my 2pence...i may as well join in!
lets get rolling then
what are the issues with caps?
1: microphony
2: parasitic inductance
3: dielectric loss & polarisation of dielectric
4: anything else you can think of...maybe cost?
how do you solve these issues?
1: epoxied caps/potted caps WHEN glued VERY well to crossover board, SHOULD be far ,less microphonic than say electrolytics, or spindly ceramics.
2: avoid WOUND caps, unfortunately that means many polyproplyene caps, which is a shame as the dielectric is one of the best
3: polyester and polypropylene are pretty comparable. I cant remember off hand, but polyester MAY be slightly better. Although MOST speakers are only LV, unless you are using ESLs, dielectric losses/polarisation may not seem very important. however the polarisation/charge 'memory' worsens with lower dielectric loss. so the perfect dielectric MAY cause the worst 'blurring'...that is IF it even occurs in the sub kilovolt range, which i doubt it does to a significant degree.
4: Price? cheap as possible given points 1,2 and 3 have been considered, AND in that order od importance. good thick leadouts...well more than SOME caps use at least...say 1mm?
My personal conclusion: high voltage polyester film caps with epoxied body. Cheap, solid, easily attached to things, and comparable dielectric performance to PP, without the parasitic inductance some 'hi end' wound pp caps suffer. the only downside to these caps are the leads which are short and normally designed for PCB mounting.
IF ONLY I COULD GET SOME CROSSOVER SIZED POLYSTYRENE CAPS...THEN ID BE REALLY CHUFFED
lets get rolling thenwhat are the issues with caps?
1: microphony
2: parasitic inductance
3: dielectric loss & polarisation of dielectric
4: anything else you can think of...maybe cost?
how do you solve these issues?
1: epoxied caps/potted caps WHEN glued VERY well to crossover board, SHOULD be far ,less microphonic than say electrolytics, or spindly ceramics.
2: avoid WOUND caps, unfortunately that means many polyproplyene caps, which is a shame as the dielectric is one of the best
3: polyester and polypropylene are pretty comparable. I cant remember off hand, but polyester MAY be slightly better. Although MOST speakers are only LV, unless you are using ESLs, dielectric losses/polarisation may not seem very important. however the polarisation/charge 'memory' worsens with lower dielectric loss. so the perfect dielectric MAY cause the worst 'blurring'...that is IF it even occurs in the sub kilovolt range, which i doubt it does to a significant degree.
4: Price? cheap as possible given points 1,2 and 3 have been considered, AND in that order od importance. good thick leadouts...well more than SOME caps use at least...say 1mm?
My personal conclusion: high voltage polyester film caps with epoxied body. Cheap, solid, easily attached to things, and comparable dielectric performance to PP, without the parasitic inductance some 'hi end' wound pp caps suffer. the only downside to these caps are the leads which are short and normally designed for PCB mounting.
IF ONLY I COULD GET SOME CROSSOVER SIZED POLYSTYRENE CAPS...THEN ID BE REALLY CHUFFED
Last edited:
If you refer to the tracks on the testing disc, no- they are pure .wav from a cd. I feel like if I alter the levels of the test tracks, they might lose some dynamic range or something else, just because I may have an inferior burning program. I'd really hate to introduce another variable....
Later,
Wolf
Wolf,
The level matching I refer to is that of the various signal paths through the x-over circuit.
Where equivalent (the same) capacitance under test has different ESR....resulting in different output level. This was hinted at earlier in this thread. It is one of the possible "weaknesses" of the test that skeptics will point to.
On Wolf's behalf, an earlier (different) cap test was performed. He was one of the respondents in the room who could hear...what was it? 7 or 8 out of 10?
The level matching I refer to is that of the various signal paths through the x-over circuit.
Where equivalent (the same) capacitance under test has different ESR....resulting in different output level. This was hinted at earlier in this thread. It is one of the possible "weaknesses" of the test that skeptics will point to.
On Wolf's behalf, an earlier (different) cap test was performed. He was one of the respondents in the room who could hear...what was it? 7 or 8 out of 10?
Hi Ed,
Not wanting to rain on your parade, but if you have enough people in that room there's a good chance someone may score 9 or even 10 out of 10. Even when double blind, and still it might not proof anything with reasonable confidence.
It's all about statistics, and depends on what the others scored as well, I'm afraid.
jan didden
There will without a doubt be individual differences in aural acuity allowing some individuals to reliably differentiate caps while others cannot. In this arena, as in many, average results cannot be generalized to any particular individual. However, average results regarding preferences for sound, color, texture, etc., may be strong enough, in some cases, to guide a manufacturer's choice of a product component.
Some of the earliest preference research out there demonstrated that the magnitude of differences in strength, temperature, and sweetness of tea determined preferences for tea. However, generalities about preferences, beyond what contibuted could not be made.
Some of the earliest preference research out there demonstrated that the magnitude of differences in strength, temperature, and sweetness of tea determined preferences for tea. However, generalities about preferences, beyond what contibuted could not be made.
I don't know, Ed. I never tested them for ESR to start with. The fact was the rated values were the same. If you are getting into ESR differences, that is more to how a cap is constructed, like the Reliable/MultiCap PPMFX for example compared to a Bennic. For this test, the capacitance values being the same is the determining factor, because the ESR is a difference to a line of products over another. IMO, there is really no way to match those, and still do this test as I have it configured.
As to the level interest of cause by a former contributor in this thread, it's not level differences that I am hearing, so it should not become a factor at all.
As to the previous tests I've participated in, I took a wire test (2003), and could easily identify which was which, much to the preplexion of the test's designer.
I took a cap-test with them a year later (2004), and didn't really test like I had hoped I would, but I was still in training and learning for my ears at that time. While I thought I could tell a difference between the caps at that time, the duration (0:30) of the test made it harder and harder to differentiate the 2 involved in the test due to listeners fatigue.
Out of ten selections I actually guessed 2 right, and 8 wrong at that time. The 2 were later in the test positioned about 2nd and 3rd from last selection. But the funny part is that another already trained participant had the exact opposite answers for all the selections. This does lead me to believe that did hear something that allowed me personally to differentiate 8 times outta 10, but I did not know what was what at that time as far as what parts contribute what character to the sound.
At the time of this previous test, I felt that mylars smoothed over nuances a bit for a more smooth and less bright sound, while the Sonicaps on the other hand sounded almost gritty they were so bright and detailed, but even then a bit sibilant. I actually thought I picked the mylar as my 'preference' portion of the test, when in fact I actually picked the Sonicap. This led me to believe I had swapped their sonic characters in the test, and actually just listened at it backwards. If this makes sense to you, you can see how I can look at it like pretty much getting 8 right out of 10.
Again- that was 6 years ago, and a lot earlier in my hobby. I play a lot with what caps I actually have in the test xover, and I do believe that a few really stand out as being different in character from the norm or the remainder.
Later,
Wolf
As to the level interest of cause by a former contributor in this thread, it's not level differences that I am hearing, so it should not become a factor at all.
As to the previous tests I've participated in, I took a wire test (2003), and could easily identify which was which, much to the preplexion of the test's designer.
I took a cap-test with them a year later (2004), and didn't really test like I had hoped I would, but I was still in training and learning for my ears at that time. While I thought I could tell a difference between the caps at that time, the duration (0:30) of the test made it harder and harder to differentiate the 2 involved in the test due to listeners fatigue.
Out of ten selections I actually guessed 2 right, and 8 wrong at that time. The 2 were later in the test positioned about 2nd and 3rd from last selection. But the funny part is that another already trained participant had the exact opposite answers for all the selections. This does lead me to believe that did hear something that allowed me personally to differentiate 8 times outta 10, but I did not know what was what at that time as far as what parts contribute what character to the sound.
At the time of this previous test, I felt that mylars smoothed over nuances a bit for a more smooth and less bright sound, while the Sonicaps on the other hand sounded almost gritty they were so bright and detailed, but even then a bit sibilant. I actually thought I picked the mylar as my 'preference' portion of the test, when in fact I actually picked the Sonicap. This led me to believe I had swapped their sonic characters in the test, and actually just listened at it backwards. If this makes sense to you, you can see how I can look at it like pretty much getting 8 right out of 10.
Again- that was 6 years ago, and a lot earlier in my hobby. I play a lot with what caps I actually have in the test xover, and I do believe that a few really stand out as being different in character from the norm or the remainder.
Later,
Wolf
Wolf, we found almost 30 years ago that polypropylene sounded more clear that Mylar in speaker tests at Dalquist, etc. In fact, someone connected with the company got mad at me for telling Walt Jung of the results, at the time. We all switched over to polypropylene for coupling caps, 25 years ago, with serious speaker designs. Mylar was found, later, to have perhaps 10 times the dielectric absorption as polypropylene and some Mylar caps had high Q ringing in the audio band. Look at the cap comparison tests done by HFN in 1985, by Martin Colloms to see measurements, if you are interested. No wonder, differences can be heard.
I benefit from all of your comments
statistical variation & spread
aural acuity (I like that one...wish I had more) I was present for the cap test 6 years ago as well. On first listen I didn't notice anything, so I blew it off...maybe too quickly
real experiences & a faithful report
clarification wrg ESR
Grinnell, Iowa
Saturday, October 16th
9:00 am,
be there
statistical variation & spread
aural acuity (I like that one...wish I had more) I was present for the cap test 6 years ago as well. On first listen I didn't notice anything, so I blew it off...maybe too quickly
real experiences & a faithful report
clarification wrg ESR
for some, sadly, yes...I look forward to the cap test, to informing myself for how my aural physical plant works in relation to others. I may be in the category that needs extended exposure to learn of the variances present, if my ears are up to the task.
Grinnell, Iowa
Saturday, October 16th
9:00 am,
be there
Last edited:
Did you have links to that test referencing that test, I know the answer is probably 2003 but Im curous only because I have put up great amount of $$$ in tests during my University life (me and 3 Engineers live together and we loved audiophile money
) I also have about $4K in wires/cables/etc that I have used to measure and test. I have yet to find a person that can actually pick out any differences during normal content playback. Im always interested in those who can actually indentify differences then the 99.9% that can not.No need to discuss in this thread further, Im just curious if there is a reference to the test.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.