The power line to your outlets should never be less than #12 in order to maintain the 20-amp (max) rating. Most commercial grade #12 wiring is THHN solid core copper. If you buy a cord with #10's, then you are effectively limiting the power draw to your equipment (conversely, it also makes sense that cord gages fatter than #12 do nothing to improve power flow from the incoming service to the user).
As far as stranded vs solid core, the only difference is the solid core wires have a lower resistivity per foot than stranded - which is why power wiring is usually specified as solid core.
Here's where I get in trouble: power cords with gages equal or bigger than the power wiring size do nothing to improve the sound of your system with respect to current flow; the sonic differences are the result of shielding RFI/EMI only. The PVC insulation, copper purity, etc. might make a difference with respect to signal carrying; however there's no effect on power delivery. |
For wire diameter, the LOWER the number, the BIGGER the wire. Therefore 10 gauge is thicker than 12 gauge which is thicker than 16 gauge. |
1) #10 is heavier than #12, so he would therefore be lowering the series resistance and voltage drop. This would NOT create a bottleneck in the system as you suggest. Wire gauges work BACKWARDS from what most people think i.e. 8 gauge is thicker / lower resistance than 12 gauge. The number may be lower, but the wire is bigger.
2) Your posting your opinion as fact. Although this is not uncommon. I probably do it myself every day and will probably end up doing it here. None the less, the other side of the coin needs to be represented. As usual, if someone disagrees with what i have to say, PLEASE post your comments. I am not above learning or making mistakes.
3) Your comments about using heavier wire than #12 being useless is incorrect. If you are putting any type of heavy load on a cable i.e. running it near its' rated current load, a heavier gauge will ALWAYS have lower voltage drop and thermal losses. The ratings on wire are a compromise at best. While they do take in such things as conductivity, thermal loss, current capacity, total length, they trade off what is "acceptable" as an industry standard for what is "optimal" from an end users point of view.
Try passing a continuous 30 amp load through some 10 gauge wire at length of 25' foot or longer and take measurements at the far end. As you would probably guess, 25' is actually a short run when you consider how house wiring is indirectly routed through conduit, etc... Then make those same measurements using 8 gauge. You will still see voltage sag using 8 gauge. Not until you're using 6 gauge will the voltage drop be to a point that is minimal.
Keep in mind that most of the devices that i work with that are pulling this type of current are RF amplifiers i.e. just like audio amps but at a higher frequency. They output wattage and low distortion is a requirement. Bottom line is that the results of voltage sag are VERY measurable in terms of maximum clean output and stability when running at or near capacity. I know this, as i get to see it all the time when testing various components on the bench.
Due to the fact that i work with electrical components that can sometimes pull HUNDREDS of amps, i've had to learn ( the hard way ) that you can never have TOO HEAVY of a wire on a high current device. That is, so long as the Device Under Test ( DUT ) is well built and actually designed to run at / near rated output for extended periods of time. Otherwise, voltage sag can act as a slight buffer, keeping the device from running on the ragged edge and blowing up. Then again, if someone is worried about optimizing the power supply feeding all of the components, i doubt that they've invested in equipment that is sub-standard or not capable of meeting spec.
4) As to your comments about using heavier gauge wire in a power cord not accomplishing anything, i disagree. Lowering series resistance i.e. "voltage drop" over any length of cable or through a connection is always beneficial. Of course, we are all shooting for optimum performance, so any deviation from "optimum" is considered a bad thing.
Whether or not someone wants to pay for "optimum" performance or the conditions necessary to achieve that may be another story. The fact that most of the benefits of "better wire" ( in any situation ) come from using logic, reasonably priced wiring and knowing how to geometrically configure it for best performance in that specific situation. There are commercially available pre-manufactured power cables that offer RFI / EMI geometries that are extremely reasonable in terms of price per foot. Surfing through the Belden website can produce some very cost effective alternatives and provide a wealth of knowledge at the same time. Knowing some basic electricity / electronics while doing this doesn't hurt either : )
5)If you're looking for the most bang for the buck, use the heaviest gauge solid core twisted pair wire that you can find. I know that several different wire / cable distributors normally stock this type of item in at least 12 gauge. You may be able to find 10 gauge, i don't know. You then run a seperate solid wire of AT LEAST equal or ( preferrably ) heavier gauge for the ground. This gives you the following benefits:
Higher capacitance per foot, resulting in greater filtration of unwanted frequencies.
Lower inductance to minimize the cables from acting like a long wire antenna.
The shortest / most direct / lowest resistance path to ground.
6) Some wire distributors will actually produce reasonable quantities of cable for you. One local distributor has their own twisting / braiding machine. You can tell them that you want to twist two or three solid gauge number 10 cables of your brand / make choice and they can do it. They will charge you for the extra labor, but it is FAR cheaper than buying some "audiophile grade" power wiring. Obviously, some places will have a minimum run in terms of footage, but you can sometimes make a deal in terms of paying slightly higher labor costs to do a short run. By studying various mass produced commercial cable designs and their benefits, you can duplicate their performance for pennies on the dollar using this method of manufacture. Sean
> |
just for my input. i think most of us who have opinions are good people. sometimes we make mistakes.. oops.. but i don't think spanking the individual publicly goes too far |
Thank god it wasn't a grounding question:~) |
Sean, thanks for the mini lesson in power cabling. I just bought a new house and plan to have a electrician install dedicated lines. I just surfed the belden site and it's all greek to me. Can you tell me exaxtly what to instruct the electrician to do? Here's what I know. Dedicated ground, 20 amp, star wired, not series. What wire (exactly) should I buy? I also plan to use better ac plugs like PS audio or the FIM. I've heard about whole house surge protection at the panel-what do I need to tell him here? Also, is there different quality in circut breakers too? If so, what's the type/brand to get? Anything else? Thanks in advance...John |
I'd like to start by saying I generally agree with what Sean is saying above. I would also like to mention that my profession is Commercial Electrician in Silicon Valley Ca. I have around twenty years experience and have installed many dedicated circuits for all sorts of sensitive equipment.
I'd like to start with a practical look at voltage drop. Why don't we plug in some real numbers that might be used for a stereo or H/T system installation. . What if we say the load is a continuos 15 amps the distance is 150 linear feet from the source (panel) copper wire is being used (k=12.9) Here's how that looks for an electrician trying to size the wire;
VD= 2 x k(resistance x circular mils/1000') x distance x load /(divided by) circular mils
VD= 2 x 12.9 x 150' x 15amps /6530(#12wire) = 8.88 or(7.4%@120volts)
VD= 2 x 12.9 x 150' x 15amps /10380(#10wire)= 5.59 or(4.6%@120volts)
VD= 2 x 12.9 x 150' x 15amps /16510(#8wire) = 3.51 or(2.9%@120volts)
VD= 2 x 12.9 x 150' x 15amps /26240(#6wire) = 2.21 or(1.8%@120volts)
I'd like to note that the allowable voltage drop from the panel to the load is 8% (NEC) So in fact a #12 wire could be used legally in this scenario.
One point I would like to make is the diminishing returns on money invested verses "bang for the buck" Running beldon wire from your panel to your load seems a little extreme to me. (If you do go this route plan on running conduit to protect the wire).
I personally would put the money into isolated grounding or dedicated neutrals.
If you buy wire and twist it together you will need to run conduit to provide protection for that wire. This may be worth the extra expense, as there appears to be some sonic benefits.
MC CABLE has the wire already twisted inside but the ground is twisted in with the hot and neutral. I'm not sure if this would hinder optimum performance? (Comments welcome)
Bumping up the wire size is a good idea. However most electricians are going to think your nuts (Myself included) if you ask them to terminate a #6 wire on a 20 amp receptacle. Especially one that is less than 200' from the panel. #8 would be the most my imagination will allow for a twenty-amp circuit.
THHN is an industry standard and personally it's all I ever use. If you go with romex I think they are making it with THHN insulation around the conductors these days though it use to be TW.
I am of the opinion that it would be better to run two or three dedicated circuits with #10 wire than one with #8 or #6 when you drop the size of the load on the circuit you will reduce the voltage drop. That is to say a 5amp load will have less VD on a #10 wire than a 15-amp load.
As far as breaker selection always match the breaker to the panel manufacturer. I personally like to use the full size breakers when ever possible.
My intention here is not to bash anyone’s comments but instead to lay down some practical wiring methods and common sense approaches. I couldn't agree with Sean more when he said "Weather anyone would want to pay for optimum performance is another story" There is a lot better ways to spend your money in this hobby so lets not be to over zealous on the dedicated circuits.
I hope this helps:~)
|
A relatively inexpensive alternative would be to run 10 ga. Romex in a 3 wire configuration (~60$ for 250' @Home Depot). Compared to 2 wire Romex, all the conductors in 3 wire are twisted, and contains two hot wires (black and a red), a neutral (white), and a bare ground. Use just the black for hot, white for neutral, and the bare for ground. The red wire isn't attached to anything, and is snipped off at each end. However, one could experiment with grounding the red wire at the panel, or independent ground, and determine if it acts like a shield drain. |
Glen, thanks for taking the time to produce the figures and data that you did. It is good to find someone that works in this specific field support some of my basic statements. My line of thinking is that the figures that you provided would probably result in a "best case scenario" and would actually measure slightly worse in a real world installation. Then again, you are the electrician and should know better. I know that many "formulas" put you in the ballpark but you have to finesse the end results from there. Much of my work is DC based and at lower voltage, making line loss even more pronounced.
There is one thing that i would like to point out about your chart though. I have no idea as to what the National standards are, but i know that here in Illinois, a standard 15 amp circuit is run using 14 gauge wire. Glen's chart starts out using 12 gauge, which is standard for a 20 amp circuit. Given the figures quoted, we can see that both gauges of wire are being run "at the ragged edge" of efficiency under those circumstances. If we look at how "in-effecient" 12 gauge is at 15 amps, think about how 14 gauge would work in that situation. This would easily put us at over a 10 percent voltage drop. This is NOT including poor connections or being tiered off of another line.
With the same line of thinking, exponentially calculate the increased losses trying to run 20 amps through 12 gauge. It puts you at about the same 10% power loss ratio. Keep in mind that once you start to heat a wire due to current flow, the wire is no longer operating within its' "window of efficiency" and loss goes up drastically after that point. All of this would meet code in a standard housing development ( at least locally here in Illinois). That is why i stated that "the ratings on wire are a compromise at best".
If you can afford to throw away 10% of the dynamic range or can enjoy your system with 10% compression due to a similar percentage in voltage drop, then stick with what MOST electricians will tell you is "sufficient". Admittedly, this would constitute pulling pretty hard on an individual circuit, but it is not out of the question. People running multiple amps, a large HT system or a LOT of front end equipment might be in this situation and not even know it.
As for me, i like "overkill" and being covered under any situation. Going "overkill" will never leave you guessing as to whether or not it might be "just a bit" better later on. You will never have to second guess yourself and say "maybe i shoulda did XXXXXXX in the first place". Peace of mind in this "over the top" hobby is not easy to come by, so take what you can get when the opportunity presents itself and is under your control.
As to wire suggestions, i'm looking into that. Belden makes a TON of commercial wires. As Glen mentioned, using some of their "fancy" wiring would be WAY overkill in my book. I have used some very simple twisted pair solid conductor wire in the past with great results. I do not know what brand it was, as my Brother used to work at Anixter and would purchase it as we needed it. It was a standard stock item for them though. It is basically standard 2 conductor THHN house wire that the manufacturer twists at the factory, saving you the time and trouble. Believe me, you DO NOT want to try to twist heavy solid conductors for ANY length whatsoever. NOT fun and plenty tough. We then ran a seperate STRAIGHT ground wire ( no braiding, twisting, etc ).
Glen's suggestions are all quite valid, especially those of doing multiple circuits instead of trying to force everything onto one big one. I think that one circuit for the front end components and one for the amp would be excellent and a BIG step up for most people, especially those in older houses. Another way to divide the two circuits would be digital / analogue, etc... If you were running multiple amps i.e. large HT systems, multi-amping, etc, you might want to even consider three or more circuits. They are not THAT costly, especially since you are doing them at one time. Believe me, you'll want to get the most out of your service call. Just ask Glen about Union rates : )
One thing to remember about all of this though: it is VERY important to remember that ALL of these circuits be tied into a common ground. Otherwise, you may end up with an even worse noise floor / hum than you had before. If you want to isolate the "audio ground" from the other ground in the house / building for noise reduction purposes, so be it. Just don't overlook the importance or safety factors involved. Sean
> |
Great posts, Sean and Glen.
Following previous advice from several of you (inc Bob C) I installed two simple 10AWG Romex lines for my 300w idle (3 amps?) Aleph monos, and two dedicated 12AWG fancy Teflon/braided Belden 83802 for analogue (tuner and pre: less than 50w) and digital (CDP: 18w). I used ACME's silver-plated duplexes mounted in those separate "outdoor" junction boxes so I wouldn't have to dig up any walls.
I then made PCs with the leftover 83802 (like those DIY Vansevers?) with Schurters and Pro-Leviton, and am VERY happy with the improved dynamics...especially at lower listening levels. Seems that here's more BREATH in the air without having to increase gain.
As the lines are only 25' long from panel to junction boxes, I imagine that 10AWG was overkill for the Aleph 2s, but I decided to save bucks there rather than use the fancier Belden 12AWG 83802.
I'm considering getting another 100' roll of 83802 and selling off lengths of it ($2.50/ft+$5sh) if any of us want to use this excellent stuff.
Referring back to the high-current example used for voltage drop calculations: is the use of 15 amps continuous current draw (1800watts!) really necssary?? Can anyone REALLY justify using any wire bigger than 10AWG? I had a helluva time using 10AWG to gang together duplexes in a junction box. Semms nearly impossible with 8AWG, and I think that 12AWG would've been sufficient. I understand vaguely that "musical peaks demands" and all that may require temporarily high current draws, but isn't it the purview of power supply caps and/or Class A operation to limit the power available anyway? IOW would anyone ever really see that 8% voltage droop with real-world amplifiers and 12AWG?
1800 watts is a lotta consumption!
I remember that my decade-old NAD amp boasted 370w/ch for 20 millisec from its 100w/ch steady-state rating, and remember that it sported a 16-18AWG PC! So what's up...I don't remember the PC getting at all warm, even on sustained organ pedal.
Sorry to drag on, but I wanted to ntroduce a real-world example into this discussion.
Thanks. Ernie |
My thinking is much like Sean. When I chose to do runs for my stereo system, I contracted 14 dedicated 20 amp circuits. This allows a dedicated run, with its own breaker, providing power to a single duplex. Often with only one male plugged into it. As Sean says this is a bit over the top, but Romex is cheap.
I sometime wonder if I should have used heavier gauge, my runs are 6, 10 and (minimum) 12 gauge, depending on the location.
The 6 gauge runs supply two 240 volt lines. These have option of 120 volt phase one and 120 volt phase two inside the same box. Consists of a Hubbell 240 Volt female twist lock plus two Hubbell 20 amp hospital grade.
The runs are all less than 30 feet, most are supplying a single device. When discussing "bang for the buck" this kind of overkill is a bargain. Between marginal and the ultimate, there is less money spent than the price for a single audiophile AC cord. |
Although I personally don't know the differance, when I was previously reading up on dedicated lines, it was suggested to use S00W 10/3 wire. Hope this helps.
Paul W |
Ernie, you have to remember that some amplifiers are not very efficient to start off with i.e. pure Class A or very rich A/B, so they pull a lot of power. Especially if we are talking a couple of hundred watts output. Couple this with in-efficient speakers, music that sounds best when played "loud" and you now have to stand on the throttle. Power consumption can go WAY up REAL fast. Having said that, i think that you did a fine job of selecting what was appropriate for your system. I see no problems with what you did and think that it is probably quite sufficient. The fact that you seperated your digital and analogue can only help things out.
As to your "real world" comments, they probably apply to the mass majority of end users. Most people are enjoying basic systems that are fine under normal operating conditions. However, i am speaking from first hand experience and i know that more than a few others can relate. Try running an entire system with a BIG amp and "thirsty" speakers from a standard 15 or 20 amp outlet and watch the lights on the equipment ( and possibly in that and other rooms ) flickering to the beat.
Keep in mind where i am coming from though. I have 6400+ watts RMS feeding my HT system, 4800+ watts RMS coming from a mono-blocked tri-amp system, 1600+ watt mono-blocks driving some very inefficient full ranges in the computer room, 200+ watts driving some monitors in the bedroom AND a "whopping" 30 wpc tube amp driving some horns in the basement. Hardly a "normal" installation, but then again, i've never been "normal" : )
As you can see, being able to deliver enough power to all of these was a major concern of mine. While Albert has me beat by a long-shot, he too has come to appreciate the differences that "add up". I know that Bob Bundus and a few others have installed similar systems and have noticed definite improvements when all was said and done. |
THANKS for the spiff responses. Haven't been able to find two conductor twisted solid core 10AWG copper. Have found stranded 8AWG two conductor twisted silver plated copper w/ teflon insulation (don't know if its FEP or TFE). Not sure if the 8AWG will connect to a Hubble 20amp outlet properly tho'.Also would like to know why twisted is better and if stranded is okay. The proposed 20amp ac line will only be for one (2 plug) outlet for a Coda 11 (100/100 pure class a) and a Classe cdp.5. Not sure if an 8' buried isolated ground rod can safely eliminate the need to share a ground w/ all the other circuits on the breaker box. If the 8AWG is(safe) overkill that's okay, eliminating the AC as a potentially weak link will be a big relief....hmmm, do fuses really sound better than breakers? |
