|
Current through coils
Reg Edwards wrote:
I am not quibbling about what EZNEC reports but what Cecil has *said* it reports. On the other hand, Cecil may be right. But I hope he is wrong! Well, let me combine my results and John P's. MHz Degrees Delay in Freq Phase Shift nanoseconds 5.5 14.13 7.14 5.89 15.68 7.39 6 16.2 7.5 7 21.36 8.5 8 29.47 10.2 9 45.92 14.2 10 88.95 24.7 11 141.38 35.7 12 163.02 37.7 13 172.28 36.8 13.7 183.82 37.2 -- 73, Cecil http://www.qsl.net/w5dxp |
Current through coils
Roy Lewallen wrote:
EZNEC doesn't report time or propagation delay. If it's needed, it must be calculated from the reported phase angles. In my case, done by installing zero ohm loads at the bottom and top of the coil for the traveling wave antenna. -- 73, Cecil http://www.qsl.net/w5dxp |
Current through coils
Cecil,
I have sent the files to you, as requested. The resonant whip ended up at about 10 feet, not 8, because I tired of the continuing the optimization. The phase shift found from the traveling wave version was about 7 degrees. The coil was similar to the coil you have been using, except squeezed down to 6 inches tall, 4 TPI. You can refine the models or perform the required adjustments as you wish to get the 8 foot version. I would not expect radical changes, but who knows? It is always possible that I made a mistake in the modeling. If so, I am confident the entire world will know in short order. 8-) 73, Gene W4SZ Cecil Moore wrote: Here's what you guys need to do to convince me that you are right. 1. Develop a coil acceptable to EZNEC that resonates an 8 foot whip on 4 MHz over mininec ground. Send me a copy of the EZNEC file. 2. Put the same coil in the traveling wave environment as I have done at: http://www.qsl.net/w5dxp/test316y.GIF and report the phase shift through the coil. Send me a copy of the EZNEC file. If you are so right and I am so wrong, that shouldn't be too difficult for half a dozen gurus (including the developer of EZNEC). |
Current through coils
"Roy Lewallen" wrote EZNEC doesn't report time or propagation delay. If it's needed, it must be calculated from the reported phase angles. Roy Lewallen, W7EL =================================== EXACTLY! Roy, you knew exactly what I was hunting for! The phase angle should indeed increase with frequency. With a true transmission line the increase should be directly proportional to frequency. But the propagation delay, with a true transmission line, should be constant versus frequency as it depends only on coil dimensions and hence on L and C. The velocity factor should also be a constant. When the angle is 90 degrees the coil is 1/4-wave resonant. When the angle is 180 degrees it is 1/2-wave resonant. The 180 degree frequency should be twice the 90 degree freq. The differences between the 'measurements' and expectations, on the basis of a true transmission line, can perhaps be explained by the very small ratio of wire diameter to winding pitch. (Who would use such a coil to load an antenna anyway?) The ratio of coil diameter to coil length is large. Therefore end-effect (as with an antenna wire) becomes quite important. I have no idea what else may have been 'connected' to the coil when inside EZNEC. Could it have affected measurements? Would it be possible to 'measure' input impedance looking into one end of a coil with the other end disconnected? Regarding the calculated values of delay time there seems to be something wrong. There is a simple relationship between phase angle and length of coil and time. What was the formula used to calculate propagation time? If the correct formula was used does this point to somthing peculiar with EZNEC's own calculations? Can't think any more. ---- Reg. |
Current through coils
Reg Edwards wrote:
(snip) What was the formula used to calculate propagation time? Glad you asked. I calculated (1/f)*(angle/360) or period of one cycle times fraction of a period. I assumed that all the phase shifts given were less than one period, total. I was hoping someone else would calculate the delays and correct any mistakes in my method or results. If the correct formula was used does this point to somthing peculiar with EZNEC's own calculations? Can't think any more. Or it points to the coil in question not acting purely like a transmission line over this range of frequencies. How would a short length of transmission line act if you parallel it, end to end with capacitance? This coil certainly has end to end capacitance. At some high frequency, that capacitance will, first resonate with the inductance, forming a trap (with a rapid phase change with respect to frequency), and above that, bypass the antenna current around the inductance. At very low frequencies, with a low impedance on each side, almost pure inductance probably dominates. In between these, (with appropriate impedance connections), there is probably a frequency range where it acts mostly like a short piece of transmission line, with approximately (a lot less approximate than +- 60%) constant delay. In such a large beast, at the intended frequency of operation, unintended parasitic effects may be as big as the intended inductive effect. These are the problems that make designing good (as in, "act as lumps of inductance") coils interesting. |
Current through coils
On Mon, 27 Mar 2006 15:17:35 -0800, Roy Lewallen
wrote: What most people would call "self capacitance" -- the equivalent capacitance from one terminal to the other. Hi Roy, Not strictly speaking. It ("self capacitance") is with respect to a very remote reference, not merely the two plate formulation of the terminal's geometries to each other (that is more part of the distributed capacitance). There are two measures of capacitance. Self capacitance is any body's capacity to store charge. You don't need a second plate to do that in the classical math - merely a reference point from which the voltage is determined (yes, another dimensionless oddity that makes this more easily said than done). Mutual capacitance, two plate construction, is the more usual form we all have come to expect - so much so that the term mutual has fallen into disuse and most express only the second, isolated term - capacitance. Tom, a week or so back, asked about the infinitesimal capacitance of a coil with 15 meters (or so) of wire. He speculated that as the "second" plate of the (mutual) capacitance was withdrawn to infinity, that it forced the value to zero. I, on the other hand showed that in the practical universe: C = 2 · Pi · epsilon0 · L / ( ln(b/a) ) a = 1m (after all, thin is relative at infinite dimensions) L = 15m b= 1,000,000,000,000,000,000,000,...000m (10³³³ meters away) epsilon0 = 0.00000000000885 C = 12 femtofarads This was certainly at the limits of my usual Capacitor Bridge to measure to this resolution 30 years ago, but time has marched on. This sized capacitance is certainly encountered every day in my new field of nanotech, and 1 femtofarad is measured by charge transfer techniques. Consider, Einstein's estimate of the radius of the Universe is roughly 10 Billion Light Years (±3dB) As this result above is vastly further away than Einstein's guess (by more than 300 orders of magnitude), lets look at again from his number: C = 12.5 picofarads Oddly enough, this value is on par with the distributed capacitance of the coil's we've been pounding away on (and even more convergent, is this is roughly the same amount of wire used in them). I extracted this correlation from reports of the coils' self resonant frequency and their inductance. Self Capacitance is nothing more than Mutual Capacitance with a second spherical plate, with a radius of this 10 Billion Light Years. However, this capacitance is the total bulk of the coil rather than that distributed to form a transmission line. Anyway, thinking of a coil in terms of Mutual Capacitance, Distributed Capacitance, Self Capacitance, Self Inductance, and Mutual Inductance is a tantalizing prospect to investigate and elevate the topic to this mythic status of transmission line - but I seem to lack the motivation to go there. The extraordinary farce is more entertaining. 73's Richard Clark, KB7QHC |
Current through coils
John Popelish wrote: In such a large beast, at the intended frequency of operation, unintended parasitic effects may be as big as the intended inductive effect. These are the problems that make designing good (as in, "act as lumps of inductance") coils interesting. Actual time delay of current through 100 turn 10tpi 2 inch diameter ten inch long inductor: http://www.w8ji.com/inductor_current_time_delay.htm 73 Tom |
Current through coils
Gene Fuller wrote:
It is always possible that I made a mistake in the modeling. If so, I am confident the entire world will know in short order. 8-) It is also possible that EZNEC cannot be used for this task. It doesn't make technical sense that the delay through a coil can go from 7 nS at 5.5 MHz to 38 nS at 12 MHz. I think that was what was bothering Reg. -- 73, Cecil http://www.qsl.net/w5dxp |
Current through coils
wrote:
Actual time delay of current through 100 turn 10tpi 2 inch diameter ten inch long inductor: Here is probably the actual time delay through that coil. W8JI wrote: "By the way, I swept S12 phase with my network analyzer on a 100uH inductor a few hours ago while working on a phasing system. The phase shift through that series inductor was about -60 or -70 degrees on 1 MHz, ... " -- 73, Cecil http://www.qsl.net/w5dxp |
| All times are GMT +1. The time now is 09:45 PM. |
Powered by vBulletin® Copyright ©2000 - 2025, Jelsoft Enterprises Ltd.
RadioBanter.com