I loved the "even experienced electrical engineers forget...." in the patent.
Where are they getting these engineers??
Sigh..
Jn
popcorn time
Well he was honest: 'we were asked to to a scientific analysis for marketing purposes'.
Jan
no, just laughed at all those "experienced electrical engineers that forget "about skin and proximity effect.
The first one of interest to me was a belden one. but they talked about oval insulation, not oval conductor.
At least there was no perpetual motion involved.
jn
I was just envious my most cited patent is for an ADSL/POT's transformer hybrid, I was in disbelief that there were no priors from 1901 or something not even one office action.
BTW the references contained at least one "speaker cable as 8 Ohm T-line" patent, even 10' matter.
Yah, Goertz... Funny how much effort was spent on that clamp to connect to the striplines.
Also, I saw no mention of amp oscillation should the load decouple below unity gain frequency.
A solution which causes a problem...
jn
ps..I'm jealous that you even have patents...for all my stuff, there is typically only one other entity on the planet that would use it, so the legal beagles here don't consider anything as worthy of the money and effort of a patent.
Last edited:
Skin Effect BS certainly runs back to 1913.
I have no idea what "overloading a conductor" is supposed to mean. Apparently not "melting"; anyway we avoid that if only for its increased resistance. Maybe he thinks the electrons start to flow outside the wire at some current density? (Of course there are no electrons, or they don't matter, only fields.)
I also do not see how his hollow conductor has more cooling effect (except very short busses).
The later letter has some practical experience, and a reason why hollow bus may be preferred (stiffer) even when skin hardly matters.
Attachments
Last edited:
I recall in '74, it was explained that copper at 60 Hz lost effectiveness beyond 4 inches diameter so for higher currents they went hollow copper with 2 inch thick walls. Saved on weight as well.
For power delivery, melting isn't the issue, reliability is. Hot conductors and connections age rapidly.
Our rule of thumb is a kilo amp per square inch copper DC, for AC we follow NEC.
Jn
Hello,
If you follow NFPA 70 (NEC) you will not burn the house down.
Many jurisdictions enforce the fine print notes; no more than 2% and 3% voltage drop over feeders and branch circuits.
If the FPN's do not get you the temperature rating of the insulation is the limiting factor. 60 degrees and 90 degrees C.
Thanks DT
If you follow NFPA 70 (NEC) you will not burn the house down.
Many jurisdictions enforce the fine print notes; no more than 2% and 3% voltage drop over feeders and branch circuits.
If the FPN's do not get you the temperature rating of the insulation is the limiting factor. 60 degrees and 90 degrees C.
Thanks DT
Sources?, power lines are 4 cables per phase, each cable is spaced with damper isolator, each cable is 3 cm diameter with a steel core, and the copper strands are not hollow...
for the copper tube two inch walls, that was my E/M professor in 1974. And it was in reference to the lower voltage/high current runs at a power generator source. What you speak of is the high tension line stuff, with quarter Meg type voltages. When I did some high power research on HV lines (IOW, I googled it
), I did come across high voltage power lines made by drawing aluminum outers and a steel core, core for strength, alum for weight. Clearly, there are many cable types out there in copper and aluminum.For our high current DC runs, back in the day code didn't cover it, so we had to interpret the intent, then approach the AHJ.
Our r and d for high temp superconducting power lines uses hollow geometry cables only, as the skin effect is rather strong when conductivity is infinite.
Jn
Saved on weight, yes. And COST!!
Superficially (especially if you think "KA/in"), current can rise as square of diameter. But heat loss (cooling) rises directly as diameter.
Once you get to hot wire, you kinda need to think KA/diameter, not KA/area.
Say you run large current in a 1" wire and it gets as hot as you think is wise. Now you need to run twice the current. A 1.4" wire has twice the copper but only 1.4 times the surface area. It will have a 1.4X higher temperature rise: too hot! What you want is twice the copper *and* twice the surface area. This leads to a 2" wire with a 1.4" hole in the middle so you are not buying 4X the copper, only 2X. (The illustration in the May letter should have shown this larger outside diameter.)
There are other ways. Two 1" wires parallel will do, it enough air can pass over both wires. A hundred 0.2" wires would also do, if they were spaced-out for ample air. Any way you slice the copper, it needs to be that 2" diameter with the copper of a 1.4" wire in it. Slicing copper is expensive, but routine. For much ordinary work we use a couple or three standard conductors carefully paralleled. But for HUGE work it may (does) make sense to have copper hollowed-out to give large surface area without paying for all the copper the diameter implies.
And yes, Aluminum is often used for big current because while it is not so good per square inch, it is lighter for the same conductivity and presently much cheaper per conductivity. The downside is more careful connections to foil the oxide film, but huge copper is huge money so there is incentive.
And no we never want conductors *anywhere* near "melting, just a figure of speech (we hope). A real drawback of hot metal is increased resistance. If you figure the dry voltage-drop at 5%, but that increases to 10% at full current and heat, your furnace which should melt a ton an hour is only melting 1,795 pounds an hour and the customer get angry. Yes, we do have examples of "doubled resistance": modern PA loudspeakers "can" survive power level where their voice coil resistance doubles. Of course voltage-drive sensitivity falls off, though on high efficiency speakers not to half; and this is hotter than any wise user will drive them, more to reduce sudden failures which spoil the show. And as you say nearly all connectors suffer from thermal cycling, and most cable is rubber/PVC which melts a lot-lot easier then the metal.
Tell me about it... I had quoted up the bog standard wire for the DC part of the machine, it was about 6 million. By the time we went to buy, the cost quadrupled.
Which is the very reason there are no Bandersnachi running around on planet Earth.(a challenge...)
For our DC needs, we'll go flat copper plate roughly 2 inches thick and 15 -20 inches wide. For long runs, 40 parallel 535 kcmil conductors for each side of the circuit. But man, so many lugs and bolts.. We do have a water cooled 30kA long run as well.
For 60 cycle, once you go above 12 to 15 kiloamps, there is a diminishing return on a solid copper conductor (4 inches diameter) as a result of skin effect at 60hz. Above that current, just go with two inch walls. Yes, the current vs diameter takes an inflection at 4 inches, but that's life..
jn
Skin effect is hardly important. As discussed, when wires get hot, you need *surface area* more than *area*.
Continue my 1" wire up to 4". By area, it would carry 16X the current; by surface area, only 4X current. So it can only carry 4X the current, if current is limited by heat. Then you want 4X the copper in 16X the area: you want 3/4 hollow. 3.46" ID 4" OD pipe, 0.27" walls.
That's just an example from an assumption that a 1" conductor gets hot at high current. JN asserts that 2" walls work for him--- he does bigger stuff than I want to work with. The implication is that conductor 2" thick and cooled on one side is about as thick as is worth doing; any thicker is a waste of copper. At some point we could write an equation comparing lateral thermal resistance to lengthwise electrical resistance, then factor in acceptable temperature and cost of metal (of various shapes).
- Status
- Not open for further replies.