Home |
Search |
Today's Posts |
|
#1
![]() |
|||
|
|||
![]()
On May 9, 1:40*am, wrote:
Tom, I thought I'd quoted some numbers in the related "Dish reflector" thread - apologies if I did not. Here are some mo * I modelled a coil as a spiral in EZNEC (40T, diameter=6", length=12", #14 copper wire) * I added a 6ft "stinger" and found the frequency where the combination was resonant: 3.79 MHz * I checked the feedpoint impedance without the coil present: 0.46- j2439 * That tells me the "lumped circuit equivalent" reactance of the coil at 3.79 MHz is +j2439 ohms * I found the frequency where the coil was resonant with no "stinger": 6.2 MHz Now I look at what ON4AA's "Corum method" inductance calculator tells me: * "Lumped circuit equivalent" reactance at 3.79 MHz: +j2449 * Self-resonant frequency: 6.3 MHz Unless I'm missing an option, if I want to predict the RF characteristics of a "bugcatcher" it seems I have *3 choices: * Use Wheeler's formula * Build a helical model in EZNEC * Use the Corum method Wheeler's formula is inappropriate at frequencies close to a coil's SRF. EZNEC and the Corum method give very close results. The Corum formulas are not difficult to use; even if they were, there is an on-line calculator which removes the need for any maths. So it seems to me the Corum formulas would be the more convenient tool to use, at least for a "first look". 73, Steve G3TXQ On May 9, 7:06*am, "Tom Donaly" wrote: You know, you haven't shown that the Corum model accurately measures the bugcatcher coil. You have stated - and I have no reason to disbelieve you - that the Corum model agrees with EZNEC. If that's the case, it's just as easy to use EZNEC, right or wrong. MoM is a method of obtaining numerical solutions to integral equations. The only reason to do that is if symbolic solutions are either too difficult or impossible to puzzle out of those same integral equations. In other words, some very deep thinkers decided that MoM would give results superior to algebraic approximations and hand waving, so they applied it to antenna analysis. I don't think it's perfect. It's certainly useful. If you think Corum is good enough for your purposes, though, go for it. Steve, this is fine for a base loading coil, but I'd suggest you try your experiment with a loading coil well up the antenna, where the coil is significantly larger diameter than the straight conductor in which it's placed. The same size coil you described (though presumably a different number of turns), placed at least half way up something like a 15 or 20 foot long thin wire, should illustrate the point. Is the EZNEC model then in such good agreement with placing a reactive load at that point in the antenna, where the reactance is from ON4AA's online calculator? If I trusted NEC to handle large steps in conductor diameter accurately, I'd suggest putting a segment in the antenna description to represent the length and diameter of the coil, with the calculated reactance placed as a load in that segment. As I understand it, though, NEC has trouble with large diameter steps. Cheers, Tom |
#2
![]() |
|||
|
|||
![]()
K7ITM wrote:
Steve, this is fine for a base loading coil, but I'd suggest you try your experiment with a loading coil well up the antenna, where the coil is significantly larger diameter than the straight conductor in which it's placed. The same size coil you described (though presumably a different number of turns), placed at least half way up something like a 15 or 20 foot long thin wire, should illustrate the point. Is the EZNEC model then in such good agreement with placing a reactive load at that point in the antenna, where the reactance is from ON4AA's online calculator? The key to understanding this question and its logical answer lies in the phase shift that occurs at impedance discontinuities. For a base-loading coil, there is only one impedance discontinuity in the system, a hi-Z0 coil to a low-Z0 stinger. That single discontinuity provides a positive phase shift at the '+' junction of the coil and stinger. coil stinger FP//////////+------------------- When a straight shaft section is installed under the coil, it introduces one additional impedance discontinuity at 'x' in addition to the '+' top of coil to stinger discontinuity. base coil stinger FP-------x////////////+--------- Because the impedance discontinuity between the base section is a low-Z0 to hi-Z0 transition, the phase shift is negative, i.e. the antenna *loses electrical degrees* at that junction. Therefore, more turns must be added to the inductor to supply the number of negative degrees lost at the base section to coil impedance discontinuity. This might best be illustrated with pieces of transmission line. Please reference my web page at: http://www.w5dxp.com/shrtstub.htm The following concepts apply to the above antennas but may be easier to understand using transmission lines. Here is a dual-Z0 stub that is physically 44.4 degrees long but is 90 degrees (1/4WL) long electrically, i.e. it is functionally a 1/4WL open-circuit stub. ---22.2 deg 300 ohm line---+---22.2 deg 50 ohm line--- The Z0=300 ohm to Z0=50 ohm transition provides for +45.6 degrees of phase shift. This is akin to the base- loaded antenna above. Here is a dual-Z0 stub with 11.1 degrees (half) of the 50 ohm line moved to the left. (The words are abbreviated because of space on the line.) --11.1 deg 50--+--22.2 deg 300--+--11.1 deg 50-- Who can tell me how long electrically is this stub using the identical feedlines from the above example? This reconfigured stub with half of the 50 ohm feedline moved to the bottom is now electrically only ~80.6 degrees long. What has happened? The new impedance discontinuity from the base section at the bottom of the coil has cost us electrical degrees by providing a *negative phase shift*. How do we solve the problem? Add some length (degrees) to the Z0=300 ohm section. If we make the 300 ohm section 38.5 degrees long, the stub will be electrically 90 degrees long once again. This is conceptually the same problem we encounter when we move the loading coil from the base location to the center location. When we move the coil up the shaft, we introduce a negative phase shift at the bottom of the coil. Therefore, we must increase the number of turns to make the loading coil electrically longer. Incidentally, w8ji knows about the coil to stinger positive phase shift and describes it on his web page. He apparently doesn't know about the opposite negative phase shift at the bottom of the coil where the shaft attaches. -- 73, Cecil, IEEE, OOTC, http://www.w5dxp.com |
#3
![]() |
|||
|
|||
![]()
Tom,
OK, I tried what you suggested. I put my loading coil midway up a 20ft vertical wire in the EZNEC model. I reduced the number of turns to lift the resonant frequency to 5.6MHz. EZNEC predicted that the magnitude of the current at the top of the coil would be 77% of the magnitude at the bottom. Then I removed the coil in the model, replaced it with a straight wire containing an EZNEC lumped load, and adjusted that load for antenna resonance at 5.6MHz again. I needed +j1630. Given the dimensions of the coil, the Corum calculator predicted a lumped circuit equivalent reactance of +j1573, and it predicted a current fall-off across the coil of 78%. 73, Steve G3TXQ On May 9, 5:35*pm, K7ITM wrote: Steve, this is fine for a base loading coil, but I'd suggest you try your experiment with a loading coil well up the antenna, where the coil is significantly larger diameter than the straight conductor in which it's placed. *The same size coil you described (though presumably a different number of turns), placed at least half way up something like a 15 or 20 foot long thin wire, should illustrate the point. *Is the EZNEC model then in such good agreement with placing a reactive load at that point in the antenna, where the reactance is from ON4AA's online calculator? |
#4
![]() |
|||
|
|||
![]()
On May 9, 1:56*pm, wrote:
Tom, OK, I tried what you suggested. I put my loading coil midway up a 20ft vertical wire in the EZNEC model. I reduced the number of turns to lift the resonant frequency to 5.6MHz. EZNEC predicted that the magnitude of the current at the top of the coil would be 77% of the magnitude at the bottom. Then I removed the coil in the model, replaced it with a straight wire containing an EZNEC lumped load, and adjusted that load for antenna resonance at 5.6MHz again. I needed +j1630. Given the dimensions of the coil, the Corum calculator predicted a lumped circuit equivalent reactance of *+j1573, and it predicted a current fall-off across the coil of 78%. Hi Steve, OK, I'm wondering now exactly what "Corum calulator" you are using that predices "a current fall-off across the coil of 78%." The inductance calculator on the HamWaves website that I thought we were talking about doesn't seem to say anything about "current fall-off" in coils, though perhaps I'm missing it. Cheers, Tom |
#5
![]() |
|||
|
|||
![]()
Hi Tom,
I should have been more explicit. I took the "Axial Propagation Factor" (4.372 rad/m) figure which was given by the HamWaves calculator and multiplied it by the coil length (155mm) to find the effective electrical length of the coil (38.83 degrees). Then I took cos(38.83)=0.779 as the fall-off in current across the coil. 73, Steve G3TXQ On May 11, 2:46*am, K7ITM wrote: Hi Steve, OK, I'm wondering now exactly what "Corum calulator" you are using that predices "a current fall-off across the coil of 78%." *The inductance calculator on the HamWaves website that I thought we were talking about doesn't seem to say anything about "current fall-off" in coils, though perhaps I'm missing it. Cheers, Tom |
#6
![]() |
|||
|
|||
![]()
wrote:
I took the "Axial Propagation Factor" (4.372 rad/m) figure which was given by the HamWaves calculator and multiplied it by the coil length (155mm) to find the effective electrical length of the coil (38.83 degrees). Then I took cos(38.83)=0.779 as the fall-off in current across the coil. Interesting. W8JI's coil through which he measured a 3 nS delay was 100t, 2" dia, 10" long, #18 wire. http://www.w8ji.com/inductor_current_time_delay.htm Converting everything to metric and entering the data into the HamWaves calculator at 4 MHz, yields a calculated delay of 21.5 nS through the W8JI coil and a VF of ~0.04 at 4 MHz. So which are we to believe? W8JI's measurements or ON4AA's calculator. There's a 7x difference between 3 nS and 21 nS. -- 73, Cecil, IEEE, OOTC, http://www.w5dxp.com |
#7
![]() |
|||
|
|||
![]()
On May 11, 2:26*am, wrote:
Hi Tom, I should have been more explicit. I took the "Axial Propagation Factor" (4.372 rad/m) figure which was given by the HamWaves calculator and multiplied it by the coil length (155mm) to find the effective electrical length of the coil (38.83 degrees). Then I took cos(38.83)=0.779 as the fall-off in current across the coil. 73, Steve G3TXQ Hi Steve, OK, so I suppose you are assuming that the current distribution will follow a cosine along electrical degrees of your antenna, with a maximum at the base/feedpoint. If that's the case, then would you not account for the bottom 10 feet of wire, about 20.5 electrical degrees? If I do that and assume 1 amp at the feedpoint, I should see about .9367 amps at 20.5 degrees and 0.5101 amps at (20.5+38.83) electrical degrees. 0.5101/.9367 would then be the ratio of currents between the ends of the coil, and that's 0.5446, only a 45.54 percent fall-off. In fact, it seems to me that the idea of cos(38.83 degrees) = .779 would imply a fall-off of 22.1%... and that tells me that perhaps I'm still not understanding your model very well. Maybe you are NOT assuming the current along the electrical degrees of the antenna, up from the feedpoint, will have a cosine distribution. At this point, I have to say that I'm just not at all sure what your model really is. Perhaps you are making different assumptions about the current distribution... Also, if you still have the model around, try adding a top hat to the upper wire. For simplicity, you can just use a simple "T" structure, where the top horizontal wire is, say, five feet long total. With such a configuration, what's the current distribution along the radiating element going to be? Of course, what I'm suggesting here is that one must be careful to test ones models at corner cases before putting too much faith in them, and even then, one must always be wary of cases where the model may go awry. Cheers, Tom |
#8
![]() |
|||
|
|||
![]()
K7ITM wrote:
OK, so I suppose you are assuming that the current distribution will follow a cosine along electrical degrees of your antenna, with a maximum at the base/feedpoint. This is a good assumption for horizontal 1/2WL thin-wire dipoles as presented by Kraus. It doesn't seem to be valid for loaded vertical antennas where there is an instantaneous phase shift at the impedance discontinuities. There is a definite change in the slope of the current profile at such boundaries. And there is the nagging current bulge in the loading coil causing a rise in current in adjacent turns. Normally a current maximum would indicate a purely resistive impedance but that doesn't seem to be the case inside a loading coil. Years ago, I gave up on the current cosine argument for loaded mobile antenna current in favor of loading the coil with its characteristic impedance and using traveling wave current to measure the electrical length of the coil. -- 73, Cecil, IEEE, OOTC, http://www.w5dxp.com |
#9
![]() |
|||
|
|||
![]()
Tom,
Firstly, I'm guilty of a "sloppy" choice of words. Whenever I've been using the phrase "drop off in current" I've meant the current at the top of the coil as a percentage of the current at the bottom. So when I've quoted 70% the current will have reduced by 30%. Apologies! Secondly, you're testing the limits of my understanding with the overall current distribution from base section, through the coil, to the top section. However I think the point is that you can't simply "add electrical degrees" through the various sections when the characteristic impedances of the sections are so disparate. That was Cecil's point in the very first posting. We also know that, as expected, summing the "degrees" for the three sections gets nowhere near a total of 90 degrees, so clearly you can't assume a cosine distribution that is contiguous across all three sections. I'll investigate what happens with a "top hat". 73, Steve G3TXQ On May 11, 6:56*pm, K7ITM wrote: Hi Steve, OK, so I suppose you are assuming that the current distribution will follow a cosine along electrical degrees of your antenna, with a maximum at the base/feedpoint. *If that's the case, then would you not account for the bottom 10 feet of wire, about 20.5 electrical degrees? *If I do that and assume 1 amp at the feedpoint, I should see about .9367 amps at 20.5 degrees and 0.5101 amps at (20.5+38.83) electrical degrees. *0.5101/.9367 would then be the ratio of currents between the ends of the coil, and that's 0.5446, only a 45.54 percent fall-off. In fact, it seems to me that the idea of cos(38.83 degrees) = .779 would imply a fall-off of 22.1%... and that tells me that perhaps I'm still not understanding your model very well. *Maybe you are NOT assuming the current along the electrical degrees of the antenna, up from the feedpoint, will have a cosine distribution. *At this point, I have to say that I'm just not at all sure what your model really is. Perhaps you are making different assumptions about the current distribution... Also, if you still have the model around, try adding a top hat to the upper wire. *For simplicity, you can just use a simple "T" structure, where the top horizontal wire is, say, five feet long total. *With such a configuration, what's the current distribution along the radiating element going to be? Of course, what I'm suggesting here is that one must be careful to test ones models at corner cases before putting too much faith in them, and even then, one must always be wary of cases where the model may go awry. Cheers, Tom |
Reply |
|
Thread Tools | Search this Thread |
Display Modes | |
|
|
![]() |
||||
Thread | Forum | |||
FA: Yaesu FT-8100R like new dual band dual recieve | Equipment | |||
FA: HTX-204 Dual Bander! Like the ADI AT-600 | Swap | |||
DUAL not duel. DUH! | Swap | |||
Dual Band HT | Swap | |||
WTB: UHF or Dual band ham rig.. | Swap |