the same in all instances), if you hear differences, then your values and parasitics comparison is questionable. Everything else is equal during the trials.
I agree with Andrew about the value of local decoupling.
But unfortunately differences can still be heard. They might be less pronounced, but they are still there.
Even behind regulators which should provide the highest amount of isolation and decoupling.
But yes, if you are on a budget, keep your best caps first and last or closest to where they have to do the biggest job, supply and decouple the load or even out rectification ripple.
Assuming a test using a design (questionable or not is irrelevant since it is the same in all instances), if you hear differences, then your values and parasitics comparison is questionable. Everything else is equal during the trials.
I agree with Andrew about the value of local decoupling.
But unfortunately differences can still be heard. They might be less pronounced, but they are still there.
Even behind regulators which should provide the highest amount of isolation and decoupling.
But yes, if you are on a budget, keep your best caps first and last or closest to where they have to do the biggest job, supply and decouple the load or even out rectification ripple.
Last edited:
he quite clearly reported that the Rifa did not perform as well as the other two that were tried.much better than the Rifa
All amplifiers need some form of current source that can react quickly to fast transients.I don't understand what you're 'getting at', Andrew
Alleged poor performance of what? The recapping efforts of 'a-n'? Rifa Caps are rubbish? All power supply caps need bypass/decoupling?
Or that dAgostini doesn't know about power supply design?
If one omits local supply rail decoupling and tries to get fast changing current from capacitors at the far end of inductive cabling, then one has failed as a Builder/Assembler.
and locating "bypassing" at the PSU is more likely to result in ringing on the supply lines than to help make the amplifier work properly.All power supply caps need bypass/decoupling
Well, there are amps that show effectively ZERO reaction to what they are supplied with, the Modulus-86 from forum member @tomchr for example.
That claim is irrelevant and out of topic
If you hear a difference between capacitors in a design, then it is the assumption of capacitor equality that was wrong.
The design is a given constant
That claim is irrelevant and out of topic.......
Hey man, why dont you let people say what they wanna say
.
Hmmm,
I had a filter blow in a SMPS and I was looking to preplace with
higher quality etc.
I found these, maybe this is what James is talking about,
made by Vishay BC...
Needs are for power supplies, with ultra-hi ripple current,
6000 hours at 105 C, Hi ripple hi reliability.
Does it matter if its for Solar?
I had a filter blow in a SMPS and I was looking to preplace with
higher quality etc.
I found these, maybe this is what James is talking about,
made by Vishay BC...
Needs are for power supplies, with ultra-hi ripple current,
6000 hours at 105 C, Hi ripple hi reliability.
Does it matter if its for Solar?
Attachments
Re:Cap burn in
To ALWEITid you allow for cap burn in?In my past experience working with tons o caps they are all different ...so boutique caps like Blackgates must reform everytime they are turned off whilst others take very little time to sound their best .Just a suggestion...
To ALWEITid you allow for cap burn in?In my past experience working with tons o caps they are all different ...so boutique caps like Blackgates must reform everytime they are turned off whilst others take very little time to sound their best .Just a suggestion...
I am interested in all new caps. I have 63V ones and most metrics are better for them (lower esr, higher ripple current etc), but the price is significantly better for the 40V ones.
And yes I was hoping for some subjective experience from using both. The whole thread is about sharing subjective input anyway
And yes I was hoping for some subjective experience from using both. The whole thread is about sharing subjective input anyway
Last edited:
Re:40 volt vs 63 volt
I conducted a quick charge and discharge comparison between a new/old stock Elna Hifi grade 100 volt dc unit vs a new Nichicon Gold tone 35 volt dc one...first I charged then up through resistance(same) to working voltage to reform them for a time and observe the miliamp meter to see current dropping to zero...then discharged them both and observed time it takes for charge voltage to drop below threshold of one volt.Then I repeat experiment but charge both to only 35 volts...out come:The Elna capacitor discharged noticeably much quicker than the Nichicon...I have to conclude that this would probably be audible to your ear...maybe if both were allowed to be formed up for a longer period of time maybe there wouldn't be much difference ...If one owned mono blocks this could be A and B'd and reported on.
I conducted a quick charge and discharge comparison between a new/old stock Elna Hifi grade 100 volt dc unit vs a new Nichicon Gold tone 35 volt dc one...first I charged then up through resistance(same) to working voltage to reform them for a time and observe the miliamp meter to see current dropping to zero...then discharged them both and observed time it takes for charge voltage to drop below threshold of one volt.Then I repeat experiment but charge both to only 35 volts...out come:The Elna capacitor discharged noticeably much quicker than the Nichicon...I have to conclude that this would probably be audible to your ear...maybe if both were allowed to be formed up for a longer period of time maybe there wouldn't be much difference ...If one owned mono blocks this could be A and B'd and reported on.
Re:63 V VS 40 V
Now if you want to play around with your equipment then lead/trace inductance can play a significant role in determining how things sound...several amplifiers I have modded have benefitted by relocating the main pwr supply caps physically closer to the devices that draw the most current(where practical)...Let's say within less than an inch ,then ,fire up the amp and listen to it...move the cap where it ties in to the trace/lead a little further away each time and listen to your circuit until you find the sweet spot(and leave it).Try it, you'll like it!
Now if you want to play around with your equipment then lead/trace inductance can play a significant role in determining how things sound...several amplifiers I have modded have benefitted by relocating the main pwr supply caps physically closer to the devices that draw the most current(where practical)...Let's say within less than an inch ,then ,fire up the amp and listen to it...move the cap where it ties in to the trace/lead a little further away each time and listen to your circuit until you find the sweet spot(and leave it).Try it, you'll like it!
- Home
- Amplifiers
- Power Supplies
- Best capacitors for power supplies update