|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
Hello,
If you want to skip straight to the question, it’s at the bottom.Purpose: I’m making several sets of balanced cables, 89259/89248 amongst them. However, my current dilemma doesn’t concern the 89259/89248, but the twisted pair I’m making from XLO 26 ga solid copper (5n or 6n pure) with solid TFE insulation. I bought it from Percy, whom I found great to work with BTW.
Background. I twisted a pair 1 meter in length and connected them to the Neutrik NC3FX-B connectors at the source end and NC3MX-B’s at the other end (I got a great deal on these: F’s were $2.60/piece & the M’s were $2.35/piece from John ext 1140 @ Full Compass). However, in this configuration I found that any number of sources added audible noise to the system, the worst amongst them were dimmers on lights. My neighbors, I live in an apartment building, have dimmers too so I can’t eliminate the problem by simply leaving dimmer actuated lights off when I listen.
Background cont’: I need to shield all my cables. Within the next day or two I’m going to order Teflon tubing in the 4 ga inner diameter (just a hair under a ¼”) from McMaster-Carr where it runs about $0.74/ft & is sold in 5/ft increments. I’m planning on running the XLO 26 ga, Kimber SF23, & 89259/89248 twisted pairs through thge Teflon tube, place a braided copper shield around that, cover w/TechFlex (got the TechFlex for $19.98 for a 100 ft roll of black ¼” from Full Compass too), then finish w/heat shrink @ the termination. I think it’s going to be a tight fit for the 89259/89248 inside the Teflon tube, but I’m confident that the Teflon tube could take a slight oval shape if the twisted pair is larger than ¼” (comments, will this work?). The primary purpose of this post is another question, however. Unlike the 89259/89248 twisted pair, the XLO 26 ga twisted pair is relatively small and would flop around in a ¼” tube.
Question (finally): I’ve considered a couple methods of keeping the smallish 26 ga twisted pair relatively centered in the ¼” Teflon tube (keeping it centered serves the purpose of minimizing the electrical interaction with the shield): 1. cotton from cotton balls placed between the tube and the twisted pair (I’m not sure exactly how I would keep the cotton distributed evenly all around the twisted pair or if the cotton wouldn’t compress when the tube was bent allowing the twisted pair to move closer to the shield) 2: getting some extra length of the 4 ga Teflon tubing which has about 1/32” wall thickness (or some other gauge--suggestions), quartering it, and placing the quartered pieces around the twisted pair. Both of these methods are basically the same that I’ve seen suggested from Chris VH’s or Jon R’s site for other, but similar, purposes. Is my logic sound (i.e. is it important to keep the twisted pair basically centered)? Has anyone tried these (can you confirm either or both methods work great in this context), or can anyone suggest superior alternative methods?
I’ll be sure to post my results if no one tries to talk me out of this.
Thanks so much for helping.
Follow Ups:
I make a cable with a similar design to what you are talking about. I have been using 30 to 32 gauge wire, but it should work with 26 gauge. I first feed the wires through 24 gauge teflon spaghetti tubing, you would probably need to use 20 or 22 gauge spaghetti tubing to easily feed the 26 gauge wire. I haven't had any problem with rattling of the wire inside the spaghetti tubes. The purpose of the spaghetti tubing is 2 fold. First it separates the wires slightly which helps the sonics, especially transients, considerably over simple twisted wires. Second it makes the twisted wires fit nicely in the outer teflon tubing.The outer teflon tubing I use is 3/16 OD standard tubing from US Plastics, 1/8 ID. Twist the spaghetti tubing with the wires inside and then wrap with a layer of thin teflon tape to hold the twists and to provide a dampening media to keep the twisted wires from rattling inside the outer tube. It should be a slightly snug fit.
Remove the braid from some coaxial cable and slide over the 3/16" outer teflon tubing. This is fairly easy to do since the braid will expand when compressed axially. Cover the braid with a layer of teflon tape to hold it in place, then put on some tekflex for the outer jacket. Attach the rca plugs, with the shield connected at only one end. The outer cable diameter is about right to snugly fit in Dayton locking plugs.
Makes a very nice sounding shielded cable. I have built with both 6N copper and 4N silver. I like the silver better although the cost is higher. Not quite as detailed and open as the silver cables using ChrisVH's design but if you need a shield these are darn good.
Thanks Miner,
I had considered individually pushing the conductors through small Teflon tubes before the bigger outer Teflon tube. I wondered what affect this would have on the twisted pair because it was my understanding that the pair were to be as close to each other as possible. Now I believe that the purpose of the tight twisting is to keep the two cables mechanically immobilized. I haven't placed the Teflon order yet so I'll add the smaller tubes and give it a shot. I've got enough cable to try several methods, and this will be one of them.Thanks again.
Thanks Miner,
I had considered individually pushing the conductors through small Teflon tubes before the bigger outer Teflon tube. I wondered what affect this would have on the twisted pair because it was my understanding that the pair were to be as close to each other as possible. Now I believe that the purpose of the tight twisting is to keep the two cables mechanically immobilized. I haven't placed the Teflon order yet so I'll add the smaller tubes and give it a shot. I've got enough cable to try several methods, and this will be one of them.Thanks again.
to make sure everything will fit as planned. McMaster-Carr sells both teflon and spaghetti tubing and their website gives the ID and OD's. You will need a little clearance to put the twisted wires, with a wrap of teflon tape, into the larger tube.Bob
You've got it right. I think I'll get as accurate of measurements as possible and then add them up for a cross section. Hopefully it will be fairly accurate, but I must err on the little too big side, rather than a useless little too small side.
Dimmers shoot out bursts of ghastly amounts of RFI at 120 Hz intervals. And in close proximity, this interference permeates just about everything, including shields on cables. (Unless the shield is Mu-metal.)Balanced cables are normally *immune* to this kind of interference. In the form of "common mode" rejection. (The interference is picked up by *both* lines, but since the signal on the two lines are in opposite phase, any common mode noise products cancel each other out.) The presence of noise on a balanced line suggests the *possibility* of the XLR connectors not being wired correctly.
Miswiring XLR connectors is a *very* common mistake with DIY- If one wires the balanced cable intuitively, based on the physical XLR pin layout (the middle pin of the symmetrical layout would erroneously suggest that pin 3 should be the ground), it's probably wrong... And if it's wrong, you will not get the benefit of common-mode rejection. (I personally learned this the hard way. And I'm pretty sure others have too.)
I've also seen the XLR connectors on a balanced cable not wired in a mirror-image at each end (viewed from the solder side), which is also a common mistake. As well as mistaking pin 1 for pin 2.
Pin 1 is the ground, pin 2 is the positive signal, and pin 3 is the negative signal. (Pins 2 and 3 may be inverted.) At first glance, this wiring layout will seem strange. You may want to check this before trying extra shielding.
XLR connectors can be very tricky- even experienced DIY'ers (including myself) miswire these things from time to time. I've also come across production cables with miswired XLRs.
Whether or not this is the cause of your noise, good luck.
Balanced cables are normally *immune* to this kind of interference. In the form of "common mode" rejection.Actually it's not the cables at all which reject common-mode noise. They play no role in themselves in this regard. What balancing of the cable does is help insure that common-mode interference remains common-mode. Any imbalance will shift some of it to differential mode.
It's the differential input that the balanced cables are ultimately feeding which is responsible for the actual rejection. And how well it accomplishes this depends on the actual design. But regardless of design, common-mode rejection typically degrades with increasing frequency. So while it's pretty easy to get good common-mode rejection at say power line frequencies, once you get up into the RF range, it's typically pretty worthless so you have to either rely on shielding, filtering and/or amplifier designs which don't have problems with RF at their inputs.
se
Granted. (Pardon my "layman's" explanation.) But still, if the XLR is wired wrong, the benefits of common mode rejection are lost. If the balanced receiver sees only one-half of the balanced signal (the other half is either floating or tied to the receiver's ground) or if the shield is carrying one-half of the balanced signal (the shield will pick up interference differently than a standard conductor), the balanced receiver will not be able to reject the interference.
Granted. (Pardon my "layman's" explanation.) But still, if the XLR is wired wrong, the benefits of common mode rejection are lost. If the balanced receiver sees only one-half of the balanced signal (the other half is either floating or tied to the receiver's ground) or if the shield is carrying one-half of the balanced signal (the shield will pick up interference differently than a standard conductor), the balanced receiver will not be able to reject the interference.The balanced receiver will still reject any common-mode noise. But yeah, if the two halves of the cable don't have the same impedance with respect to ground, then there will be some mode conversion going on and the receiver will just pass it on.
se
I used two pieces of cotton sash cord as a filler on my unshielded twisted pairs. I then wraped the four pieces with teflon tape and then tech flex. I did it primarily for cosmetic reasons. I doubt that it would fit in the tubing you have now but you might add shrink / shield / shrink on top of it. You can get 100' of sash cord for a couple of bucks at the Depot.
Gene
Thanks Gene W,
I'll be trying any number of methods to see what works. I had thought of some sort of cord instead of "loose" cotton. It would be easier to attach and wouldn't compress like I would anticipate cotton from cotton balls would.Thanks again.
Just use the "heavy wall" Teflon tubing from McMaster-Carr. This is the same outer dia., but small inner dia. It is a little stiffer, but is still workable.
Thanks audioengr,
I considered this. US Plastics has several wall thicknesses to chose from as well. I suppose for the 26 gauge cable in question it's sort of a no-brainer. I was reluctant to go with the thicker wall because I wasn't sure what the "flexibility index" would be. Oh well, why am I so worried? My amp is almost 100 lb's and the DA is said to be 40 or so. The preamp is lite (i.e. like 8 lb's or so), but I've got one of those fancy pants magic bricks that were all the rage a decade ago to put on it in the unlikely event the cables actually are too stiff.Thanks much for your input.
I actually use the thicker one in one production cable. It is flexible enough.
I agree with Jon- you don't want the pair flopping around inside the tubing. Try using a thick polyester yarn spiralled around the pair to use as spacing around the pair when it's inside the tubing. This is essentially how I make the DIY AES/EBU cable. Good luck.
Thanks Chris VH,
I'll give that yarn a try. It isn't a very expensive proposition either, which is nice. I'm also going to try the sectioned Teflon. If I hear a difference I'll let everyone know.
Thanks for all your time/effort too. I've read and reread the info on your site considering various IC's. I'll give your specific recipe a try in the near future.
One of the problems with trying to use tfelon tubing as a container, is that in order for the wires to run inside of it, there MUST be some clearance, or you will never get the wires down the tube for any significant length. Even three feet is a long way when you are pushig and proding, etc.The problem with tubing large enough to pull this off, is that now the wires will rattle around inside the tubing, and motor/generator distortions could occur, due to the hot wire interacting with the shield, as well as the twisted pair being allowed to move.
Perhaps an oversized teflon tubing, and wrap the twisted pair with enough teflon plumbers tape to bulk it up, keep it slippery, but somewhat compressable.
Whatever you do, you want to keep the twisted pair in a tight twist, to minimize any potential relative motions.
You will also want to cover the outer braid with HS, to keep it from moving.
Jon Risch
Thanks Jon Risch,
Do you happen to know if the 89259/89248 twisted pair will fit inside the 0.208" inner diameter of the Teflon tube (URL to source below)? I've added the two core diameters and it would appear that it won't work w/o ovalizing the tube, however when twisting the pair some of that foamed PFE must compress atleast a little.Back to the XLO 26 ga. In similar fashion to the bulking up w/plummers tape, perhaps I could run some very small gauge TFE tubing (e.g. 28 or 24 ga) along side the twisted pair securing it w/plummer's tape (similar to the sliced foamed PFE suggested in your recipe)?
With regard to the HS recommendation, are you stating that the entire length of each IC should be heat shrinked (i.e. a one meter pair would require 2 meters of HS)?
Thank you very much. I have benefited a great deal from all of your time & effort.
You can order it from US Plastics they are a little cheaper then the hacked up cores from coax, and its no work :-)dee
;-D
Thanks Penguin,
That's very helpful. In the case of the 89259/89248 would I be better off skipping the 4 gauge tube all together? If so it'll save me $1.48/pair ft of expense. The only thing it contributes is a little greater space between the shield and the conductors, but I wonder if the conductors being somewhat loose inside it doesn't more than offset whatever gains would be realized by the additional spacing? It sounds to me like the twisted pair might not even fit through there anyway.
John's recipe is pretty clear about the importance of spacing the shield away from the conductors and letting it fold into the creases of the twist. I found the 1/8" od teflon tube fit extremely well and the quad twist holds together very nicely. I have not tried to force the thing inside of a tube but i would suspect it is rather difficult. You can still make the twist under $5 a foot so a three foot pair cost less than $80 with decent locking connectors, $150 if you use the WBT crimps. I think it is not a bad deal.
dee
;-D
Penguin,
I was assuming, maybe not the smartest thing, that with the two conductors + the two additional small guage tubes that the shield wouldn't be able to get into the creases. I'm really not on a total budget kick, but if the twisted pair/additional small guage tubing combo won't even fit through the 1/4" tube then what would the point be to begin with?I like to plan this stuff out ahead of time as much as possible, but sometimes you [read me] just have to go forward and see how it all turns out. I'll probably order enough 1/4" Teflon to do the two pairs regardless. If it fits, I use it for this project. If it doesn't, maybe I'll use it for some other project in the future.
Absolutely we want to minimize the interaction of the shield/conductors. I'm just trying to figure out the best way to do it, & a way that can be accomplished w/o too much trouble.
Cousin... Are we related ? :-)The idea is that you use the 1/8" in a star quad twist configuration to space out the shield. Is saves you from cramming the thing into a tube. The two cores and the tubes line up something like this on the cross section
oO
ooYou get the picture. The whole thing slips into the shield. It is a lot easier than trying to thread the twisted cores into the tube.
dee
;-D
Re: Got it! Thanks.
I do not believe that you will be able to "ovalize" the teflon tubing enough, and the foamed teflon does not give hardfldy at all, it is very tough and stiff.Yeah, the smaller tubing will wotk better for the smaller wires.
Yes, full HS coverage would be best, although you can leave small gaps to help flexibility, and essentially use bands of HS every few inches, with just a 1/4" to 1/2" gap at each HS band.
Jon Risch
Thanks.
The dimmers ect mainly create noise on your AC line, so instead of the shielding, perhapse a good power filter, like the JR line filter would be better.
I do not like shields. But, for shielding from EMI/RFI, the small teflon tube spacers sound like the best idea to space the shield off. I would get as much distance as possible there, so you may want to use a larger mesh shield, with the teflon tube spacers left whole or uncut. You could wrap everything with teflon tape, then the shield. Best, Greg R.
Thanks Greg R,
I was trying to avoid the shields... but the noise doesn't come exclusively from the AC line. I have two pair of AQ Viper balanced, and when I use those the system has a deep black quiet quality regardless of dimmer usage (though I still have mine off when I listen). Interestingly, I have a 1 meter pair of interconnects made from solid 20 ga silver conductors slid through TFE dialectric, non-twisted, non-shielded, terminated w/single-ended WBT connectors and the noise isn't quite as bad.Nevertheless, I still plan to build JR's line conditioner/surge protector in a few weeks. I'll probably go w/DIYCable's because I like equipment to look finished (thus the use of TechFlex). Do you have any other sources for a finished looking (I'll build it) JR conditioner?
Thanks again.
Is it possible that the AQ shield is in part makeing the highs quiet because the shield is interacting in a negative way with the signal wires? I'm sure your correct that the shield may be reduceing noise, but aren't the highs and transient response also being reduced a bit?
The 20 ga untwisted and unshielded ICs should be very noisey if you have a large EMI/RFI problem. These ICs tend to support the case that the black quiet backgroung produced by the AQ ICs are being produced by an overall reduction in the highs, due in part to the interaction of the shield with the signal wires. So, I'm guessing you have alot of noise on your AC line, and would benefit the most with an AC filter.
I think you will be surprised at how nice it sounds with the JR filter! I painted mine black, then sprayed some clear acrilic on it for protection. Looks great down there in the shadows. ;-) BTW, a large braided cord powering the JR filter helped alot. Best, Greg R.
As far as the highs being quiet, I suppose I would have to say that I'm not aware of it. Of course, I'm building some cables for comparisons. I'll let you know what I hear. My speakers have the Scan-Speak Revelator 9900 tweeter, and any deficiencies in the HF should be readily apparent.I have experimented with ferrite cores around the power cords. Unfortunately, when I get on these binges where I'm making changes, I make a lot of changes... with the result that I don't always take proper care to evaluate each change individually. The cores didn't prevent the noise, however.
Thanks for the input.
This post is made possible by the generous support of people like you and our sponsors: