Jump to content

Are Your Capacitors Installed Backwards ??


Kreg

Recommended Posts

 Yes It was just inside the speaker but the principle should still hold true with resistance. My primary reason for doing this was because I had read that it reduced electronic interference from drivers and other conductors in close proximity. I see no reason not to do it I just did not see any reason to do it either. 

 I am interested in what happens with interference by the way. I bought a balanced 1/8" jack to dual xlr's to feed into my xilica. Since my S-MWM system is not ready to go yet I went ahead and ran it into my crown xli800 and there was a noticeable drop in hum over the cheapo 1/8 to RCA cable I had been using.

  I had a set of Fortes in the shop one day. I had never switched from the Forte to the MCM 1900 before and the guy wanted to hear them. We had the Fortes up pretty loud. Now when I plugged into the MCM it was more than loud. Same power supplied to both and the perceived sound level from the MCM's was far greater. I think the Forte needed that big wire more than the MCM did. The idea of twisting that you suggest does interest me though I just have not found a study that convinced me to do it.

 I completely understand tolerance stack up and have to deal with it in machining. As I get my system up and running I will in time try some of these things out. That comment makes sense and is why I am going with biamping and DSP for the first time. Cables will come later if at all. I get to a certain point I won't care about further improvement's anyway.

  • Like 1
Link to comment
Share on other sites

11 hours ago, Jeffrey D. Medwin said:

Deang,   

 

Goodness, you are over - reacting

 

I have NO interest in building crossovers for Joe Public. 

 

As you have, or will will find out, the public mostly 

 

The GTO I use is now built by Kemet, $50 or so a cap, 50 part minimum order.  WIMA GTO, none in stock, special order.

 

Jeff 

  I checked the other day and Mouser shows stock in the 5 ufd,1500 volt dc, Wima GTO. Most values were out of stock. 

  Looked them up to learn something. Not much gleaned from the spec sheets.Good to see the inductance and phase plot here. 

  Reading here I saw that most members who mention speaker wire used

larger ga than I do. I think the 16 ga in the Canare 4S8 is fine. But slept on this. 

  Rather than chop off a couple 10’ sections of 4S11 I will connect the 4S8 as a quad. And whack off a couple sections of 4S6 for the mid/tweeter. 

  The latest Klipschorns and La Scala’s have Star Quad wiring. It has better noise rejection than twisted pair construction.

  The Quad connected 4S6 is 17 ga. I like smaller ga wire. I fact just thinking about it makes me feel better. 

  HAVE Inc has great prices on these. I think you have to order a 100’ of the 4S6. I purchased a 100’ to internally rewire my La Scala ii and Heresy iii. But once I saw how the wire attached to the crossover board in the LS ii backed off. I assumed posts or connectors. Not hardwired to the circuit board. 

Link to comment
Share on other sites

1 hour ago, Jeffrey D. Medwin said:

Differences in uF values of caps is very MINOR sonically, compared to the types used.  Maybe a 10 to 1  sonic and importance difference  between these two parameters. 

 

Caps that don't cut it sonically, don't ever stay intact in any "C" position, no matter how well they are matched.

 

It takes $$$$ to bypass at a highest performance level, open up your wallets...IF you wanna do it right. 

 

And on that note I think I am done.

  • Like 1
  • Haha 1
Link to comment
Share on other sites

2 hours ago, Jeffrey D. Medwin said:

Differences in uF values of caps is very MINOR sonically

I can't not make a comment here. THE CAPACITANCE OF A CAPACITOR DOES NOT MATTER? Oh I'm sorry is "very minor" You heard it here first folks.

Now I 'm questioning the voltage rating of a capacitor, does that matter or have any meaning?

Link to comment
Share on other sites

a few questions which might (?) help Jeff's POV (and tax his typing) -

 

How many amplifiers have you built over the years?

 

In retrospect, which topology worked the best sonically

for the least $ investment up to this point? - RC coupled? Transformer coupled? Loftin-White direct coupled? (with Jeff's extremely detailed advice via "snail mail" I built a 5842q parallel 2A3 SE amp about 26 years ago)

 

(just yes or no) Is there a pecking order of topology for single ended vacuum tube amplifier audio path?    How about the power supply ?

 

When are push pull tube amps advantageous over their

single ended relatives?

 

Do you keep a "control"  - reference amp to A-B as the development process moves onwards as to not be overly influenced by "expectation bias" ?

 

fwiw with Jeff's experience, I would not discount some of his ideas as long as the speaker type is fixed.   (I think he's cognizant of RC time constants - lol)

 

 

 

Link to comment
Share on other sites

I concur that this is going nowhere, but I just wanted to both follow up on an idea expressed in an earlier post and to state for the record that I know what "Mil-Spec" means.  Regarding the latter, the point I was getting at earlier was that although military specifications are usually quite good, the things they're specifying do not necessarily represent the cream of the crop, and I'd be surprised if NASA, for example, doesn't go beyond them with their own specifications in several areas (like wiring/terminations).

 

Regarding the "follow-up," I whipped up a spreadsheet to determine the reactance of each individual capacitor in various combinations (1C; C/2 + C/2; and C/2 + C/3 + C/6) at half-octave steps throughout the audio band.  I had the spreadsheet use the individual reactances mathematically combined into voltage-dividing networks and figure the voltage drop against an 8 ohm load at each of the frequencies.  The result is that the three curves corresponded exactly, as opposed to my earlier thinking-out-loud that they may not.

 

So it seems as though if one could obtain two capacitors of half-needed-value with the same (or lower!) ESR each than that of the appropriate single capacitor, advantage may well be gained by using the two paralleled instead of the single unit.  (Any inductive part of the ESR would be improved upon for the same reason as the resistive part.)  Or in combining several to arrive at a more correct value for the purpose at hand.  (Heck, maybe even precisely combining value-wise different types for their perceived audio qualities.)  That scenario may well be a more worthwhile rabbit to chase than precision lengths of exotic wire, as one example.

 

I was a bit surprised by the results but admit I'd never given it any thought before this thread started to run its course, so at least no long-held beliefs have been undermined.  Whew!

 

On 2/19/2019 at 4:33 PM, Jeffrey D. Medwin said:

I use multiple paralleled caps because NO one film cap can do the job, to my satisfaction, all by itself for the entire musical bandwidth..

 

I do not particularly CARE if the computed value is precisely met.  About ten to twenty times more important, to me,  is the quality and values of the caps, in a bundle, and what it sounds like to me.   

 

Different uF values of film caps each play a different part of the musical frequency spectrum.

 

For the same exact uF value, different qualities are displayed by different cap manufacturers.  

 

I do this I usually listen to music playback, and when I hear a range that has a suck-out, or a mild depression, I add the right uF value , to maximize that area which needs help.   You have to know, gained from experience, what uF values of a bypass cap effects what musical range

 

If I think the transition of the mid bass to midrange is deficient, I know from prior experimenting, and I  will use a 0.68 uF value.  Typically, ONLY a  DynamiCap E, is optimal in that range, ......actually its been UNbeatable in that range, music - performance wise. 

 

A 0.22 uF value will primarily effect the middle midrange, but there is probably only three Manufacturer's  0.22s I really like and will use.   

 

For a hole or depression in the lower treble, I know to try adding a 0.068. 

 

There are additional smaller values one may typically apply, going up the spectrum.

 

Its all done by ear, and by prior experiences. 


Mr. Medwin, each time you simply add another capacitor to the bundle to address some perceived "hole or depression" somewhere in the audio band (which is indeed what you said there), all you're doing is lowering the crossover point in terms of what effect that "C" position is supposed to be doing.  Since you're not doing anything with the associated inductor (I'm assuming by what I can see from your photos that it's a 2-pole crossover) to match the change in capacitance, I doubt the lowering of the crossover frequency is being effected as smoothly as it could.  I may be mistaken about what you'd said.  Perhaps you're taking some care to maintain the overall amount of capacitance.  But the present quote sure doesn't sound like that's the case.

 

In fact, let's look at your given examples.  0.68 uF has a reactance of 10k ohms at 22 Hz decreasing to 10 ohms at 23 kHz.  0.22 uF is 32k ohms and 31 ohms at those two frequencies.  And for 0.068 uF the values range from 100k ohms at 22 Hz to 100 ohms at 23 kHz.  I don't have an inkling which frequencies correspond to the three areas you mention so cannot even begin to offer what might be a ballpark reactance for any of those capacitor values there.  Nonetheless, I don't see how it's possible you're getting improvements for "suckouts" or 'holes" anywhere by adding those caps to your crossover high-pass "C" bundle.  All than can be happening is an overall lowering of capacitive reactance at that "C position."

 

You remind me in some respects of one of my brothers.  He's known for coming up with some real oddball ideas, like the time he considered putting up a wind-powered generator and gearing it up (overdriving it) to produce even more power from the same amount of wind (and selling the excess back to the power company).  I tried to explain to him that it just doesn't work that way, but, oh well.  He's my brother and I love him anyway.  Thankfully he never followed up on the idea.

 

Oh, you don't have to answer, but I wonder whether 57 1/8 includes or excludes the wiring from the output transformers and inside the speaker cabinets.

Link to comment
Share on other sites

15 hours ago, Jeffrey D. Medwin said:

 

 

Hello Glen:

 

(1)  57 1/8th inches comes from the Lab of Robert Fulton. 

 

 

Jeff 

 

 

 

 

  I will check my existing cables to make see if they are at least 114.25”. And make sure the new ones are close. 

  I was planning on 120” but that was not a hard number. Perefect length was enough to make the span from amplifier to speakers.

  Being a EE I would assume that length in a copper cable is some percentage of the wavelength of a particular frequency. Since the relative velocity of the EMF wave is an approximation in copper cables a transmission line the significance of 57.125 inches long is tough to pin down.

  The most common number I see for copper cable is 0.7. Propagation coefficient. I suspect that is precision to one decimal point. 

  If larger cables improves the system and having the speaker cables a multiple of 57.125 inches long is audible who knows what’s next? 

Link to comment
Share on other sites

23 hours ago, Jeffrey D. Medwin said:

Deang,   

 

Goodness, you are over - reacting

 

I have NO interest in building crossovers for Joe Public. 

 

As you have, or will will find out, the public mostly 

 

The GTO I use is now built by Kemet, $50 or so a cap, 50 part minimum order.  WIMA GTO, none in stock, special order.

 

Jeff 

   I see a lot of talk on these ideas. But I am a man of action. These biwire Star Quad speaker cables are exactly 114.25” long. 

  They do seem to sound better than two single runs of 16 ga. The initial transient on percussive attack is better defined. Whether half the resistance on the woofer cable, the Star Quad geometry, or being a multiple of 57.125” in length I cannot say.

  BTW, how is the speaker cabled measured? From tip to tip, binding post to binding  post, or insulation to insulation?

  It is dark, sneaking over to Country Place Country Club to dive for golf balls in the water traps. Gators move slow when the water is this cold.

2380AB0C-5FB6-4F53-8174-50C494A3345A.jpeg

Link to comment
Share on other sites

I would say "center of inner lug radius to center of inner lug radius".  But that discounts internal wiring of both the amp and speaker... hahaha!

 

If you're going that route, you've got to be sure the lengths are all within +/- .005 (inclusive) or you're wasting your time.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

×
×
  • Create New...