gary f said:
No question about it. I've lashed up highly recommended circuits only to be left wondering 'what could they possibly be thinking?' With time I become more convinced an uncomfortable number in the tube domain equates excess higher harmonic distortion with 'fast sound'. And you can't discount shear enthusiasm of the new. It's the Internet, you expected maybe unalloyed truth and wisdom?
gary f said:
I don't think it is about being honest or not. People who report these things are honest, I'm sure (except maybe a couple
). It's just their perception and perceptions are real.Unless you take seriously the possibility that your perceptions may not jive with the world as objectively experienced (which, I know, is a difficult subject in itself), you will not question your own perception so you make no progress towards "objectivating" your experiences.
Your experiences thus remain totally personal to yourself and are of only very limited value to anyone who IS looking for repeatable, consistent experiences.
Jan Didden
Salas- just letting you know I'm still fooling with the silver vs. brass test. IMO, what I'm testing is very far away from normal cables in terms of dielectric and resistance, and I got distracted tonight making dielectric absorption measurements of various caps. Probably should look at cables as well.
FWIW, I still don't believe caps make much difference inside the feedback loop of an amp. Unfortunately, with few exceptions, all the caps we care about are outside the loop. The typical large value low frequency roll-off cap in the feedback divider will affect amp performance- any error in the divider is an error in the amp output. Any caps prior to the base of the first diff amp transistor are outside the loop- including the RF suppression cap. Obviously, all coupling caps are outside the loop. I don't yet suscribe to the "PS caps are part of the signal path" philosophy, so have no problem with electrolytics there, plus any bypass that presents a low impedance at HF, regardless of whether the phase angle suggests it's an inductor or not.
If DA is a problem, there are no suitable large value choices. I used to think Oscons were extremely good- ultra low dissipation. Unfortunately, they have the worst DA of any electrolytic I measured. No idea about Black Gates, as I don't have any to test. For up to 10uF or so, polystyrene can't be beat. Not perfect, but very close. Polyester (Mylar) is so-so. Polypropylene is much better, but not as good as styrene. If that's all one can get, is it good enough? Silver-mica is not very good. Ceramics are worse- didn't test any NPOs, but I suspect they suffer from DA too. No surprises here- my results are the same as everybody else who's tested this.
The big question is how serious is DA, and it probably depends on just where in the circuit it is, and what the surrounding impedances are. I remain skeptical, but an awful lot of people seem to correlate this with sound quality. I have a lot of listening still to do, along with silver vs. brass!
FWIW, I still don't believe caps make much difference inside the feedback loop of an amp. Unfortunately, with few exceptions, all the caps we care about are outside the loop. The typical large value low frequency roll-off cap in the feedback divider will affect amp performance- any error in the divider is an error in the amp output. Any caps prior to the base of the first diff amp transistor are outside the loop- including the RF suppression cap. Obviously, all coupling caps are outside the loop. I don't yet suscribe to the "PS caps are part of the signal path" philosophy, so have no problem with electrolytics there, plus any bypass that presents a low impedance at HF, regardless of whether the phase angle suggests it's an inductor or not.
If DA is a problem, there are no suitable large value choices. I used to think Oscons were extremely good- ultra low dissipation. Unfortunately, they have the worst DA of any electrolytic I measured. No idea about Black Gates, as I don't have any to test. For up to 10uF or so, polystyrene can't be beat. Not perfect, but very close. Polyester (Mylar) is so-so. Polypropylene is much better, but not as good as styrene. If that's all one can get, is it good enough? Silver-mica is not very good. Ceramics are worse- didn't test any NPOs, but I suspect they suffer from DA too. No surprises here- my results are the same as everybody else who's tested this.
The big question is how serious is DA, and it probably depends on just where in the circuit it is, and what the surrounding impedances are. I remain skeptical, but an awful lot of people seem to correlate this with sound quality. I have a lot of listening still to do, along with silver vs. brass!
This wasn't a terribly rigorous test, but each cap was attached to the input of a Keithley electrometer and charged up to 10V for a minute. It was then shorted out for about 200mS and the voltage rise read after 30 seconds. All the caps were small radial electrolytics between 47 and 330 uF, 25-100V. I tested old and new Panasonics, Illinois, Siemans, Rubycon, Nichicon, and an Nbccon photoflash cap, just for the heck of it. There were 85C units, 105C units, and unmarked ones from an old Radio Shack assortment. There wasn't much difference between most of them. Voltage rise was to about 3%, with the exception of the Rubycon, which was about 2%. The photoflash cap was closer to 4%, but the big loser was a Rubycon non-polar cap. It went well over 10%. I also had a small oil filled can of unknown history. It was essentially perfect, as was an Italian motor capacitor, probably polypropylene. If I were stuffing a board and needed a small electrolytic, I'd probably grab the Rubycon first, and given a choice I'd tend towards the newer parts, rather than caps that had been sitting around for years- better DF, if not DA. Now, most of these parts are typical of what commercial equipment manufacturers use, better than you'd find in consumer products. Still, nothing to write home about. Any thoughts as to what one might choose from, say, the Digikey or other catalog, that would be any better?
Andrew- the focus here was small electrolytics, though I did toss in a couple films that were sitting on the bench. Next I'll gather a group of films and compare those. I don't expect to see anything different than everybody else who's tested them Now, I do have some parts that aren't commercially available- 15uF polystyrene mil surplus parts, some nice 0.47uF polystyrenes, and more easily available, some low value Illinois polypropylene units for HV/HC use. I don't have many polypropylenes, as I've always been able to find styrenes, and not all that many NPOs, but I'll search my stock.
Now, I do have some parts that aren't commercially available- 15uF polystyrene mil surplus parts, some nice 0.47uF polystyrenes, and more easily available, some low value Illinois polypropylene units for HV/HC use. I don't have many polypropylenes, as I've always been able to find styrenes, and not all that many NPOs, but I'll search my stock.Didn't get much done last night, as both power and phone were out due to a snow/slush storm. Went through a bunch of caps this morning and suspect one can group things into a few simple groups, not including electrolytics. Worst is silver mica. In spite of it's excellent DF, DA is lousy. Next is old polyester like the ubiquitous CDE "WMF" caps. If DA really matters, these are to be avoided. A bit better is what I'd call Japanese mystery film. It's probably a better grade of polyester or close cousin, but my data sheets aren't specific. I was particularly interested in the Panasonic ECQ series, as I have a lot of them, but they're sort of middle of the pack. Polycarbonate is only slightly better. Then we have a large jump to polypropylene, which is very good, but not as good as polystyrene, which is about a factor of two better yet. I didn't have any NPOs that were large enough to be useful, or to get good numbers out of without messing around with shielded enclosures and such. The numbers here can't be compared to anything else, as my test doesn't adhere to any standard, but the relative performance of each cap can be compared. My assumption is that the value of the cap doesn't have much to do with this, as it's more a material property. That tends to hold true for DF as well. I only labeled the caps I could confirm with data sheets, but it's easy to guess the rest. Now the question remains, do these numbers have anything to do with sound quality?
ITW 3.0uF "MP" 0.02%
Sprague 1.0uF "431P" 0.19%
Sprague 6.0uF "430P" 0.14%
CDE 2.0uF "WMF" 0.40% polyester
Illinois 0.47uF "PPA" 0.04% polypropylene
Electrocube 15uF "230BIE" 0.24% polyester
ASC 3.0uF "AYC2" 0.018%
EC 0.47uF "ST12" 0.010% polystyrene
Electrocube 5uF "650BIC" 0.09% polycarbonate
Shizuki 1uF 0.13%
Panasonic 2.2uF "ECQ-E" 0.13% "high dielectric plastic film"
Panasonic 0.47uF "ECQ-V" 0.16% "plastic film"
Sangamo 68000pF S.M. 0.50% dipped silver mica
GR 0.2uF S.M. std. 1.4% stacked silver mica
Illinois 10uF "MW-R" 0.35% polyester
Yellow block 3.3uF 0.26%
Corning glass tempco exceeds DA
ITW 3.0uF "MP" 0.02%
Sprague 1.0uF "431P" 0.19%
Sprague 6.0uF "430P" 0.14%
CDE 2.0uF "WMF" 0.40% polyester
Illinois 0.47uF "PPA" 0.04% polypropylene
Electrocube 15uF "230BIE" 0.24% polyester
ASC 3.0uF "AYC2" 0.018%
EC 0.47uF "ST12" 0.010% polystyrene
Electrocube 5uF "650BIC" 0.09% polycarbonate
Shizuki 1uF 0.13%
Panasonic 2.2uF "ECQ-E" 0.13% "high dielectric plastic film"
Panasonic 0.47uF "ECQ-V" 0.16% "plastic film"
Sangamo 68000pF S.M. 0.50% dipped silver mica
GR 0.2uF S.M. std. 1.4% stacked silver mica
Illinois 10uF "MW-R" 0.35% polyester
Yellow block 3.3uF 0.26%
Corning glass tempco exceeds DA
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.