Nelson Pass said multiple times Elna silmics are one of the best electrolytic capacitor he's heard and measured... so guys please I need a break now .
Just to throw oil on the fire ( a word to word translation from french ).
I do not trust results of a blind test.
I only trust results of double blind tests.
In the case we argue here. Double is that the guys who present the amps to a listening team, do not even know wich is A or B.
There is a third party that knows. These guys have wrapped the two amps in an anonymous package.
So people who present the two packages don't know wich contains A or B.
The listening party will vote on package 1 or package 2. Statistics will be made about 1 and 2 And lastly the third party will reveal wether 1 and 2 are A and B or are B and A.
In case, I do not explain well, look at how double blind tests are performed in the medical domain.
I do not trust results of a blind test.
I only trust results of double blind tests.
In the case we argue here. Double is that the guys who present the amps to a listening team, do not even know wich is A or B.
There is a third party that knows. These guys have wrapped the two amps in an anonymous package.
So people who present the two packages don't know wich contains A or B.
The listening party will vote on package 1 or package 2. Statistics will be made about 1 and 2 And lastly the third party will reveal wether 1 and 2 are A and B or are B and A.
In case, I do not explain well, look at how double blind tests are performed in the medical domain.
Last edited:
in a true blind test you won't see what amp starts A or B you will just hear music starting and transitioning from one amp to another when one of the guys clap his hands so in a test round there are just a to b or b to a after that the music is stoped and the listener must tell which sounded better the first one or the second one so you can repeat that round as many times ...
I also have a well treated room and very high end speakers.
For example a few days ago I changed the 220uf caps from my NAD 306 that are located on power amp boared crossing from Cerafines 25v to KZ 50v.
I also posted a photo with the caps locations.
With Cerafines the sound was a little distorted and blured in high mid-high frequency so my ears did get some irritation
So I swapped the Cerafines out and with KZ the sound is now with smooth high frequency also the mid is very smooth but now the sound is very analytical but is cold .
So just by replacing those two 220uf did make a huge difference
Attachments
Wow!
I assume you are just an ignorant troll!
NAD306 high end? Shame on you, although I quite like NAD kit!
The circuit you have highlighted seems to be part of power protection / switch on delay. If you think that will have a effect on the sound you are really off your trolly and can be safely ignored.
I assume you are just an ignorant troll!
NAD306 high end? Shame on you, although I quite like NAD kit!
The circuit you have highlighted seems to be part of power protection / switch on delay. If you think that will have a effect on the sound you are really off your trolly and can be safely ignored.
is not so simple as you think those caps are affecting the sound I have a few amps and 306 is one pretty good amp
Hi robert2017,
You do realize that ESR is a useless parameter, generally speaking. If you want to know about the quality of a capacitor, you measure dissipation. Of course, you can get cheapy ESR meters (what a great marketing job for these!), but you cannot buy cheapy meters that accurately measure dissipation. 1/dissipation =Q of course.
Just so you know, dissipation is a measure of how much signal energy is lost inside the capacitor into the dielectric. This is normally highly non-linear (= distortion) whereas ESR is just resistance (impedance properly).
Time to buy some equipment actually useful for the task at hand now. You still have an uncontrolled test. The equipment I use is rather long in the tooth, but it's an HP 4263A. It's test fixtures cost more than the ESR meter does. By quite a margin in fact and I have three different test fixtures.
Again I say to you, study the relevant theory at least before you begin messing around with stuff you don't have a good handle on, then talking down people who actually do know what they are doing.
You can get an HP 4263A cheap these days as it is an old instrument. I bought mine brand new when they were first introduced. I used an HP 4261A before that. Do yourself a favour and grab the 4263A and hopefully with the Kelvin test lead set, or the fixture that is mounted directly on the front of the instrument.
-Chris
You do realize that ESR is a useless parameter, generally speaking. If you want to know about the quality of a capacitor, you measure dissipation. Of course, you can get cheapy ESR meters (what a great marketing job for these!), but you cannot buy cheapy meters that accurately measure dissipation. 1/dissipation =Q of course.
Just so you know, dissipation is a measure of how much signal energy is lost inside the capacitor into the dielectric. This is normally highly non-linear (= distortion) whereas ESR is just resistance (impedance properly).
Time to buy some equipment actually useful for the task at hand now. You still have an uncontrolled test. The equipment I use is rather long in the tooth, but it's an HP 4263A. It's test fixtures cost more than the ESR meter does. By quite a margin in fact and I have three different test fixtures.
Again I say to you, study the relevant theory at least before you begin messing around with stuff you don't have a good handle on, then talking down people who actually do know what they are doing.
You can get an HP 4263A cheap these days as it is an old instrument. I bought mine brand new when they were first introduced. I used an HP 4261A before that. Do yourself a favour and grab the 4263A and hopefully with the Kelvin test lead set, or the fixture that is mounted directly on the front of the instrument.
-Chris
Thanks for the photo.
And you didn't put your foot in it! 😀
It must be tempting to change out all the capacitors, but which ones take priority?
I agree with you but disagree when you say that you can define all parameters because the actual science won't know anything .
Or maybe you don't know everything' 🙄. Don't you think is possible.
A cap may sound fat , another cap will sound cold and detailed. other will sound restricted in 3 D image but warm and detailed.......
I mean there is more than dissipation which is also important .
The BIG BIG problem with the science is that if the science can't explain something the science conclude that thing don't exist.
So you will learn that don't exist .
10 or 30 years later when they explain that something the science conclude that things actually exist
So now you will learn that actually exist .
This is how science works ..sorry
Last edited:
I think that Robert means 'everything' and not 'anything' - may be a language problem here.
right my mistake.
I think you should start by changing those little 1uf to 22 uf because those already will have 10% or more V-loss especially the 1uf .
Also when I measured the 1000uf ones they where around 700uf and the 500 uf ones around 350uf or something .
If you can find 2 Black Gates and replace those 1000uf with that and start from there
Up to 47 uf usually I combine KZ with Silmic beneath 47uf KZ with FineGold
Last edited:
Robert it appears that you have some appreciation for the need to conduct controlled experiments to provide evidence for your hypothesis about audible differences between capacitors. This is an encouraging sign. That is the method used by the engineers and scientists who are responding to your posts. The physics and electronics that they are suggesting you learn was developed through this very experimental approach. This hard won knowledge might not capture every possible phenomenon affecting the musicality of amplifiers but it has succeeded in describing many. Those who propose new previously unidentified factors or influences carry the burden of demonstrating their existence through controlled and repeatable experiments. Others are not required to prove that such factors don't exist, you are required to prove that they do. One blind listening test doesn't get that done.
Hi robert2017,
It is important that the replacement part fits properly in the space where the original capacitor did. I can't stress this enough.
As far as testing capacitors go, testing capacitors for dissipation weeds out many bad parts, some of which have big names and are expen$ive audiophile approved ones. Another point made earlier bears repeating. Just in case you are dabbling with any hand made capacitors, they are 100% of the time wanting (being extremely polite) in their characteristics.
There are specific capacitor locations that are sensitive to the sound quality in most amplifiers. Others just need to be in good condition. Then there are capacitor types that can make a word of difference. You can actually hear these positive changes, and ... wait for it ... you can measure those changes too.
Keep in mind that I (and most other good technicians) both measure performance, and listen to it. I don't grade my own work, but have others weigh in on their impressions. I never tell them anything, including whether I've changed parts or not. Keeps them honest.
Anyway, back to the main point. Installing parts that do not fit into th eoriginal place is an instant fail on the modification or repair. These jobs are the #1 cause of lifted and broken traces. The work you did in the picture is a great example on how to rweck traces on a PC board.
What is the most expensive component in any device that uses them?
The circuit board. Treat it as if it is the one thing you should not damage - ever.
-Chris
It is important that the replacement part fits properly in the space where the original capacitor did. I can't stress this enough.
As far as testing capacitors go, testing capacitors for dissipation weeds out many bad parts, some of which have big names and are expen$ive audiophile approved ones. Another point made earlier bears repeating. Just in case you are dabbling with any hand made capacitors, they are 100% of the time wanting (being extremely polite) in their characteristics.
There are specific capacitor locations that are sensitive to the sound quality in most amplifiers. Others just need to be in good condition. Then there are capacitor types that can make a word of difference. You can actually hear these positive changes, and ... wait for it ... you can measure those changes too.
Keep in mind that I (and most other good technicians) both measure performance, and listen to it. I don't grade my own work, but have others weigh in on their impressions. I never tell them anything, including whether I've changed parts or not. Keeps them honest.
Anyway, back to the main point. Installing parts that do not fit into th eoriginal place is an instant fail on the modification or repair. These jobs are the #1 cause of lifted and broken traces. The work you did in the picture is a great example on how to rweck traces on a PC board.
What is the most expensive component in any device that uses them?
The circuit board. Treat it as if it is the one thing you should not damage - ever.
-Chris
I am glad to see a comment about damaging the PC board.
Anatech comment is full of good sense and experience.
It really takes a high quality board and a skilled repaireman to change components. A difficult bit is when the track is on the component side, there is a high risk when lifting the component to lift the track too.
So, changing caps over and over again, I cannot beleive the PC board can survive it.
I know what I'm doing . The cap placement looks very bad but there is a trick . After the cap was soldered I applied a special glue on the back leads so they are glued on the front board . If you watch closely the photo you can see that there is glue there . I did this to change those caps on the fly .
Also many times the cap will not fit .....so without special glue on the front board there is great risk of the leads messing the board because they are not fixed .
Exactly! This was pointed out to Robert2017 previously on another forum. His argument as to how replacing these caps altered the sound: Because they connect to other parts and "interact".
Robert2017 should just be ignored. What's annoying about this is having to correct all the misinformation people like him spread. NAD gear is a cut above the standard consumer fair, but swapping out some or all of the electrolytics with audio grade can not cause the changes he claims.
If Black Gates made such a huge improvement, they would still be in production. Every high end company would be buying them. Robert, resistors make a difference too. Time to replace those cheap ones in your NAD with expensive audio resistors?
Have you considered the increase in inductance due to the long leads and the potential for them to act like aerials?
Hi PKam,
Absolutely! But we haven't got through the question of capacitors yet.
Hi robert2017,
Glueing the parts isn't an answer. It also make further service on the unit more difficult to impossible without damage to the PC board. Normal sized parts are available just as high (or higher) quality than the originals. The oversized can also radiate or pick up noise or signals from other parts of the circuit. Highly undesirable, and certainly audible if your listening skills are even half as good as you claim them to be.
Oversized components are a failure. Plain and simple as that. This is also true where components are wired point to point.
-Chris
Absolutely! But we haven't got through the question of capacitors yet.
Hi robert2017,
Glueing the parts isn't an answer. It also make further service on the unit more difficult to impossible without damage to the PC board. Normal sized parts are available just as high (or higher) quality than the originals. The oversized can also radiate or pick up noise or signals from other parts of the circuit. Highly undesirable, and certainly audible if your listening skills are even half as good as you claim them to be.
Oversized components are a failure. Plain and simple as that. This is also true where components are wired point to point.
-Chris
- Status
- Not open for further replies.