Audio Asylum Thread Printer
Get a view of an entire thread on one page
|For Sale Ads|
In Reply to: RE: DC LINK capacitors in Speaker Crossovers posted by airtime on March 11, 2017 at 09:05:54
Most people are about two to three decades behind, in capacitors for audio. ( See next paragraph ). The DC LINK cap will filter AC and DC, just like any film cap on the list you referred to.
There are a new wave of HIGH performing industrial-use film caps, often meant for pulse duties, that I have used in amplifier power supplies the last 30 months, with wonderful results.
I am just asking if anyone has used one type ( DC LINKS ) in speaker crossovers. When you read the specs on them, they exhibit low ESR, and have high peak current capabilities, which are good characteristics in a film cap, that needs to do audio.
With my post, I am trying to find out, who in the audio universe, has taken the time, had the imagination, to experiment with same, and WHAT are their impressions, results.
Thank for asking, lets see who has done what with DC LINKS ! Will I be the first on my block ????
I say give it a try. That's the only way DIYers can find out if these actually work out. But PLEASE do let us know what how it works out.
I've had a number of things that looked good on paper not work out in real world applications. But you never know???? But there are things that shouldn't have worked and did. Like the Dynaco A-25 speaker. And they've STILL selling them over 60 years now.
Off the subject a bit but why shouldn't the Dynaco A25 speaker not have worked? It did as you seem to admit and if some theory said it shouldn't have, the theory had to be incorrect.
The theoretical problem is in the power response. A two-way with a 10" driver covering the midrange needs directivity control to have a reasonably uniform power response. I know the crossover point for the A-25XL was 1500 Hz, and I think the A25 might be higher but I'm not sure. That's about an octave above the point where 10" cones start becoming directional. The result is a dip in the upper midrange in the power response which gives the speaker a mellow and laid back character. The large-ish tweeter was also directional at the top of its range, so I would imagine it lacked energy in the top two octaves. I've not heard the original A25, only a Seas kit.
By late 1960s standards, it may have been a very fine speaker. But by modern standards, it's a stupid design and I don't know why people keep cloning it. Nostalgia I guess.
The A-25 is hardly a stupid design. I bought a pair of the original design in 1970, and the A-25 sounded great then and still does. This is not nostalgia since I have a second pair of original A-25s that I bought on EBay 20 years ago and I listen to them all the time in my video system. Maybe the later versions didn't sound as good. I don't know because I never heard them, but don't dismiss the original.
FWIW, I tried the Wima DC Link caps in my electronics and did not like them at all.
OK, I take back the stupid remark. However, there is a good reason why you don't see 10" two-ways anymore except PA speakers with horn tweets.
Loudspeaker design has advanced a lot in the last 50 years and I don't see any reason to clone such an outdated design unless except for nostalgic reasons. The Seas kit is rather expensive too.
Was this in power supply applications, or as coupling caps?
Power supply. The Wima caps were cold and unmusical even after lots of hours.
I would have asked you what you like better, but see in your post to drlowmu that you use Black Gate in the power supply. Since they're unobtainium , as Eli Duttman likes to say, I guess I'll still be looking for Plan B .
Thanks again for your reply. I thought about you yesterday, while warming up my Last PAS.
The Blackgate WKZ caps still reign supreme as far as I am concerned. I am hoping that Audio Note will soon release its replacement for the WKZ caps and that it will sound as good as if not better than the originals. The Audio Note Kaisei caps are ok but not a real alternative to the WKZ in my opinion.
In the meantime I have used Jensen radial 100uf+100uf caps in a couple projects with decent results. It is a distant second to the WKZ caps but if I can't use a WKZ the Jensens are my first choice. There are many folks who like ASC oil caps for power supplies. They just haven't worked out for me at all but for those who do like the ASC sound, I recommend GE 97F oil caps instead since they are less colored than the ASC.
The only Blackgates I ever purchased are FK 1000 µF 50V, for the power supply of a McCormack SS phono stage I was modifying. They were too large to fit, so I had to go with something else which would.
I take that back, since I also bought a couple of the 4.7 µF bi-polars everyone was raving about. They never got used either. After seeing what these things now sell for, maybe I should consider it!
I've tried clip-leading the ASCs into a tube amp's existing power supply circuit, and liked what they did, but always wondered how much of that had to do with the extra 60 µF per cap storage capacity.
At the risk of totally hijacking Jeff's speaker crossover thread, maybe we should move this over to DIY tube or Tweakers.
Thanks for the head's up on the Jensens.
It is not my position to discuss this, but I know of a VERY good example, of just the opposite of the results ( and build technique ) in single ended amps, that Mr. Vorhis, who is a friend of mine, is posting.
This may be my thread, but I can not take the liberty to discuss proprietary subjects up here, any more than I have.
Let me just say this. The technology and design of capacitors have not stood still. There are capacitor designs / implementations made today, that are considerably better than WKZ's, etc what we had 20 to 60 years ago !!
Also, only a few people in Audio know (a) what they are and (b) know how to employ the newer products.
Most DIY audiophiles I survey, on different audio Forums, are up to 60 years behind the times, in their thinking, and in their doing. This thread's responses, IMHO, is but one good example.
The answer to these questions is wisely, "Try it for yourself, and see". Fortunately, a couple of Wima DC links in the power supply will be much more affordable to try than a couple of V-Caps would be in coupling duty.
I'll give them a try in my next project, thanks!
Some of those doddering 60 year out of date thinkers also have years of experience and have seen the "new flavor of the week" come and go.
If you want more scientific testing info on a product then do more posting over at the DIY type forums. If you want to know more of how a particular thing rings out in actual use and how it ends up sounding then post here.
It's not that these guys are 60s out, but more that these guys have tried things that have come down the audio path road FOR the last 60 years.
And I apologize for not being supportive originally and I do look forward to hearing YOUR opinion on these caps.
Listening experience? Me ?? 72 years ago, I was listening to an ALTEC 604, my Dad's, in our home, as an new-born infant. I was there at home, as a six or seven year old when Dad brought in a second 604B ( new model) driver. So, I have LOTS of experience.
These new 4 pin WIMA DC LINKs will work out fine, not by themselves alone, but WHEN PROPERLY BYPASSED, ( as most, if not all film caps require ), for full musical expression. Implemented properly, they out-perform musically the old Blackgate WKZs, and at a fraction of the cost.
Sorry, I am not free to discuss bypassing properly.
Could be cuz most Audio Weenies are aged hobbyists,
arguably even into their Dotage :-) .. simply aren't interested in the latest claimed wrinkle.
I know I'm not.
Few Millennials (as a wide spectrum grouping) could care less about Audio 'stuff'.. in general.
Best sell yer wares..quickly :-)
I have had good luck with the FK series, and to a lesser extent the VK series, but not so much with other BlackGates.
That's good to know, since I've got several places I could use them.
I appreciate your feedback.
Here's a link to a description of the Audio Note Kaisei caps. At the bottom of the page there is a hint of a true BlackGate replacement but it won't be available any time soon.
I could see what you are saying - and hearing - with a DC LINK all alone.
I always use a DC LINK as the " main cap", one in a bundle of caps, to get a certain total wanted uF value. Across it ( bypassing the DC LINK ), I put carefully selected additional film caps, much smaller in uF values, to voice the circuit, particularly if its a Single Ended circuit, until it sounds correct by ear. This is precisely how one gets a wideband-sounding single ended circuit.
Usually, one would add four to seven additional film caps, specific values, high in quality, all done by ear, as part of a bundle.
IMHO, no one cap, on its own, used alone, can do what a carefully selected bundle does, to play wideband, correct sounding, etc in a SE amplifier circuit. Same would apply to a speaker crossover.
Jeff, As you know, we disagree on the matter of parallel caps or cap bypassing. In my experience, a single very high quality cap always sounds better than a cocktail mixture of different caps, no matter what their value or brand. I can always hear their individual contributions and the result is a less coherent, less organic sound. Bypass caps can make the sound seem faster and more spectacular but it's not as natural. Maybe someone with different priorities would reach different conclusions but that's what I hear.
To determine the truth in 2017, take a pair of you latest monos up to Montana, and compare them directly to Dennis' amps, on his 604 MLTLs.
Multiple bypassing is an ART, and it is often expensive, to bypass with multiple caps, properly. Dennis is the only one "I" personally know of, in ALL of single ended audio amplifier building, who knows how to do it really well, and he has LOTS of experience doing so !!
Jeff Medwin `
Some day, it would be nice to see the "guts" of his 2A3 monos to understand what you're saying.
I don't know how many changes Dennis has been since this picture.
Pretty damned unremarkable looking, isn't it?
I mean, for a guy who makes all kinds of grandiose claims........I expected to see something special.
Have Fun and Enjoy the Music
"Still Working the Problem"
re-read my post below. I've added more useful
info to it.
Thanks for posting the pix of one of my old amps.
That one had very good bottom-end response, and
followed the musical groove with alacrity, although
further improvements have been made to date.
It exhibited an extended high-end response that
was down about 1.5 to 2.0 db at 40KHZ, depending
on the output tube chosen. Rolloff STARTED at about
15KHZ, but was down very little at that point, with
a Single-Plate 2A3..
Perhaps some comments on the Buddy Amp that you
post can be of some help to forum members:
Here goes-- I hope I don't leave anything out!:
See that long black A.C. wiring going into the
rectifier filament transformer? Why not locate the
transformer between the A.C. and the tube sockets,
greatly shortening the Line Power leads. This keeps
crap from the line out of the aluminum chassis.
Two rectifier tubes in parallel? No. Use ONE.
You can't use a 5R4. The filament has too small
a surface area, and can't deliver enough instantaneous
pulse current on top of the idling current..
Use a MODERN 5U4GB. (Example: JJ 5U4GB, or
I can only guess, but it looks like the filter
cans are probably 70uf or so. C1 and C2. Choke-
input is nice here, but those caps should be
about 35uf. Oils are OK here.
See that big Oil Can as C3? Better get THAT
out of there or your Highs are GONE!
Looks like you have a nice Plate Choke on the
driver tube-- watch its value carefully. Make tests.
What is that red cap? Perhaps 4uf Dynamicap?
Excellent! Needs a few more small bypassers to
extend H.F., or you'll lose some highs here....
Driver plate resistor. Get one that can pass
high-energy pulses accurately. I suggest the
best 2-watter that money can buy.
I don't see any form of voltage regulation on the
driver plate supply-- can't tell exactly from the
picture, but it looks like only a voltage-dropping
This CANNOT work with a high plate-resistance tube
with high gain and low plate current.. The power supply
there must be stiff, and have lots of current available.
A series voltage dropper resistor KILLS the driver stage
thru voltage sag, and current starvation. You won't get
enough power to properly drive the 2A3 grid. The amp will
be a "wimp"-- both power-wise, and bandwidth-wise.
Driver cathode bypass: Get that Solen out of
there and get something good-- another Dynamicap
with extensive, careful bypassing, or use an
Audionote copper foil, etc. Don't do it and kiss
your highs and musical swagger goodbye....
That ground on the cathode bypass is crap--
all that other junk on that ground rail before
it gets to the driver tube, which is the most important part
in the amp! Re-locate the whole layout/grounding scheme.
See the driver tube filament leads? Those run right
parallel to The cathode SIGNAL leads from the 2A3 filament
into the filament transformer. THIS layout has DESTROYED
H.F. response in the amp! Re-Do the whole layout!
2A3 cathode bypassing: Get that SOLEN OUT OF THERE!
Cheapo! Another place for a bypassed Dynmicap. Or kiss
some more highs goodbye...
2A3 cathode resistor: This is a high-current pulsed
application, so use either two exactly the same resistors
(Mills, etc.) in parallel or use one 50 watt heat-sinked
inductive wirewound power resistor. Smaller sizes WILL
result in reduced dynamics and bandwidth.
See the 2A3 filament transformer mounted right next to
the filament transformer for the driver tube? The 2A3
filament trans contains the cathode signal on the 2A3.
DO NOT mount ANYTHING near the 2A3 filament trans. Re-do
See how close that 120VAC pilot light is to those two
transformers? What do you think the 120V is doing to
that whole area? Re-do the layout! Put all 120VAC into
one end of the chassis, isolated from all else.
Grounding system as a whole: It can't be a series string.
It needs to spread-out in increments from a center point,
that is bolted to the chassis at a strategic location only.
Re-do the layout, starting at the A.C. line ground
2A3 and driver tube sockets: These will sound better when
they have some physical means of isolation from the chassis.
Aluminum chassis: This metal will sink high frequencies into
attenuation, but will also add a "whitish" (old audiophile term)
coloration to all music. This is probably a result of aluminum
reacting against the steel in transformers and chokes.
For this reason, wood is becoming popular as chassis material.
Steel is better-- much better. It is the same thing as the
transformers-- that is very good, but transformers have to be
elevated above the steel on non-steel mounts. This is to prevent
transformer vibrations from shaking the steel chassis, and
smearing/reducing high frequencies.
Look at my older amp picture. In it, you'll find all of these
things, and much more, done right.
There is one mistake in my old chassis-- much has been improved
since then, but there is one glaring mistake in it: the H.V.
output leads from the plate transformer are twisted together.
That mistake wasn't made in the Buddy Amp.
All of us can still learn... That includes myself.
Thanks for sharing your observations Dennis.
This 2a3 amp uses a very similar circuit without as many bypass caps.
I think this layout is much better WRT keeping the AC away from the signal path and keeping all the wires away from....everything (something Dennis claims to be very important and I don't disagree).
Have Fun and Enjoy the Music
"Still Working the Problem"
Very neat and orderly with short leads for each connection. If I recall correctly, didn't this have some pretty severe HF rolloff?
Yes but so does Dennis' amp.
A 7b4 with a 280k ohm plate resistor has an output impedance of somewhere near 50k ohms.
Dennis says the HF starts to roll off at 15kHz using modern output tubes which have less interelectrode capacitance than old tubes.
Dennis said with a Sylvania 2a3 the roll off starts are 9kHz and with a RCA 2a3 the roll off starts at 11kHz.
That is pretty much in line with the math.
A 50k ohm drive impedance won't drive much output tube Miller capacitance.
The other issue is the current delivery capability (or lack there of) of a 7b4 running less than 1ma.
That Miller capacitance will need to be driven and that takes current. Current that the 7b4 is not able to deliver.
The other problem with the design (my friends and/or Dennis') is the power supply.
The low value caps and low value chokes with low DCR makes for a very unstable power supply.
It's not stable with the amp at idle. When you stat playing music through it all hell breaks loose.
But if some people like the sound of that sort of thing...... who am I to complain?
It's just that for an amp that costs 18k per pair, with "best amp ever made" verbiage surrounding it........I would expect more.
Have Fun and Enjoy the Music
"Still Working the Problem"
No thanks. No need to travel to Montana. I am quite comfortable using a single .1uf CuTF V-Cap as a coupling cap and unbypassed 100uf BG WKZ caps in the power supply. They give me the sound I like.
Exactly! We each have different wants when it comes to musical reproduction... even when if/when we think we are simply accurately reproducing what is recorded. Hell, we listen differently to live music, let alone reproduced!
Some folks get it. Others don't, confusing their opinions for some truth...
Okay, Tuesday night rant finished. Back on topic now.
"Confusion of goals and perfection of means seems to characterise our age." Albert Einstein
Tre, who is typically technically correct, offered you a good answer.
I am going to try them, just wondering who else has already experimented with them, specifically in speakers.
Thanks for your thoughts, encouragement, appreciated.
Post a Followup:
Post a Message!
This post is made possible by the generous support of people like you and our sponsors: