Search the Community
Showing results for tags 'cables power diy'.
Found 1 result
There has been a lot of posts recently in other threads about making your own DC cables for LPS-1, JS-2, Y adapters etc. I'm going to attempt to give some theory behind DC cables and CA and some hints for making your own, and some results of my own experience. So what is the big deal, DC is, well, DC! Resistance should be the only thing that matters, right? Wrong! Particularly for power digital circuits some AC issues actually do make a big difference. This is due to the fact that the load current of most digital devices is not constant, it varies all over the place, and can do so very quickly. This rapidly changing load current causes a voltage to develop across the inductance in the cable. No matter how good the regulator in the power supply is this noise will still be there. This means that both resistance and inductance of the cable matter. Cable capacitance is a good thing, but the ranges available in cables have a very small effect so so you can pretty much ignore the cable capacitance. Are there "cable parameters" that DON'T matter for a DC cable? Yes, anything that primarily affects distortion is not an issue, such as stranding vs solid core, conductor material and plating. The difference in resistance between silver and copper does produce lower resistance than copper, but just making the copper conductor slightly thicker does the same thing. Thus I do not consider the vastly higher price of silver worth it. The net result is that stranding doesn't matter, OCC etc doesn't matter, dialectric has extremely small effect. So what DOES really matter? The wire gauge determines the resistance and the geometry of the wires (how they are arranged in the cable) makes a HUGE difference in the inductance. So things to optimize for, thick wires and use the proper geometry. The geometry is actually far more important than the wire gauge, although you don't want to go with really thin wires either. There are 4 common geometries in use in DC cables: parallel conductors (zip cord) coax twisted pair starquad They are listed in decreasing inductance, zip cord has about 10 times the inductance as starquad for equivalent gauge. Coax and twisted pair are pretty much the same at about 1/3 the inductance of zip cord. All can be shielded or not shielded. The shield does not really make things better for most applications of DC cables, but if it is done wrong (which is extremely prevalent) it can make things worse. Thus I recommend using unshielded cable for most situations. I hope from the above you come to the conclusion that using starquad is a good thing when making your own DIY DC cable. So what is starquad? It consists of four conductors, the whole group of which is twisted in the cable. This is NOT two twisted pairs. They are not woven or braided, the four conductors are arranged in a square. Diagonally opposite conductors are connected together at both ends of the cable. That is it, not very complicated. There are many articles on the net covering starquad with nice pictures. (sometimes it is called star quad, sometimes star-quad, they are all the same thing) So what is special about starquad? Well of course you have two wires in parallel which halves the resistance, but the big thing is the magnetic field, it forms in such a way that the inductance is about 3 times lower than coax or twisted pair, AND it produces an inherent very good shielding effect, without using a shield! So to make your own starquad cable you need a cable with 4 conductors, with the whole thing twisted, preferably unshielded. There are many of these on the market. They do not have to be specifically marketed as starquad. Many cables that say "starquad" have wires with only two colors, diagonally opposite wires have the same color, that makes it easier to figure out which ones to connect together. But cables with different colors for all for wires also work well, you just have to figure out which ones are diagonally opposite each other. So what gauge to use? A 24 gauge starquad microphone cable is probably too small, and a 14 gauge speaker cable is probably going to have wires that are impossible to connect to the barrel connectors most components use. So somewhere in between is probably a good choice, 20 or 18 AWG wires are probably the sweet spot. Belden 8489 and West Penn 244 are 18AWG 4 conductor unshielded cables that work very well. Canare makes a starquad speaker cable called 4S6 which has 20AWG wires which should be a very good choice as well. Alex likes to use a cable with a shield which I think just makes things more complicated than they need to be. If you want to make your cable without soldering anything there are barrel plugs with screw terminals available at many places. These terminals will accept a single 18AWG wire nicely, but it gets very tricky to put two 18AWG wires into each connector (which is what you have to do for starquad). The 20AWG 4S6 might be a better choice when using these. If you want to make a Y cable where you put TWO cables into one plug, you may just barely be able to get 4 20AWG wires into each connector. Last night was the first time I was actually able to try some of this myself. Since I am on the road now without my soldering equipment I decided to try a starquad cable with the screw terminals. My entire setup right now is an LPS-1 powering a SqueezeBox Touch (SBT) with Senheiser phones plug in to the SBT. The current setup used a simple DIY 16 AWG zip cord cable with soldered on barrel plugs. So I found a small amount of 244 on ebay and a bunch of screw terminal connectors on Amazon. They were all here yesterday. Then started the fun, the two 18AWG wires do NOT fit well in those screw terminals. Every time I pushed them in a strand or two would not go in and curl off to the side shorting out to the other terminal. I had to try it about seven times to get all 4 wires properly inserted into the plug. The wires are pretty stiff so I had to really tighten those screws down tight to keep them from pulling out. It only took five tries on the other end of the wire. I finally got it together and plugged it in, WOW this was a big improvement! I thought this system sounded amazing, but with this DC cable it was dramatically improved. And that less than 10 dollars! Next I'm going to try the 20AWG 4S6 cable and see how that sounds. John S.