Iron-core versus air-core woofer inductor intermod test.

[Deang]

Bob, Hopefully, Al was being somewhat tongue in cheek, you're not. So what is your scientific explanation for those lossy PIO caps not sounding any different to you than the motor runs -- that the ears don't care unless the measureable differences are light years apart? So then, an ALK (or anything else for that matter) built with motor runs or PIOs should sound the same then if built with Solens or Hovlands or whatever.

[Al Klappenberger]

Dean,

I suspect that a chain is only as strong as it's weakest link. The A and AA just isn't worth the fancy caps. My Universal replacement network on the other hand.......

AL K.

[markus111]

----------------

Yes! metal near an air core will mess it. Case in point: hat stupid steel crew on the AA tweeter filter inductor!

Al K.

----------------

Does this include the squawker horn? Is it only ferrous metal that will mess it up? I can remember an article in Speaker Builder where the author took a pair of Advents (?) and unhooked the tweeters from the rest of the network, and unhooked the woofer. He was actually able to drive the tweeter inductor with inductive coupling from the woofer inductor! The unhooked tweeter volume was only 20db down from when it was hooked up! His point being that there is significant coupling present in some production speakers. He modded them by separating the tweeter and woofer sections of the crossovers, and he says that he could tell a difference in the sound. I wish I could find the original article - I'm pulling all of this from memory.

Mark

[BEC]

Dean,

I think there is a threshold on the ESR being something that can be heard. The Jensens come in at around 0.4 ohms. I know that I can hear a bad cap cutting back the highs at around 0.5 to 0.6 ohms on an instantaneous comparison with a switch. This effect is magnified somewhat in the type AA since there is an extra cap in series and the ESRs will add. I usually find the two 2 uf caps in old original AAs to be at around 0.8 to 1.0 ohms each. I am still searching for any two caps that test the same and sound different. I have not found that yet. Unless and until I find that, I will hold that a good cap is a good cap. I will also hold that any listening test of crossover components other than an instantaneous and blind comparison is not a good test.

Bob Crites

[Deang]

"I am still searching for any two caps that test the same and sound different."

Have you looked under the workbench?

Unless and until I find that, I will hold that a good cap is a good cap. I will also hold that any listening test of crossover components other than an instantaneous and blind comparison is not a good test.

A good measuring cap is a "good cap", I'm not disputing that. The switch is cool, but I don't agree it's necessary to make a test valid.

[Al Klappenberger]

Guys,

I think I agree with Bob on this one. I believe that "Q" or ESR is all there is. I don't think there is any magic UNmeasurable quality to a cap that is audable. I know about stuff like linearity and such. I just don't buy it. The Jensen caps, for example, sound "mellow" simply becasue they are lossey and kill the highs. Listening tests are almost always useless unless you can "flip a switch" between two things and have somebody else flip the switch while you listen. The human mind hears what it want to hear. I can't overemphesize this point! This is the very reason I will almost never say that this sounds better than that. I will always site measurements. I simply do NOT believe my own ears.

Al K.

[Deang]

"I think I agree with Bob on this one. I believe that "Q" or ESR is all there is. I don't think there is any magic UNmeasurable quality to a cap that is audable. I know about stuff like linearity and such. I just don't buy it."

But "linearity" is measurable, so why not factor it in? "Capacitors may also change capacitance with applied voltage. This is another major source of non-linearity when building low distortion filters. In the case of some types of audio equipment, capacitor non-linearity in the signal path is the dominant source of distortion." http://www.answers.com/topic/capacitor

I'm not going to pretend I understand all of the implications related to the more complex technical explanations and measurements. I'm clearly out of my realm here, and any attempt to offer up my own technical explanations would be intellectually dishonest. However, I think it needs to be said that there are those just as astute in this field who are clearly in disagreement with you. Maybe instead of saying there is something going on that goes beyond the measurments, I should have said the most used and common methods of measuring do not fully account for all of the perceived differences in sound. Aren't you guys willing to at least admit the possibility of this?

"The Jensen caps, for example, sound "mellow" simply becasue they are lossey and kill the highs." Listening tests are almost always useless unless you can "flip a switch" between two things and have somebody else flip the switch while you listen.

Well then, there is a problem. Bob, using this very method, claims he heard no perceptable difference between the GE cans and the Jensens. Bob says the GE cans have the lowest measurable ESR (and therefore the highest Q) of any capacitor he has measured to date, and the Jensens the highest. So, how do we account for Bob's perception that there is "no difference", and your perception that the Jensens sound "mellow" because of their higher loss factor? Did you listen to them, or simply measure, and because of your findings determine that they would roll off the highs?

"The human mind hears what it want to hear. I can't overemphesize this point! This is the very reason I will almost never say that this sounds better than that. I will always site measurements. I simply do NOT believe my own ears."

I agree with the former, but not the latter. I think maybe you have both unwittingly boxed yourselves in. You say the reason I hear differences is because I've predetermined in my own mind that there are differences. What if I say you have possibly corraled your hearing by predetermining that no measurable difference means no difference -- so you simply don't hear them?

Though it's on a much higher order of magnitude, and so probably not the best example to use, I think what you said about the Trachorn is interesting.

"Here though is an example of an area of acoustics where looking at frequency response curves and distortion measurements tell you very little. While the measurement on the new horn are slightly better then the stock horn the improvement in realism will knock your socks off!"

So, little measurable difference, but big acoustic difference. So Al, do we trust the measurments, or the ears we shouldn't trust? Don't get angy, I clearly understand the acoustic plot comparing the output through the K-400 and Trachorn show much greater differences than the measurable difference of capacitors. I'm just trying to point out that there are some inconsistancies with all of this, and that measurments alone shouldn't be completely substituted for good listening and an open mind.

[Al Klappenberger]

Dean,

A lot of seeming contradictions there!

I have never been able to measure any distortion contributed by a capacitor. I'm quite sure that it's pretty inconsiquential compared to the distortion generated by a moving diapram in a speaker though! It's simply not an issue whereas ESR and Q certianly are. What that does to a filter is VERY measurable.

When the subject gets arond to the acoustics of a horn, I will tiptoe quiely away! I wouldn't know what to measure. Here though, the difference between a good horn and a bad one is will beyond the differences between one capacitor brand and some other. Even a tin-ear could tell the difference!

Al K.

[Arkytype]

Dean, Bob, Al, et al: (sorry, couldn't resist)

Guys,

Thanks for the enlightenment & confusion :>) about an interesting topic--the "sound" of passive components.

The subject of measurement vs. audibility is familiar to many of us whether it involves cables, capacitors, amplifiers or other devices in the signal path.

IMHO, to properly assess the audibility of Component A vs. Component B, a double-blind test using an AB-X switch box is the only "fair" way to eliminate ones prejudices. Not knowing if you are listening to Component A or B (or no change) when you press the button levels the playing field and usually quiets the BSers.

Recently, when I received my Tripathi-based TEAC A-L700P amp a few weeks ago, I was convinced it had deeper bass than my 50 WPC McIntosh amplifier. After manually swapping amps several times it became clear a switch box was needed to make a faster A-B comparison.

I built an A-B box that would allow me to not only switch between two amplifiers, but also had inverting polarity terminals for testing the audibility of absolute polarity.

After using my SPL meter to carefully match the gain of both amps, I proceeded to A-B the TEAC and McIntosh. I was surprised to hear NO audible difference between the two. At first, I thought the box wasn't switching properly. No matter the program source (pink noise to Pink Floyd), both amps (to my ears) were identical. The deeper bass initial impression?---obviously a subjective evaluation on my part. Could it be I wanted to hear deeper bass?

Having satisfied (for now) the issue of amplifier differences, I tried "A-Bing" for the audibility of polarity. The tweeter outputs from my ALK ES-400 networks were fed to the A-B box and the outputs were connected to the Beyma CP-25 tweeters. In position A, the polarity of the signal to the tweeter was normal, and in position B it was inverted. BTW, this test was done using each Klipschorn and then both. Again, pink noise to Pink Floyd and no audible difference.

With the Ivie IE-30 SPL at the listening position and listening to pink noise, I could see (and hear) the comb filtering (from the acoustical combining of the squawkers and tweeters) in the 1/3-octave display. Changing to the opposite polarity only slightly shifted the display. Moving the B&K 4133 mike left and right a foot or so easily duplicated the polarity change seen switching from A to B or B to A.

The SPL meter detected a change, but with the short wavelengths involved, there was probably an infinite family of destructive and constructive interference going on between the two squawkers and tweeters.

The next test was more revealing. Swapping the polarity of the squawkers was plainly audible. Finally, a comparison that could be heard!! Simple physics would demonstrate the wavelength emanating from the squawkers vary from about three feet to three inches. Our ears are far enough apart to make polarity inverting audible.

Everyone has done the next A-B comparison. Swap the polarity of one bass driver while playing source material with deep bass. You should hear a pronounced difference when both drivers are properly connected. This is usually referred to (incorrectly) as "phasing" your speakers.

Getting back to the audibility of Component A vs. Component B: If one accepts the premise that we don't have an "audible" memory, stopping a test to change out a capacitor, inductor or other device would seem to negate the purpose of the test.

If (for example) Capacitor A caused a subtle high frequency roll off and Capacitor B didn't, an A-B test (or A-B-X test) would (or should) readily identify Capacitor A. If, on the other hand, it takes a few minutes to manually swap components, could one hear the same roll off?

If our quest is to realistically re-create in our listening room a live performance, then the efforts of Al, Dean, Bob and many other Forum members will continue to advance the art.

I'm only suggesting that we might need to re-think how we arrive at the perceived audible differences between two components (active or passive) by changing our testing methodology from a subjective/opinionated-based one to a more objective/scientifically-based one.

Lee

P. S. I'm aware my swich box is A-B only and my testing was not douoble-blind. My purpose was to hear differences, not which sounded "better". I'm working on an A-B-X box design.

[Al Klappenberger]

EVERYBODY...

Everyone should read Arkytypes posting above very carefully. The A-B box Lee talks about is no toy! I have pasted up 4 pictures he sent me of it soon after building it. This is what is needed to actually compare two things properly. It will do a lot to keep the listener from fooling himself, which is NOMALLY the case. It's human nature and we all do it. This is the only way to do it properly!

Al K.

[Arkytype]

Al,

Thanks for the kind remarks. The box was constructed from spare parts needing a good application. The simplest version to build is using a 4PDT relay which you control from your listening position with a pushbutton.

If you have a Tripathi-based amplifier, you cannot use a common ground as the outputs are floating. The 4PDT relay allows for switching of the individual left and right channel grounds.

In addition to a simple A-B comparison, the other circuit in my box allows a polarity inverting function with the ability to select a combination of which relays are slaved together. This allows independent polarity swapping of both squawkers or both tweeters.

If there's enough interest, I can post a schematic for the DIYer.

While this is not an A-B-X box as described in my earlier post, it will help you make informed comparisons.

Lee

[Deang]

I had a really cool page and half response typed out in Word, and decided why bother. Instead, I thought it might be more beneficial if I supplied some tools for you guys to help you get up to speed.

http://www.superduperinc.com/F-G_Pages/fd53.htm

[BEC]

----------------

On 6/5/2005 12:09:09 AM DeanG wrote:

I had a really cool page and half response typed out in Word, and decided why bother. Instead, I thought it might be more beneficial if I supplied some tools for you guys to help you get up to speed.

http://www.superduperinc.com/F-G_Pages/fd53.htm

----------------

Dean,

Go ahead and post it. I can probably come up with a good word and half response.

Bob

[scriven]

Finally!

Dean posts a reference that I understand!

[Deang]

LOL. Don't feel bad. I don't understand half of what I post either.

Bob, I'm sorry. But Bullsh!t counts as either one word or two.

I spent the night at work cutting into capacitors in between phone calls. There sure didn't appear to be anything going on between them construction wise that would account for any difference in sound. I did see some interesting things. For example, Solen says they "solder on" their leads. What they actually do is mash the lead into the paste, and then apparently drop a big blob of solder on it. When I pulled on the lead (after removing the epoxy cap), it came right off with the blob of solder attached to the lead. I couldn't help but laugh when it happened. It was the only cap that had any solder at all, the Dayton's and Auricaps were both pressed in. On the metallized film side of things, the Dayton's were obviously the thinnest, and you could actually see through it. The Solens and Auricaps were so close in consistency (the way they "crinkled" and stretched before snapping), that I couldn't make a call on which was thicker. I also read through all of the Madisound threads on the recent double blind cap tests they did. At any rate, all of the rolled metallized types appear to be identical in construction.

[Deang]

I was cleaning up the garage after work yesterday, and when I picked up the loose film laying around, the piece I had cut off from the Solen and was using for comparison turned out to have two layers. It really seemed like a single layer when I was fiddling with it. When I picked it up, I noticed a little bit of separation on the end of the strip, so I peeled them a part. The only thing I can figure is that the high humidity/heat we've been having here allowed the top layer to lift a little. I'm really suprised they didn't separate when I was doing the ever scientific "crinkle test". At any rate, it ends up that thickness of the material is identical to the Dayton's ("see-through" thin).

Friday night I'm going to substitute the two 20uF Auricaps in my ALKs with Solens. With only half the film thickness of an Auricap - I'm only expecting half the sound.

[3dzapper]

Dean, Why don't you cut up a few Teflon caps and compare them for us.

Seriously, I won't chop them up but I can imagine, after your descriptions, that the Kimber caps I just got could be used for window panes. They are that small for the capacitance value.

Rick

[Deang]

Think those are little? Yeah, compared to a film and foil like the Hovland or Theta, they are. Get that same value in a Solen, and you'll need tweezers to pick it up.

[BEC]

A trip down memory lane.

[Daddy Dee]

I remember this thread.