On 07-27-09, Volleyguy wrote:
"Tempo Electric has changed their ratings based on what has been said here by many of us..."
I just want to de-lurk to belatedly clarify this point. We didn't actually change the ratings, themselves, but we did change the way the rating tiers were labeled.
Based on comments earlier in this thread, there was an implication that caps on Tier D were grossly inferior to the ones higher up, when they are really good enough for general use by the typical DIYer. Not everyone can spend the sort of money that Duelend or VCap Teflons, among others, cost, so the vast majority of users will probably find caps on this lower tier to be the most affordable for their projects.
To remove the possible stigma, we renamed the top three tiers AAA, AA, and A, respectably, since that's where the really outstanding performers ended up. We then changed the name of the fourth level from D to B. The differences among the first three tiers were often subtle and you really need to be at the outer fringes of perfectionist audio in both equipment and listening skills (I modestly include my partner, myself, and our listening panel in that category) to appreciate the improvements that these caps can provide. Generally speaking, you also need deeper pockets and a higher level of obsession than the typical DIYer may have, as well.
In addition, for the vast majority of listeners, the top tier caps may be financial overkill. If you can live with the trade-offs mentioned on Tier B (formerly Tier D), then it's possible to elevate mundane equipment to something far better without busting the bank.
A point which many DIYers may overlook, is that by carefully matching inexpensive caps that otherwise didn't fare so well when compared to the very best, it's still possible to achieve a sound that's surprisingly good. For example, in his personal phono stage, my audio partner, Dr. Arthur Loesch, paired the hard sounding Electrocube with the soft sounding Vitamin Q and achieved surprisingly good results. It's not the way I'd prefer to do it, but the sound was far better than the relatively low cost (or individual flaws) might suggest. In the production version of the same circuit, we had to spend 10 times as much to hear a 5% improvement (advancing from 95 to 100), but that's what perfectionist audio is all about.
As for the next round of comprehensive tests, Arthur and I are working on a set of aftermarket cross-overs for the Tannoy Monitor Gold drivers, so we will eventually revisit the speaker caps which were overlooked or unavailable earlier, including the Duelend CAST PIO series. We're also working on a couple of new electronic designs, where overlooked lower value brands will be compared with the best of the earlier rounds. Given the snail's pace at which things happen around here, this could easily be as long as a year from now (September, 2010).
For those who are curious, we agree that the Dueleund resistors are the best we've heard to date for use in loudspeaker x-overs (we rate Caddock MS and MV series as the next in line), but we're also interested in trying Intact Audio's (Dave Slagle's) latest type which are custom wound with copper magnet wire, based on the desired resistance at a given temperature.
http://www.intactaudio.com/res.html
Thanks! And now back to our regularly scheduled discussion thread...
//Joseph Levy |
Thanks for your comments!
I don't want to beat this topic to death, but the system used in labeling each level will always carry some sort of psychological baggage with it.
As an editor, I never cared for the original A, B, C, and so on, because of the association with the grading system used in many schools where grades C and below are considered failing (or close to failing). Though they are not the very best in an absolute sense, for the average DIYer, who may otherwise be clueless, the caps on the tiers formerly called C and D are far from failures. You'd be surprised (well, maybe not) at the number of folks who still think that all passive parts sound alike and for whom spending more than $2 on a capacitor is stretching the budget.
The idea was to give the results some sort of pecking order, to clearly highlight the very best that we heard, and to try and avoid stigmatizing some otherwise decent products. Many listeners in our audience simply don't share the level of obsession that you (Volleyguy) and I (we) have, so I wanted to cut the typical reader some slack and suggest a range of alternatives to "cost no object" parts without completely denigrating their value. Hence, the decision to relabel the tiers. Perhaps Greek letters are the answer.
Also, in our designs, we usually mix caps (as well as comparable resistors) from the top three levels in order to cancel out any sonic signature (and they all have one). An argument could be made, however, that Duelund components are the exception to this rule, since (so far) no one has been unable to detect any audible signature worth mentioning.
In the future, if (for example) we feel that the Duelund CAST PIO blows the VSF out of the water, we're more likely to drop the VSF down to AA, and so on, rather than create a AAAA tier. But who knows? Maybe a class by itself is where the CAST belongs. We have some very interesting candidates lined up for the next round, including the Audio Note Silvers. I'd love to have some samples of the Duelund and Jensen silvers for comparison, too.
BTW, I still haven't read more than 10% of this thread, but will peek in from time to time to get a sense of what new discoveries may surface. Apologies if I don't respond or comment for awhile, even months. There are just not enough hours in the day!
Happy Listening,
Joe |
There's quite a lot to address here, so I may just comment on one issue at a time.
In terms of materials, all caps contain two layers of foil, which can be silver, copper, tin, or aluminum. That's the material which is electrically charged. The insulating material, which separates the foil layers, can be paper (treated with wax, oil, or other substances), plastic (such as Teflon, polystyrene, polypropylene, mylar, or polyester), something more exotic (silk), or some combination of the others. More expensive caps use individual, interleaved (or inter-wound) layers of conductors and insulators. Those are the "film-and foils." The less expensive ones spray or coat the conductive material onto the insulator. Those are called "metalized."
So, because they perform totally opposite functions, foil vs. plastic really isn't the correct vocabulary. What we're looking at is a wide mix of material combinations and fabrication techniques, some of which sound better than others when used in perfectionist audio components.
It's only been in the past 30 years that the concept of capacitors having distinct sound characteristics has come about. Traditionally, the best insulators, like Teflon and polypropylene, have been considered the best sounding, but they are also the most difficult and expensive to wind. Exotic combos, like the Jupiter or Duelunds are very recent developments.
What sets the Duelunds apart is their completely different approach, both in materials and fabrication. For the curious, there are papers by both Steen Duelund and Harvey Rosenberg (who was an early champion of Mr. Duelund's research) which can be found online. They go into much more detail about the theory behind his approach than one can easily summarize.
The other big variable is high voltage vs. passive circuit applications. Our experiments showed that the same cap used in high voltage electronics often sound very different when used in a passive speaker crossover (and vise versa). Also, putting them together in different combinations in the same circuit can greatly alter the end result. Throw in all of the other parts that make up a circuit (tubes, transistors, resistors, inductors, transformers, copper or silver wire, solder, and so on) and there's no end to it.
Basically, we found that certain caps (as well as other components) consistently sounded better when used in certain positions, combinations, or passive vs. active circuits. The Duelund caps and resistors are certainly at the top of the list, but others worked well, too. The issue of resonance in a capacitor is also worth considering and this may be a characteristic of how tightly a cap is wound, as well as the type and thickness of the insulating material. This may be why the "new" VCap TFTF sounds better than the original -- the Teflon insulation is substantially thicker.
For a long time we thought that foil-type inductor coils, such as those made by Alpha-Core, were the best to use. However, Dave Slagle pointed out to us that, abstractly speaking, these were built like capacitors (layers of foil wrapped with concentric layers of plastic insulation) and that we should stick to wire-wound inductor coils, like the Solen air-core. I actually did a direct comparison of Duelund's original wire-wound, iron-core, toroidal inductor with the Alpha-Core and preferred the Alpha-Core. This just shows that the ultimate proof is in the listening. Along those lines, my partner, Dr. Loesch, has historically preferred oil caps, but in our tests none of them really sounded as good as the film-and-foils.
My point is that there's really no "evil" material and that the reasons why we prefer one material or fabrication technique over another cannot always be applied as a general rule. To give you a more specific answer, as of this writing, I can't say that plastic, per se, has a sound that we can personally distinguish as plastic. Of course, we've never tried the Duelund caps or resistors in our electronics (the high voltage Duelunds weren't on the market when we were voicing our amp and preamp), so that view may change when we do. |
Further thoughts:
While it's easy to become overly focused on one component or another, the audio system always needs to be considered as a unified whole. From the time the stylus enters the groove or the laser beam hits the pits, until the signal exits the speaker, there needs to be a philosophically balanced approach to the entire signal path. Until you've addressed the basics of every component in the audio stream in turn, it's often difficult to single out one factor or another in considering which upgrade will make the greatest sonic difference.
In other words, if you start with a flawed circuit design or mis-matched components, then even the best parts can never make up for this. You need to continually re-evaluate your overall strategy. BTW, I'm speaking abstractly here, so don't take this as being critical of your or anyone else's choices.
If you're starting from scratch, there are some premises which will set on the right track to begin with. For example, tubed phono tends to sound better than solid state, but there's a trade-off in the form of tube noise. In amplifiers, push-pull amps that use EL84 tubes and 5AR4/GZ34 rectifiers are all generally musical and easy to modify. Ditto with 2A3 or 300B SET amps. If your budget is limited, there are many $200-$500 sleepers out there, as well, including the Dynaco PAS-3, ST-35, plus any number of Fisher and Scott amps (look for the ones which use EL84's). For under 10% of the price, Tannoy Monitor series drivers from the 1960s-80s in custom cabinets can equal or better any $50K speaker on the market today.
Get the basics right and it's easy after that. Even without tweaks, you'll still have music. |
Volleyguy wrote, "I had noted that even one poly cap in the crossover and the sound changed to plastic..."
Let me elaborate further. As mentioned above, to talk about "foil caps" vs. "poly caps" is to use a mixed metaphor. That is to say, all poly caps contain foil, so it's not an accurate way to differentiate them. Polypropylene and polystyrene, among other plastics, are dielectrics (insulators) which actually sound different from each other, while foil (which can be any number of metals) is a conductor. What you probably meant to say is that your argument is about the use of a paper dielectric vs. a plastic dielectric.
However, if you look at our survey, there are paper caps and plastic caps which both rate high and low. While the Duelund (paper) scores at the top, the Jupiter, also paper, does not. The same is true of the VCap TFTF (plastic), which scores at the top, and the CRC Teflon, which scores quite low. While the dielectric may be the same in each pairing, the sound is quite different. Note, however, that the Jupiter uses aluminum foil, which is also used in a few others that scored much lower, like the Jensen aluminum. Rather than single out a specific element for cheers or jeers, the difference is often in the way that element is used as a part of the whole.
What we were more concerned with, and which we were listening for, were issues like veiling (did it seem like a curtain, however slight, was separating us from the musicians), air (could we hear the space between the performers or envision the performance space, itself), detail (how clear were words, how closely could we follow individual instruments or voices), glare (did the extreme treble become shrill or congealed), tone (did violins sound like violins, pianos like pianos), and so on. That's the sort of vocabulary we used in our evaluations.
The question always was did the listening experience sound like live music or was something in the way. The concept of a plastic or paper sound was never mentioned and never considered as an issue.
|
I'm starting to read some of the recent posts and see that many of my comments have already been covered to one degree or another. For example, the link between Steen Duelund and Harvey Rosenberg, the Fisher EL84 amp, and so on. I was also going to comment that larger caps sound better, but that's been mentioned, too.
A few more random notes:
You're right about the cheap parts in most speaker x-overs. However, keep in mind that most x-overs have too many parts to begin with, so it's no wonder that cheaper parts are common. Commercial products are built to a price point and that's always been a problem for the audio perfectionist.
Regarding your EL84 amp, I'm not sure exactly what was done or who did it, but we spent a long time (years) mixing and matching different parts, including tubes, caps, and resistors, before hitting on the magic combination. Regardless of what we personally liked, to get the sound you love requires going through this type of round-robin on your own while listening in your own system. In order to understand how different parts affect the sound, I always caution people against making permanent changes without doing a personal comparison first.
As the amps and other components in our system became more refined, we went through these comparisons at least three times for every component. We typically made only one change at a time, which is why it took about eight years to feel that everything was fundamentally right. That was two years ago and we're still making improvements, mainly in the front end.
Also, running pentodes as triodes doesn't really change the sound that much, it mainly reduces the power. If you want a true triode sound, then it's best to use a true triode amp. When properly implemented, the EL84 is still the best of the pentodes and a good, musical performer. It will benefit from a tube rectifier, however, such as the 5AR4/GZ34. I recently bought an Olson Classic 40 amp (GZ34 + EL84's) for $200 and even with the crappy mods that the previous owner did, it still sounds musical.
|
Volleyguy wrote: "It is those very Teflons that I am looking at... I am keeping the [Duelunds in the] speakers but am not sure about the amp."
When we ran most of the tests, the VSF was the only Duelund cap available and its size made it impractical to audition in our electronics. At some future time, I'd like to acquire small values of the newer Duelunds and see how they sound in active circuits, too.
"As far as for inductors would not foil resonate much more than wire if the insulation and structure could not hold the foil dead still?"
One would think that, but the proof is in the listening. That's why I always say keep an open mind and forget about preconceptions. Duelund abandoned the wire-wound, iron-core inductors in favor of foil, which they presumably think sounds better. Note that their foil inductors are highly damped and not loosely wound, though that's the way capacitors are fabricated, too. I'm referring back to Dave Slagle's argument that foil inductors that are built like capacitors should not sound as good as air-core, wire-wound inductors.
"Where the oil caps tested in a all oil circuit?"
We never tried that, but it's another variation that someone could explore.
"Steen did feel that he could hear plastic in the cap?"
That may be, but it never crossed our minds or came up as a topic for discussion. In other words, we never detected a "plastic" sound, just specific differences between different caps.
"I often wonder if SET movement has not been aided by the internet."
In the 1970s and 80s, when the commercial American audio scene evolved into acoustic suspension speakers driven by high powered solid state amps (like the Dyna 400, the Ampzilla, or the Phase Linears) and high-powered pentode amps (conrad-johnson and Audio Research), the SET movement was thriving in Japan. The Internet certainly accelerated the trend, though small journals like Sound Practices were important, too.
"Only in a non commercial enviroment can one find out that vintage Tannoy's can compete with speakers costing 50k."
What excites me about the Tannoy x-over project is that there is only one small value cap (about 1.5uF) directly in the signal path. This will open up a much wider range of candidates to test in speakers, including some of the silver foils which we just could not afford to purchase in a 10uF value. This should allow us to compare the various Duelunds, Audio Notes, Jensens, VCaps, REL Caps, and others, in a way that we were unable to do before.
Joe
|
It's a 15" 3859 dual-concentric which, in the mid-1980s, was originally sold in either the Panther or Cougar stage monitor cabinet. A variation was briefly (and unsuccessfully) marketed for home audio and my partner has come up with his own design, loosely based on the Monitor Gold x-overs.
The actual value in question can vary from 1.5uF to 2.2uF. It's run in parallel with a resistor and I believe that by varying the value of the resistor we can get the cap down to 1.5uF. While there is a resemblance in principle, it differs somewhat from the published Tannoy x-overs of that period.
Ultimately, we'll be applying what we learn to the Monitor Golds, as well.
Joe
|
Just a minor revision. Taking a look at the schematic for the proposed Tannoy x-over, there's also a 10uF value at the beginning of the signal path, along with a 5.6uF and 20uF across the HF and LF sections, respectively.
The plan is the use the best we have on hand to start with, which includes the 10uF VSF, and continue from there. The small 1.5uF (or so) value will give us a chance to try a wider range of products without breaking the bank, after which we'll have the option of upgrading the larger values with the top performers to further judge the overall effect.
Joe
|
