Great post! Does it make a difference that the signal and ground wires are different lenghts?
26 responses Add your response
Onhwy61, I tried the experiment first with just 2 wires, equal length, with no tubing or teflon tape just to see if I liked the sound. It seemed fine that way, and I noticed no RF problems. But, I was a little concerned about the fine wires hanging around the back of the amp being too delicate and easy to snag. So I wanted to use a tube for each wire, but that was too cumbersome and hard to attach to the RCA's. Then I figured that if I used one tube with teflon tape and wrapped the ground wire in a spiral, I could get a single normal looking IC with some RF protection and reduce the capacitive/inductive effects of 2 parallel wires running full length close together. I didn't notice any change in sound with the slightly longer ground wire. And it cleaned up the package nicely, so now I don't have to worry about it. Also, since the 1/4" tube is way oversize for the wire, the signal conductor has air dielectric for about 1/8" before there is any effect from the other materials(Yes, I know it is not centered and touches the sides here and there, but, hey for the most part it has air dielectric.)The break-in time was pretty short, only a couple of days and it smoothed out. Since I am running direct source to amp, this is the only set of IC's in the system, so it is easy to judge the differences. It was one hell of a lot better than the basic cables that I was using temporarily. Mids and bass were better and the highs benefitted the most - less rollof and clearer. This system now throws a soundstage that is deep and wide, with great imaging way outside the speaker boundaries. What can you lose for a couple of bucks? Try it, you might like it.
The late Harvey Rosenberg of Listener and NY Audio Labs popularized a similar design many years ago. You could just as easily twist the two wires around each other, creating a twisted pair. Since they are enameled, you don't have to worry about them shorting out. If you are worried about something like that, try using the individual conductors from some CAT 5 cable. The better stuff is Teflon insulated and it is solid core 24 gauge. It comes pre-made as multiple twisted pairs, so you would only have to terminate the wire into the RCA's.
David Magnan was the first one to try using MEGA thin wire for interconnects. He did this after studying time delay distortions and skin effect. He has since gone from using small gauge wires ( 36 gauge ) to conductive paint. Since the conductive paint literally has no depth, it is all "skin" for increased surface area.
As to the bottom line, i too agree that you can make some pretty phenomenal interconnects for next to nothing. Sean
what connectors do you recommend? also, instead of putting the signal wire down the middle, what about wrapping the signal wire around the tube as well, spaced out so they don't touch? and is there any advantage to using teflon tape coated polyethylene tubing rather than plain old teflon tubing? trying to figure what route to go. must upgrade from monstercable and don't want to go broke.
Good question. I did not try wrapping both conductors, evenly spaced, around the tube. I did read about that method though. The reason I put the signal conductor down the middle of the tube is because it is a much oversize tube and I wanted the "air dielectric" around the signal wire. I didn't find any teflon tubing in my hardware store, so I used polyethylene. If you can find teflon, then I think that would be better, or at least just as good. I think the "air dielectric" idea actually works, and I used the same oversize tubing on my 22ga solid core speaker wires with a separate tube for each conductor. I spaced them apart with 2" spacers and got very good results from them too. You can experiment with this stuff pretty cheaply, so you can try different methods. I am not claiming this is the greatest, but it does sound real good for very little money.
This is basically the Chris Venhaus design, which can be found here: http://www.venhaus1.com/diysilverinterconnects.html. This particular recipe uses thin (28 - 32 gauge is recommended) silver wire.
Based on this design, I built an XLR interconnect (the pin1 -> pin1 wire loosly spiralled around the outside) just a few days ago and I'm about to assemble an RCA interconnect as well. And with some braided sleeving to cover it, it looks just as professional as anything out there.
It's not broken in yet (and silver is notorious for being slow at that), but from briefly listening to it, I can say that it sounds at least equal if not better than my HT ProSilway mkII (tad more detailed and extended).
Hi to all!
I've tried a similar recipe too. Wound two magnet wire of 20 (or 18 gauge-not sure) around a polyethilen core in 1.25 inch spacing, used teflon tape, cotton around it, heat shrink tubing, Canare gold plated connectors and ended up with a very "serious" looking cable. But the sound of it isn't what i'd expected. Too dark, not very detailed, bass not very defined, too much of it's own caracter I would say. What is good about it is a pretty expansive(but a bit diffuse) soudstage and pretty "big" (but soft) bass. Might be good for taming a very shrill system.
Perhaps the awg of the wire is to hefty and the main reason for the poor "sound"? Made another pair with smaller distance between wires - but this sounded even worse. Curiously, I found that the directivity with this wire also plays a role. Certainly good for the money, but as well nothing more...
So I continued to use my good old "reference" Monster M1000i cable...
I have posted this as a waypoint to the (possibly) wrong direction - using to much awg wire.
I love "making" cables so I will continue - can't help it!
I know only that he recommended 12 AWG stranded wire.
In the designs I have made I used 13 or 11 AWG stranded wire(had them available, no other "special" reason).
You might, as well, use some shilded cable for the ground wire(as I've done in some designs). Connect the shield only at one end with the shield of the "main" Cable.
I can't specifically tell you what (soundwise) is the difference between Shilded and non shielded ground wire as I haven't use them on the same designs.
There are some other recipes that I found work very good.
Go to the www.tnt-audio.com website and search for the TNT TTS(The Twisted Snake) DIY mains cable. It's a "must try" if you're in DIY! I realy have made many different cables but like them very much in my current setup.
I could not find "the" recommended cable, but made them of similar cables. It's the design that counts the most, I believe.
By my experience cables with several smaller AWG conductors "sound" better then a similar/same construction with only two or three bigger AWG wires!
The latest, and I'm afraid best(in my current setup), design I have made 2 days ago...
I've been reading about the Virtual Dynamics power cables on another forum (and their website) and wanted to give a try to something as close to them as i could get.
I used 7 AWG(!!!!) solid core cables, shielded them (so I got an "coaxial" cabel) for the positive and negative wire. For the ground wire i used a simple stranded 11 AWG wire. No twisting.
It's a dinossaur of a cable! I don't have to tell you how difficult it was to terminate this "cable"! It's stiff as...I don't know what to compare with. And It's a pain in the ass connecting it... But when finaly done this task I was surprised with the result! Even not "burned in" yet it sounds amazing! I've never heard such a "complete" cabel! Dynamics, soundstaging, you name it - sound amazingly effortless and lifelike! I forgot to mention I "put" them onto my CD Player. Immagine - this AWG wires for "only" a CD player!!!
But it's to unpractical to use. I will try to make some smaller AWG (10 or 11) cables to make them a little more user friendly.
According to Walt Jung and Richard Marsh, solid core 20 gauge is linear out to 20 KHz. Above that point, one may begin to experience skin effect. This will vary with the shape of the conductor though, so some conductors may be better suited than others. Wide and flat solid conductors with rounded edges will work best but then you have to start worrying about high capacitance as you bring the conductors closer together.
While high capacitance isn't a big deal for speaker cables, it is for an interconnect. Goertz finally realized this and came out with their Micro-Purl series of interconnects after already having their "wide, flat & stacked" interconnects. While the Micro-Purls still use flat, wide conductors stacked on top of each other and twisted, they made one very important change. They inserted another unused flat ribbon between the two active conductors. This raises the inductance / lowers the capacitance and makes the cable more usable universally. On top of this, the same cable can be used in a balanced mode now, as you have three conductors at your disposal rather than just two. Too bad they didn't put the Micro-Purls into a better dielectric, as they might have had a real winner with this design.
In the same article mentioned above, Jung and Marsh stated that their tests showed that the use of multiple 24 gauge solid core wires in parallel was the best way to go. This offered low series resistance with no phase or skin effect problems in or directly above the audible range. My experience is that this type of design works VERY well. That is, so long as one uses good quality materials and pays attention to the electrical characteristics of the cable.
In my experience, using just one conductor of this or smaller gauge typically results in a noticeably lean tonal balance. The end result of using multiple wires of light gauge in parallel is a far more even tonal balance without intruding on the high frequency performance that the smaller gauge originally offers.
Magnan is the company that uses conductive paint for some of their interconnects. The drawback to this is that it has a very high series resistance, which in my opinion and experience, is not good thing to design into a cable of any type. Some of the other Magnan's that i have make use of 36 gauge solid core conductors. As a general rule, you'll find that higher grade interconnects avoid using stranded wire. While one can use stranded wire, problems with skin effect and strand jumping will become more apparent at a much lower frequency. Sean
PS... As i've mentioned before, don't shoot me if you disagree with what i've posted. I'm only the messenger. Take your beef up with Marsh & Jung.
Sean, I agree with the 20ga wire being ok for skin effect in the audible range. I used that or 22ga for speaker wires on low power systems., but my trials have shown that in IC's, the 26-30ga range works best overall. I have not had any bad effects from leanness, but they aren't exaggerated on the bottom end like Cardas, either. Just fairly neutral and detailed. I think the 30ga for short interconnects and 26 for longer ones is fine. By the way, there are now quite a few Audiogoners that are using these DIY cables, and they all gave very nice reports to me about the sound. Nobody has complained yet. I have even made several sets for those who didn't want to DIY. I make them to custom length and use whatever plugs they want, and they look and sound great. I now use the TechFlex cable coverings and they look real professional. The sound is very competitive with some stuff that is over $500 per meter pair. I only charge $39.95/pair plus whatever plugs they want to use. It's still a bargain for those who don't want to make them themselves. Not getting rich, but I am helping some members to get good cables for cheap prices. And for those that are handy with a soldering gun, the formula to make them is posted in the thread. No trade secrets, just good sound.
This is Chris's receipe from his web site under DIY fine silver interconnect. I've seen this few years ago and it is still there. You should give him some credit.
Btw, depends on what you're looking for. 30 gauge only works well with neutral to warm system. A bright system will suffer. If you have a bright solid state amp, I would suggest at least 26 gauge or larger.
S23chang, this is NOT Chris's design. Chris's design is a double-helix configuration which has both wires located on the outside of the teflon tube. Mine has a center conductor that travels inside the tube, and an outer conductor that wraps around the outside of the tube. My center conductor has a true air dielectric, and the outer conductor has a teflon dielectric. My cables use different wire gauges than Chris's. So, the entire design is different. Different configuration, different wire spacing, different wire gauge, different dielectrics, different everything. I even have different wire lengths, where my ground wire is longer than the hot center wire. Yes, they both use wire, tubes and RCA plugs. That is where the similarity ends.
Some may like my cable better, or some may like Chris's cable better, but they are not even close to being the same design. If I had used Chris's design, I would certainly have mentioned it, and gave him credit for it. I have no need to steal anyone else's ideas.
Twl: I agree with what you are saying, but in the long run, almost all cables / cable designs are plagiarized from someone / something down the line.
As far as i know, the first person to build cables using similar design philosophy to what you are talking about and what Chris' design is based on was David Magnan. From my experience and what i've read over the years, he was the first to actually build and market cables trying to reduce skin effect by utilizing "scrawny" wire while trying to minimize dielectric absorption through air insulation. Once he set forth the basic idea / criteria, others followed suit. As you can see, there are more than a few dozen ways to get the job done and some may work better or are easier to produce than others.
For sake of clarity, Jung and Marsh also did some research in this area and published their findings, but i don't which of these two groups was first to publicly display their findings. Even with their relatively similar idealogies on the subject, they both suggested different ways of achieving what were basically the same goals. Hence the similarities in comparing your design to those of Chris VH. In effect, they are what is commonly called "the same difference" i.e. a couple different roads that lead to the same destination with each route each providing different scenery along the way : ) Sean
Sean, your probably right. There may not be much, if any, difference in sound between these cables. But since all things are different, it is likely that some small differences will show up.
I mean, all the cables need to have some kind of wire between the RCA plugs, and they are either straight, twisted, or some kind of combination. They will have some kind of dielectric, and some kind of wire gauge. I guess that all these things add up to whatever difference in sound we hear from them.
Whatever the reason, this is a pretty darn good sounding interconnect, and it doesn't cost much. I think that is the key.
Twl: I built a set of cables incorporating some of the ideas that you discussed. The wire i'm using is 24 gauge Teflon insulated silver plated copper. I'm using two per polarity for an equivalent 21 gauge conductor. I have multiple spools of this stuff, so i figured i'd give it a try. The two "positive" conductors are wrapped around each other in a twisted pair configuration and the two "negative" conductors are also wrapped in a spiral pair. The positive conductors are run in a straight line down one side of the tube and the negative are spaced on the opposite side of the tube in a straight line. Capacitance is pretty low, appr 8 pf per foot, so i know that should make Audioengr happy : ) The RCA's are locking models that have a Teflon dielectric.
I finished them a few days ago and threw them on my burner. After letting them "cook" for about 36 hours, i gave them a quick listen. They were not bad at all. I compared them to some of the Magnan's that i have ( Models 2, 3 & 4 ) and they help up pretty well. As you may know, the Magnan's use very fine wire ( something like 36 gauge ) and try to achieve an air dielectric. I did not measure any of the Magnan's, so i don't know what their electrical characteristics are like. They've always come across as being very smooth, airy, liquid and detailed but a little on the lean side.
I also tried them against some Goertz Silver Sapphire's, which are similar to their speaker cable design. These are flat ( NOT twisted ) and make use of 18 gauge pure silver ( not plated copper ). This design is VERY, VERY, VERY high in capacitance and i was wondering what to expect. In most cases with typical SS line level gear is concerned, highly capacitive interconnects are a "no-no".
In both cases, the Magnan's and the Goertz sounded slightly more liquid with a more cohesive presentation. The upper mids on my "home-brews" were slightly subdued, resulting in a presentation that was not quite as spacious sounding and slightly lacking in overall clarity and transparency. Female vocals were a little less intelligible and slightly more congested sounding. Frequency extension was quite good on both ends of the audio spectrum. The soundstaging on all of these cables was excellent and beat out several others that i tried in the same location of the system.
I put the cables back on the burner and am going to let them cook for a full 30 days. It is quite possible that they will sound different after "fully cooking" as i'm sure i caught them "in transition". I'm hoping that the upper mids fill out a bit in the process. I may have to tweak the design a bit more otherwise.
I've also got another set of homebrew interconnect cables that i'm working on, but haven't gotten them fully assembled yet. I thought i had them finished up, but forgot to do something and had to pull them apart again. I'm following the design that Jung & Marsh laid out in their research, so we'll see how that goes. For a three foot pair of stereo interconnects, you end up using 96 foot of wire !!! Needless to say, there is a LOT of cutting and soldering going on. Once i get them put back together, they'll go on the burner for 30 days of "punishment" and then i'll give them a listen. Sean