|
using coax shield to create a loading coil ?
The effect of the radials is a surprise. I would not expect that short radials
would work well at all? The system predicts about 9 Ohms. That is closer to what is expected and over 6 db better then the 50 Ohm reading. I will try experimenting and let you know. Thanks - Dan Frank wrote: That's correct Dan. I just wanted to systematically build up the antenna, adding a component at a time, to note where the major losses are. This was the first trial with no loading -- except for copper conductivity. From the other model you sent me it seems that any other attempts are redundant. The major losses are due to ground loss, as expected. Unfortunately this can only be overcome by increasing the length, and number, of radials -- something that is pretty well known. Also inductive loading of the radials does not seem to have any effect, except for marginally decreasing the antenna efficiency. I have been interested in installing a short monopole for 160m, so am very interested in your results. I have a fairly large lot (visible on "Google Earth), so am not so restricted in radial length. 73, Frank "dansawyeror" wrote in message ... Frank, I tried the nec below. The result was resonant at 21.9 and about 34 Ohms. I am not competent at reading nec cards yet, however the model editor does not show any coil loads. That could explain the frequency? Dan Thanks - Dan Frank's Basement 2 wrote: Dan, here is a preliminary run on a 12 ft monopole model structured as follows: base at 6 ft, 10 x 6ft radials. All #14 AWG. Ground - perfect, frequency 3.8 MHz. Zin = 0.968 - j1847.55 ohms; Efficiency = 87.4 % (structure copper loss); Gain = 4.15 dBi; Take-off angle = 0 deg; Gain at 27 deg elevation (expected TOA with real ground) = +3.09 dBi. I will try successive modifications to approach a practical model. The code I used, modified so it should run in 4nec2, is shown below. 73, Frank CM 75 m Vertical 12 ft high CM base 6 ft up, 10 X 6 ft radials CM copper conductivity CE GW 1 24 0 0 18 0 0 6 0.0026706 GW 2 12 0 0 6 6 0 6 0.0026706 GM 1 9 0 0 36 0 0 0 2 GS 0 0 .3048 GE 1 GN 1 EX 0 1 24 0 1.00000 0.00000 LD 5 1 1 144 5.8001E7 FR 0 11 0 0 3.5 0.05 RP 0 181 1 1000 -90 0 1.00000 1.00000 EN "Frank's Basement 2" wrote in message news:dhmKf.6088$_62.3050@edtnps90... Dan, The lumped inductance of 4 +j1750 comes from your previous comment about the inductance range from 60 - 90 uH. I just chose the mid range value of 75 uH at 3.8 MHz. To be exact 2*PI*f*L = 1791 ohms. The real part of 4 ohms is based on an approximate Q of 400. Incidentaly I am working at another location this morning. The computer is an old 600 MHz machine, with 384 MB of RAM, and Windows ME OS. The NEC code here takes 17 seconds to run. 73, Frank "dansawyeror" wrote in message ... I see the length is set to 1.8 meters already. A 2 meter elevation minimum is needed to lower ground effects. How is the lumped inductance set of 4 Ohms and 1750 Z? What impedance does that translate to? How did you calculate this value? Dan Frank's Basement 2 wrote: Hi Dan, thanks for the interesting info. You did not specify dimensions, but from your comments it appears you are using a vertical about 23 ft high. Such a monopole would have a 3.5 ohm input impedance when placed above a perfectly conducting ground, and gain about +4.5 dBi. Adding a center loading coil raises the input impedance to 11.5 ohms, and gain +2.6 dBi. Base loading provides an input impedance of 5.5 ohms with almost the same gain as center loading (Q = 400). Adding ten, 6ft radials, at 3" above an average ground, the input impedance increases to 40 ohms, and gain -6.3 dBi. Adding lumped element loading coils, (75 uH, Q = 400) in each radial (antenna base end) drops the input impedance to 37 ohms, and gain -6.4 dBi. Don't know why this does not agree with Reg's program. Probably I made some fundamental error with the NEC model. Included the code below, so you may see an error I missed. 73, Frank CM 75 m Vertical 23 ft high CE GW 1 64 0 0 23 0 0 0.25 0.0026706 GW 2 12 0 0 0.25 6 0 0.25 0.0026706 GM 1 9 0 0 36 0 0 0 002.002 GS 0 0 .3048 GE 1 GN 2 0 0 0 13.0000 0.0050 EX 0 1 64 0 1.00000 0.00000 LD 5 1 1 184 5.8001E7 LD 4 1 33 33 4 1600 LD 4 2 1 1 4 1750 LD 4 3 1 1 4 1750 LD 4 4 1 1 4 1750 LD 4 5 1 1 4 1750 LD 4 6 1 1 4 1750 LD 4 7 1 1 4 1750 LD 4 8 1 1 4 1750 LD 4 9 1 1 4 1750 LD 4 10 1 1 4 1750 LD 4 11 1 1 4 1750 FR 0 11 0 0 3.5 0.05 RP 0 181 1 1000 -90 0 1.00000 1.00000 EN Frank, Good morning. Let me start at the beginning. I have a loaded vertical on 75 meters. The combination of the antenna and ground measure about 40 Ohms at the antenna. The models all show such an antenna over a perfect ground should have a radiation resistance of between 3 and 4 Ohms. That says the antenna system is less the 10% efficient. This then is a journey to reduce ground resistance. Attempts to add radials and wire mesh to the ground have had very little if no effect. This leads to Reg's c_poise model. It predicts a coil in the range of 60 uH to 90 uH tuned to a 2 meter by 18 mm 'wire' will have a total resistance in the 2 to 4 Ohms range. Together this should result is a 8 Ohm system. The ratio can be directly inferred as an performance improvement of 5 to 1 or 7 db. This is worth some effort. To answer your question the first step will be one coil and one radial. The objective is the get the antenna system close to 10 Ohms. From there I will experiment with adding radials and coils. I am not sure what to expect. Thanks - Dan Frank wrote: Not sure I understand what is going on Dan. Are you planning on loading each radial element? Frank "dansawyeror" wrote in message ... These results were from Reg's c_poise program. The band is 75 meters and the coils were about 70 uH. The coils were a relatively large diameter, on the order of a meter. The wire lengths were about 20 meters. By varying the length the coil, the coil wire may be varies from 1mm to 12mm. Richard Clark wrote: On Sat, 18 Feb 2006 08:20:38 -0800, dansawyeror wrote: The devil is in the details. Modeling shows large coils with 1 mm wire have a Q in the range of a few hundred. On the other hand a coil with 12 mm tubing has a Q of about 2000. The R of the 1 mm coil is about 6 Ohms while the 12 mm coil is on the order of 1 Ohm. Given these model results it says there is a significant difference between 1 mm and 12 mm coils. Hi Dan, In the details, indeed. What is the LENGTH of wire in this 6 Ohm resistor? What is the LENGTH of wire in this 1 Ohm resistor? How many turns are in these "large coils?" What is their diameter? What is their solenoid length? Without these details, there is nothing said that is significant. 73's Richard Clark, KB7QHC |
using coax shield to create a loading coil ?
I am using 4nec2 and am getting errors from the GM card. Wasn't there an issue
with these being a decimal instead of an integer? Yes, it was and still is. 'Original' Nec2 does not understand the Necwin+ syntax for the last field in the GM card when the 'x.y' format is used. Within (modified) Necwin+ the ITS (imov)field with the x.y format does move all (preceding) wires with tags from 'x' up to tag 'y'. With default Nec2 the x.y field is rounded to an integer and all wires with tags from 'x' up to the one just preceding the GM card are moved, so no explicit 'end-tag' is specified. When using Nec4 one can specify both the start- and stop tag- and segment-numbers on four separate fields. Arie. |
using coax shield to create a loading coil ?
"Roy Lewallen" wrote in message ... Reg Edwards wrote: Roy, you seem to have forgotten proximity effect. . . . Forgotten? I just didn't see what relevance it had on the difference in Q between an inductor made from a braided coax shield and one made from solid tubing. And I can't see from your posting anything which adds to that discussion. But maybe I'm missing something? Roy Lewallen, W7EL ======================================= Yes, Roy, you are missing something. |
using coax shield to create a loading coil ?
Dan,
Will be interested to know how you get on with the antenna. I suspect that adding loading coils to radials is about the same as adding loading coils to any part of an antenna system. They do nothing to effect the radiation efficiency, only modify the input impedance. 73, Frank The effect of the radials is a surprise. I would not expect that short radials would work well at all? The system predicts about 9 Ohms. That is closer to what is expected and over 6 db better then the 50 Ohm reading. I will try experimenting and let you know. Thanks - Dan Frank wrote: That's correct Dan. I just wanted to systematically build up the antenna, adding a component at a time, to note where the major losses are. This was the first trial with no loading -- except for copper conductivity. From the other model you sent me it seems that any other attempts are redundant. The major losses are due to ground loss, as expected. Unfortunately this can only be overcome by increasing the length, and number, of radials -- something that is pretty well known. Also inductive loading of the radials does not seem to have any effect, except for marginally decreasing the antenna efficiency. I have been interested in installing a short monopole for 160m, so am very interested in your results. I have a fairly large lot (visible on Earth), so am not so restricted in radial length. 73, Frank "dansawyeror" wrote in message ... Frank, I tried the nec below. The result was resonant at 21.9 and about 34 Ohms. I am not competent at reading nec cards yet, however the model editor does not show any coil loads. That could explain the frequency? Dan Thanks - Dan Frank's Basement 2 wrote: Dan, here is a preliminary run on a 12 ft monopole model structured as follows: base at 6 ft, 10 x 6ft radials. All #14 AWG. Ground - perfect, frequency 3.8 MHz. Zin = 0.968 - j1847.55 ohms; Efficiency = 87.4 % (structure copper loss); Gain = 4.15 dBi; Take-off angle = 0 deg; Gain at 27 deg elevation (expected TOA with real ground) = +3.09 dBi. I will try successive modifications to approach a practical model. The code I used, modified so it should run in 4nec2, is shown below. 73, Frank CM 75 m Vertical 12 ft high CM base 6 ft up, 10 X 6 ft radials CM copper conductivity CE GW 1 24 0 0 18 0 0 6 0.0026706 GW 2 12 0 0 6 6 0 6 0.0026706 GM 1 9 0 0 36 0 0 0 2 GS 0 0 .3048 GE 1 GN 1 EX 0 1 24 0 1.00000 0.00000 LD 5 1 1 144 5.8001E7 FR 0 11 0 0 3.5 0.05 RP 0 181 1 1000 -90 0 1.00000 1.00000 EN "Frank's Basement 2" wrote in message news:dhmKf.6088$_62.3050@edtnps90... Dan, The lumped inductance of 4 +j1750 comes from your previous comment about the inductance range from 60 - 90 uH. I just chose the mid range value of 75 uH at 3.8 MHz. To be exact 2*PI*f*L = 1791 ohms. The real part of 4 ohms is based on an approximate Q of 400. Incidentaly I am working at another location this morning. The computer is an old 600 MHz machine, with 384 MB of RAM, and Windows ME OS. The NEC code here takes 17 seconds to run. 73, Frank "dansawyeror" wrote in message ... I see the length is set to 1.8 meters already. A 2 meter elevation minimum is needed to lower ground effects. How is the lumped inductance set of 4 Ohms and 1750 Z? What impedance does that translate to? How did you calculate this value? Dan Frank's Basement 2 wrote: Hi Dan, thanks for the interesting info. You did not specify dimensions, but from your comments it appears you are using a vertical about 23 ft high. Such a monopole would have a 3.5 ohm input impedance when placed above a perfectly conducting ground, and gain about +4.5 dBi. Adding a center loading coil raises the input impedance to 11.5 ohms, and gain +2.6 dBi. Base loading provides an input impedance of 5.5 ohms with almost the same gain as center loading (Q = 400). Adding ten, 6ft radials, at 3" above an average ground, the input impedance increases to 40 ohms, and gain -6.3 dBi. Adding lumped element loading coils, (75 uH, Q = 400) in each radial (antenna base end) drops the input impedance to 37 ohms, and gain -6.4 dBi. Don't know why this does not agree with Reg's program. Probably I made some fundamental error with the NEC model. Included the code below, so you may see an error I missed. 73, Frank CM 75 m Vertical 23 ft high CE GW 1 64 0 0 23 0 0 0.25 0.0026706 GW 2 12 0 0 0.25 6 0 0.25 0.0026706 GM 1 9 0 0 36 0 0 0 002.002 GS 0 0 .3048 GE 1 GN 2 0 0 0 13.0000 0.0050 EX 0 1 64 0 1.00000 0.00000 LD 5 1 1 184 5.8001E7 LD 4 1 33 33 4 1600 LD 4 2 1 1 4 1750 LD 4 3 1 1 4 1750 LD 4 4 1 1 4 1750 LD 4 5 1 1 4 1750 LD 4 6 1 1 4 1750 LD 4 7 1 1 4 1750 LD 4 8 1 1 4 1750 LD 4 9 1 1 4 1750 LD 4 10 1 1 4 1750 LD 4 11 1 1 4 1750 FR 0 11 0 0 3.5 0.05 RP 0 181 1 1000 -90 0 1.00000 1.00000 EN Frank, Good morning. Let me start at the beginning. I have a loaded vertical on 75 meters. The combination of the antenna and ground measure about 40 Ohms at the antenna. The models all show such an antenna over a perfect ground should have a radiation resistance of between 3 and 4 Ohms. That says the antenna system is less the 10% efficient. This then is a journey to reduce ground resistance. Attempts to add radials and wire mesh to the ground have had very little if no effect. This leads to Reg's c_poise model. It predicts a coil in the range of 60 uH to 90 uH tuned to a 2 meter by 18 mm 'wire' will have a total resistance in the 2 to 4 Ohms range. Together this should result is a 8 Ohm system. The ratio can be directly inferred as an performance improvement of 5 to 1 or 7 db. This is worth some effort. To answer your question the first step will be one coil and one radial. The objective is the get the antenna system close to 10 Ohms. From there I will experiment with adding radials and coils. I am not sure what to expect. Thanks - Dan Frank wrote: Not sure I understand what is going on Dan. Are you planning on loading each radial element? Frank "dansawyeror" wrote in message ... These results were from Reg's c_poise program. The band is 75 meters and the coils were about 70 uH. The coils were a relatively large diameter, on the order of a meter. The wire lengths were about 20 meters. By varying the length the coil, the coil wire may be varies from 1mm to 12mm. Richard Clark wrote: On Sat, 18 Feb 2006 08:20:38 -0800, dansawyeror wrote: The devil is in the details. Modeling shows large coils with 1 mm wire have a Q in the range of a few hundred. On the other hand a coil with 12 mm tubing has a Q of about 2000. The R of the 1 mm coil is about 6 Ohms while the 12 mm coil is on the order of 1 Ohm. Given these model results it says there is a significant difference between 1 mm and 12 mm coils. Hi Dan, In the details, indeed. What is the LENGTH of wire in this 6 Ohm resistor? What is the LENGTH of wire in this 1 Ohm resistor? How many turns are in these "large coils?" What is their diameter? What is their solenoid length? Without these details, there is nothing said that is significant. 73's Richard Clark, KB7QHC |
using coax shield to create a loading coil ?
Frank's Basement 2 wrote:
I suspect that adding loading coils to radials is about the same as adding loading coils to any part of an antenna system. They do nothing to effect the radiation efficiency, only modify the input impedance. Wherever did you get that idea? A dipole made out of 80m hamsticks is less than 1% efficient. -- 73, Cecil http://www.qsl.net/w5dxp |
using coax shield to create a loading coil ?
"Cecil Moore" wrote in message . com... Frank's Basement 2 wrote: I suspect that adding loading coils to radials is about the same as adding loading coils to any part of an antenna system. They do nothing to effect the radiation efficiency, only modify the input impedance. Wherever did you get that idea? A dipole made out of 80m hamsticks is less than 1% efficient. Well, I guess I should have said: "... do nothing to effect the radiation efficiency, except possibly reduce it". Don't know anything about "Hamsticks", but they must have lousy loading inductors! Frank |
using coax shield to create a loading coil ?
Wherever did you get that idea? A dipole made out of
80m hamsticks is less than 1% efficient. Well, I guess I should have said: "... do nothing to effect the radiation efficiency, except possibly reduce it". Don't know anything about "Hamsticks", but they must have lousy loading inductors! Frank To be exact; a 16 ft dipole would need to be loaded with inductors of Q = 78 for a 0.99% efficiency. The gain is therefore -18.3 dBi. Frank |
using coax shield to create a loading coil ?
Frank wrote:
To be exact; a 16 ft dipole would need to be loaded with inductors of Q = 78 for a 0.99% efficiency. The gain is therefore -18.3 dBi. Sounds about right. Hamsticks are about 12 dB down from a good screwdriver on 75m. -- 73, Cecil http://www.qsl.net/w5dxp |
using coax shield to create a loading coil ?
"Cecil Moore" wrote in message
. com... Frank wrote: To be exact; a 16 ft dipole would need to be loaded with inductors of Q = 78 for a 0.99% efficiency. The gain is therefore -18.3 dBi. Sounds about right. Hamsticks are about 12 dB down from a good screwdriver on 75m. -- 73, Cecil http://www.qsl.net/w5dxp I wonder what the efficiency of a Miracle Whip is? 73, Frank |
nec simulation - unexpected result ??
All,
I have been experimenting with various loaded antennas to use in my relatively limited space. For this I assumed the two arms of a dipole must be identical to support resonance, this assumption has not been supported by modeling. Actual model runs show that if the two arms of a dipole are close then there is sufficient interaction that they will combine to form a single resonance. The model below shows a simple example of this. The loads and length of the arms are not equal, however nec predicts a single resonance at about 3.5 MHz. Changes of 10 to 20 percent around resonance seem to create one resonance. Is there an explanation for this? Thanks - Dan kb0qil CM 75 m loaded dipole CM copper conductivity CE GW 1 31 4 0 8 0 0 8 .001 GW 2 11 0 0 8 -6 0 8 .001 GE 0 LD 4 1 16 16 3 2500 LD 4 2 5 5 3 2000 EX 0 1 30 0 1 0 GN 2 0 0 0 13 5.e-3 FR 0 1 0 0 3.543 0 EN Frank wrote: "Cecil Moore" wrote in message . com... Frank wrote: To be exact; a 16 ft dipole would need to be loaded with inductors of Q = 78 for a 0.99% efficiency. The gain is therefore -18.3 dBi. Sounds about right. Hamsticks are about 12 dB down from a good screwdriver on 75m. -- 73, Cecil http://www.qsl.net/w5dxp I wonder what the efficiency of a Miracle Whip is? 73, Frank |
nec simulation - unexpected result ??
dansawyeror wrote:
I have been experimenting with various loaded antennas to use in my relatively limited space. For this I assumed the two arms of a dipole must be identical to support resonance, this assumption has not been supported by modeling. Is there an explanation for this? An electrical 1/2 wavelength conductor is resonant no matter where you feed it. Even if you don't feed it anywhere, it is still resonant. -- 73, Cecil http://www.qsl.net/w5dxp |
nec simulation - unexpected result ??
"dansawyeror" wrote in message
... All, I have been experimenting with various loaded antennas to use in my relatively limited space. For this I assumed the two arms of a dipole must be identical to support resonance, this assumption has not been supported by modeling. Actual model runs show that if the two arms of a dipole are close then there is sufficient interaction that they will combine to form a single resonance. The model below shows a simple example of this. The loads and length of the arms are not equal, however nec predicts a single resonance at about 3.5 MHz. Changes of 10 to 20 percent around resonance seem to create one resonance. Is there an explanation for this? Thanks - Dan kb0qil No matter what Dan, you should see only one resonance at the overall length of a half wave. Not that resonance has any bearing on antenna efficiency. Also; your NEC model has uneven segmentation, which does produce significant errors. Interesting to note that your antenna is also resonant at 7 MHz. 73, Frank |
nec simulation - unexpected result ??
"a 1/2 wave segment is resonant no matter where you feed it." That allows for a
large single coil to 'tune' one arm of an antenna and for the other to be adjustable. Simulation predicts the impedance will change when it is not feed at the center, it appears to go up as the feed point is moved. I will play with the segmentation and see what happens. Thanks - Dan Frank wrote: "dansawyeror" wrote in message ... All, I have been experimenting with various loaded antennas to use in my relatively limited space. For this I assumed the two arms of a dipole must be identical to support resonance, this assumption has not been supported by modeling. Actual model runs show that if the two arms of a dipole are close then there is sufficient interaction that they will combine to form a single resonance. The model below shows a simple example of this. The loads and length of the arms are not equal, however nec predicts a single resonance at about 3.5 MHz. Changes of 10 to 20 percent around resonance seem to create one resonance. Is there an explanation for this? Thanks - Dan kb0qil No matter what Dan, you should see only one resonance at the overall length of a half wave. Not that resonance has any bearing on antenna efficiency. Also; your NEC model has uneven segmentation, which does produce significant errors. Interesting to note that your antenna is also resonant at 7 MHz. 73, Frank |
nec simulation - unexpected result ??
On Thu, 23 Feb 2006 21:40:31 -0800, dansawyeror
wrote: "a 1/2 wave segment is resonant no matter where you feed it." Hi Dan, I don't know where to start on that one. 1/2 wave "segment?" And then to partition (into what? it is already describe as A segment) for a feed - that is resonant irrespective of where it is fed? Any wire is resonant, further elaboration does nothing to change that one obscure characteristic - and in fact, any wire is multi-resonant. That allows for a large single coil to 'tune' one arm of an antenna and for the other to be adjustable. Then it ceases to be "a 1/2 wave segment" unless the frequency is adjusting with the length - this would seem to be obvious, but what end is served in saying it? What distinguishes an arm from a segment? Simulation predicts the impedance will change when it is not feed at the center, Simulation should. it appears to go up as the feed point is moved. In distinct contradiction to most OCF dipoles - odd. In fact one of the hallmarks of the OCF is being resonant in many ham bands where the standard dipole does not. 73's Richard Clark, KB7QHC |
nec simulation - unexpected result ??
dansawyeror wrote:
"a 1/2 wave segment is resonant no matter where you feed it." That allows for a large single coil to 'tune' one arm of an antenna and for the other to be adjustable. Simulation predicts the impedance will change when it is not feed at the center, it appears to go up as the feed point is moved. I will play with the segmentation and see what happens. Absolutely true! But, what does feedpoint impedance have to do with resonance? ... NUTTIN! |
nec simulation - unexpected result ??
Amos Keag wrote:
"But, what does feedpoint impedance have to do with resonance?" Imagine a whip worked against ground. It is resonant at 1/4-wavelength where it presents a low impedance. It is resonant again at 1/2-wavelength where it presents a high impedance. Best regards, Richard Harrison, KB5WZI |
nec simulation - unexpected result ??
dansawyeror wrote:
Simulation predicts the impedance will change when it is not feed at the center, it appears to go up as the feed point is moved. An off-center-fed dipole will match 300 ohm twin lead if fed at the correct point. This is covered in my 1957 ARRL Handbook. -- 73, Cecil http://www.qsl.net/w5dxp |
nec simulation - unexpected result ??
"dansawyeror" wrote in message
... "a 1/2 wave segment is resonant no matter where you feed it." That allows for a large single coil to 'tune' one arm of an antenna and for the other to be adjustable. Simulation predicts the impedance will change when it is not feed at the center, it appears to go up as the feed point is moved. I will play with the segmentation and see what happens. Thanks - Dan Dan, As for NEC segmentation. "0.05 wavelengths per segment is preferred, but can be as long as 0.1 wavelengths. Segments shorter than 0.001 wavelengths should be avoided". From L. B. Cebik's "Basic Antenna Modeling: .....". In most cases all segments within a structure should have equal length segmentation. Where antenna models become very large, with 1000 segments or more, there are work-arounds which allow for uneven segmentation without introducing errors. The problem with large numbers of segments is that processor time increases dramatically. I see nothing wrong with using one loading coil in a dipole. The effect is simply the same as an off-center-fed dipole. 73, Frank |
nec simulation - unexpected result ??
Richard Harrison wrote:
Amos Keag wrote: "But, what does feedpoint impedance have to do with resonance?" Imagine a whip worked against ground. It is resonant at 1/4-wavelength where it presents a low impedance. It is resonant again at 1/2-wavelength where it presents a high impedance. Best regards, Richard Harrison, KB5WZI Resonance has NOTHING to do with impedance. Resonance is resonance; it has a harmonic response. Feed point impedance is the load presented to a transmission line when you want to make a wire, any wire, resonant or non resonant, into an antenna. I can feed any antenna with a single wire against ground, I can feed the same antenna with 50 ohm coax, 70 ohm coax, 90 ohm coax, 72 ohm balance line, 300 ohm balanced line 450 ohm balanced line 600 ohm balanced line. None of these transmission lines changes to resonance or non resonance of the antenna. Resonance is determined by the physical characteristics of the antenna. Generally these include the antenna length and the length to diameter ratio. PERIOD. |
nec simulation - unexpected result ??
On Fri, 24 Feb 2006 11:48:22 -0500, Amos Keag
wrote: Resonance has NOTHING to do with impedance. Resonance is resonance; it has a harmonic response. Hi Amos, Resonance is the absence of reactance, or more properly its term is 0. As reactance is fully part of the specification to impedance, resonance has a very unique relation: r ±j0. You can take a dipole that exhibits this unique characteristic at regular intervals of frequency - notably at harmonics (in a perfect world, not so necessarily in life). You can also take that same length of wire and shift the feedpoint such that its resonance (still that same characteristic loss of X with some remaining R) changes in frequency - as does the spectrum of other resonances which are sometimes no longer related by harmonics. Taking as an example, an 11 segment 3mm wire 37.9M long in free space and feed it in the conventional way (in the middle) and its resonances may be observed at: 3.8 MHz¹ 7.95 MHz² 11.75 MHz¹ 16.25 MHz² 19.65 MHz¹ 24.65 MHz² 27.55 MHz¹ Or feed it at 68% along its length (or segment 8) and observe: 3.8 MHz¹ (with a Higher R as I had incorrectly argued with Dan) 5.65 MHz² 7.85 MHz¹ 12.45 MHz² 15.65 MHz¹ 17.55 MHz² 19.65 MHz¹ 25.55 MHz² 27.55 MHz¹ where strictly speaking MHz¹ is resonance and MHz² is anti-resonance A curious property has emerged, we now have 9 resonances (speaking largely) where formerly we had 7 in exactly the same span of frequency for the same piece of wire. Further, we also have the anti-resonance of the standard dipole at 8 MHz replaced by a resonance in the OCF dipole. To roll back the calendar 10 years or so, this is also the hallmark of fractal antennas in that they exhibit more resonances than found in "conventional" dipoles. There are certain lengths of wire, with certain offsets of feed that offer fairly good overlaps with Ham Bands that are not otherwise found in common dipoles. I am at a loss to specify those "certain" characteristics, and it is arguable that feeding an offset dipole can be successfully achieved without some effort in isolating (choking) the feedpoint from the driveline - a distasteful reality conveniently discarded in modeling. 73's Richard Clark, KB7QHC |
nec simulation - unexpected result ??
Amos Keag wrote:
But, what does feedpoint impedance have to do with resonance? On a standing wave antenna, like a center-fed dipole, the feedpoint impedance is (Vfor+Vref)/(Ifor+Iref) where Vfor is forward voltage, Vref is reflected voltage, etc. and the plus sign denotes superposition, i.e. phasor addition. On a wire dipole, the resonant feedpoint impedance will occur only when all the phases line up, i.e. If Vfor is at zero degrees, Vref must be at 180 degrees, Ifor must be at zero degrees, and Iref must be at zero degrees. That way, we get minimum voltage divided by maximum current with a resultant phase angle of zero degrees. Eureka! The dipole is resonant because the feedpoint impedance is purely resistive. -- 73, Cecil http://www.qsl.net/w5dxp |
nec simulation - unexpected result ??
On Fri, 24 Feb 2006 19:28:37 GMT, Cecil Moore wrote:
Amos Keag wrote: But, what does feedpoint impedance have to do with resonance? On a standing wave antenna, like a center-fed dipole, the feedpoint impedance is (Vfor+Vref)/(Ifor+Iref) where Vfor is forward voltage, Vref is reflected voltage, etc. and the plus sign denotes superposition, i.e. phasor addition. On a wire dipole, the resonant feedpoint impedance will occur only when all the phases line up, i.e. If Vfor is at zero degrees, Vref must be at 180 degrees, Ifor must be at zero degrees, and Iref must be at zero degrees. That way, we get minimum voltage divided by maximum current with a resultant phase angle of zero degrees. Eureka! The dipole is resonant because the feedpoint impedance is purely resistive. Cecil, are you saying that a resonant dipole must have a low impedance, and that modes where the feedpoint impedance is purely resistive but high are not "resonant"? That seems to be what your formulae above and explanation suggests. Owen -- |
nec simulation - unexpected result ??
Amos Keag wrote:
Resonance has NOTHING to do with impedance. jX is only zero at resonance. -- 73, Cecil http://www.qsl.net/w5dxp |
nec simulation - unexpected result ??
Owen Duffy wrote:
Cecil, are you saying that a resonant dipole must have a low impedance, and that modes where the feedpoint impedance is purely resistive but high are not "resonant"? That seems to be what your formulae above and explanation suggests. Yes Owen, that's what I am saying. When I was at Texas A&M in the dark ages, we called the feedpoint impedance of a one- wavelength dipole an "anti-resonant" impedance. It is explained at: http://dx.doi.org/10.1036/1097-8542.041800 "Antiresonance - The condition for which the impedance of a given electric ... system is very high, approaching infinity." Semantics strikes again. To distinguish the left-most low resistance point on an SWR circle from the right-most high resistance point, we mid-20th-century Aggie EEs called the leftmost point, "resonant", and called the rightmost point, "anti-resonant". If you and I were ever to agree on definitions, I have no doubt that we would also agree on concepts. -- 73, Cecil http://www.qsl.net/w5dxp |
nec simulation - unexpected result ??
On Fri, 24 Feb 2006 20:52:11 GMT, Cecil Moore wrote:
Semantics strikes again. To distinguish the left-most low resistance point on an SWR circle from the right-most high resistance point, we mid-20th-century Aggie EEs called the leftmost point, "resonant", and called the rightmost point, "anti-resonant". If you and I were ever to agree on definitions, I have no doubt that we would also agree on concepts. I am sure we are talking the meaning of the terms (semantics) rather than the underlying concept. Narrowing the term resonance to only apply to the resonance that exhibits a series resonance equivalent behaviour seems to me to unnecessarily limit the meaning of resonance (though I note it is used in optics to some extent). I think of the high impedance of a dipole with zero reactance at some frequencies also as a resonance, and I think you did too when you said recently to Amos "jX is only zero at resonance." If that is to mean that jX is "only ever" zero at resonance, then if jX is zero, you have resonance, whether R is high or low. On that basis, one would have to say that a full wave centre fed dipole exhibits (at the feed point) resonance similar to a lossy parallel tuned circuit and should be considered a resonant radiator. Owen -- |
nec simulation - unexpected result ??
Amos Keag wrote:
Resonance has NOTHING to do with impedance. Resonance is resonance; it has a harmonic response. . . . Resonance has everything to do with impedance. Resonance is defined as any frequency at which the impedance is purely resistive; that is, where the reactive part of the impedance is zero. And that is all resonance is. You can change the resonant frequency of an antenna by simply adding a series or parallel inductor or capacitor at the feedpoint. This doesn't change the antenna characteristics in any other way than to alter the feedpoint impedance. Roy Lewallen, W7EL |
nec simulation - unexpected result ??
Owen Duffy wrote:
On that basis, one would have to say that a full wave centre fed dipole exhibits (at the feed point) resonance similar to a lossy parallel tuned circuit and should be considered a resonant radiator. I know that is what your gut feeling wishes were true. But a large portion of the RF engineering community considers "anti- resonance" to be the exact opposite of "resonance" and indeed it is the exact opposite on a Smith Chart, being the opposite side of the SWR circle. Semantics strikes again. I'm sure that our Russian counterparts have a completely different word for exactly the same effects. -- 73, Cecil http://www.qsl.net/w5dxp |
nec simulation - unexpected result ??
Roy Lewallen wrote:
Resonance has everything to do with impedance. Resonance is defined as any frequency at which the impedance is purely resistive; ... In the distant past, when I had a dinosaur for a pet, resonance was defined as the frequency at which the impedance is a purely low impedance. The frequency at which the impedance was a purely high resistance was known at the anti-resonant point, the exact opposite of resonance, and indeed, it was the exact other side of the SWR circle on a Smith Chart. -- 73, Cecil http://www.qsl.net/w5dxp |
nec simulation - unexpected result ??
Egad. Calling it antiresonance is asking for (communications) trouble,
since not eveyone uses the same terms. Just call it "half-wave resonance" and "full-wave resonance". I don't think I've EVER heard anyone call a parallel-tuned circuit "anti-resonant." I do regularly hear people distinguish between series and parallel resonance, however. I'm not likely to soon adopt "antiresonance" for either condition, as it sounds way too much like something opposing resonance. Cheers, Tom |
nec simulation - unexpected result ??
K7ITM wrote:
I'm not likely to soon adopt "antiresonance" for either condition, as it sounds way too much like something opposing resonance. In a transmission line with reflections, antiresonance is indeed plus or minus 90 degrees from resonance and "never the twain shall meet". Resonance and antiresonance cannot, by definition, occur at the same point, i.e. if a point is antiresonant, it cannot, by definition, be resonant. Quoting "Transmission Lines and Networks", by Walter C. Johnson, PhD. (one of the heavyweight gurus of the mid-20th- century) page 156: "When the lossless line is an odd number of quarter wavelengths long, the sending-end impedance is theoretically infinite (inversion of the receiving-end impedance). The actual impedance, considering losses, is a very large resistance, and the line is said to be ANTIRESONANT." (Capitals substituted for italics for obvious reasons) So your argument is with Walter C. Johnson, PhD, ex-chairman of the Department of Electrical Engineering at Princeton University, not with me. -- 73, Cecil http://www.qsl.net/w5dxp |
nec simulation - unexpected result ??
If the feed is 'changing the resonance' then there is a problem with the
feed!! Richard Clark wrote: On Fri, 24 Feb 2006 11:48:22 -0500, Amos Keag wrote: Resonance has NOTHING to do with impedance. Resonance is resonance; it has a harmonic response. Hi Amos, Resonance is the absence of reactance, or more properly its term is 0. As reactance is fully part of the specification to impedance, resonance has a very unique relation: r ±j0. You can take a dipole that exhibits this unique characteristic at regular intervals of frequency - notably at harmonics (in a perfect world, not so necessarily in life). You can also take that same length of wire and shift the feedpoint such that its resonance (still that same characteristic loss of X with some remaining R) changes in frequency - as does the spectrum of other resonances which are sometimes no longer related by harmonics. Taking as an example, an 11 segment 3mm wire 37.9M long in free space and feed it in the conventional way (in the middle) and its resonances may be observed at: 3.8 MHz¹ 7.95 MHz² 11.75 MHz¹ 16.25 MHz² 19.65 MHz¹ 24.65 MHz² 27.55 MHz¹ Or feed it at 68% along its length (or segment 8) and observe: 3.8 MHz¹ (with a Higher R as I had incorrectly argued with Dan) 5.65 MHz² 7.85 MHz¹ 12.45 MHz² 15.65 MHz¹ 17.55 MHz² 19.65 MHz¹ 25.55 MHz² 27.55 MHz¹ where strictly speaking MHz¹ is resonance and MHz² is anti-resonance A curious property has emerged, we now have 9 resonances (speaking largely) where formerly we had 7 in exactly the same span of frequency for the same piece of wire. Further, we also have the anti-resonance of the standard dipole at 8 MHz replaced by a resonance in the OCF dipole. To roll back the calendar 10 years or so, this is also the hallmark of fractal antennas in that they exhibit more resonances than found in "conventional" dipoles. There are certain lengths of wire, with certain offsets of feed that offer fairly good overlaps with Ham Bands that are not otherwise found in common dipoles. I am at a loss to specify those "certain" characteristics, and it is arguable that feeding an offset dipole can be successfully achieved without some effort in isolating (choking) the feedpoint from the driveline - a distasteful reality conveniently discarded in modeling. 73's Richard Clark, KB7QHC |
nec simulation - unexpected result ??
Dot wrote:
Looking at your definitions I would suggest that "resonance" is really the point at which the antenna mimics a series resonant circuit, exhibiting a low impedence and "anti-resonance" is the point at which it mimics a parallel resonant circuit, exhibiting a high impedence. True, but I cannot take credit for the definition which comes from "Transmission Lines and Networks", by Walter C. Johnson, PhD, guru and chairman of the Department of Electrical Engineering, Princeton University during the 1940's and 1950's. -- 73, Cecil http://www.qsl.net/w5dxp |
nec simulation - unexpected result ??
Purely resistive load = resonance ............ PERIOD
Cecil Moore wrote: Owen Duffy wrote: On that basis, one would have to say that a full wave centre fed dipole exhibits (at the feed point) resonance similar to a lossy parallel tuned circuit and should be considered a resonant radiator. I know that is what your gut feeling wishes were true. But a large portion of the RF engineering community considers "anti- resonance" to be the exact opposite of "resonance" and indeed it is the exact opposite on a Smith Chart, being the opposite side of the SWR circle. Semantics strikes again. I'm sure that our Russian counterparts have a completely different word for exactly the same effects. |
nec simulation - unexpected result ??
Cecil Moore wrote:
In a transmission line with reflections, antiresonance is indeed plus or minus 90 degrees from resonance and "never the twain shall meet". Resonance and antiresonance cannot, by definition, occur at the same point, i.e. if a point is antiresonant, it cannot, by definition, be resonant. One more thought: In a transmission line with reflections, a voltage node is located at a point of resonance. A voltage anti-node is located at a point of anti-resonance. Makes perfect sense to me. -- 73, Cecil http://www.qsl.net/w5dxp |
nec simulation - unexpected result ??
Amos Keag wrote:
If the feed is 'changing the resonance' then there is a problem with the feed!! Not at all. As you can see at: http://www.qsl.net/w5dxp/notuner.htm, I use "the feed" for the specific purpose of changing the resonant frequency of the antenna system. The impedance transforming series- section is really a series stub which resonants the entire antenna system. -- 73, Cecil http://www.qsl.net/w5dxp |
nec simulation - unexpected result ??
On Fri, 24 Feb 2006 20:12:42 -0500, Amos Keag
wrote: If the feed is 'changing the resonance' then there is a problem with the feed!! Hi Amos, Where do you see that in the data? Or are you mis-interpreting the distinction between the choice of the feed point with attaching a feed line? The model distinctly lacks a feed line, or may be presumed to have a feed line that is completely isolated from the antenna(s). The wire retains its original fundamental resonance - within a couple dozen KHz, a negligible difference. The remainder of its resonances are strictly governed by the selection of the feed point's position along the length of the wire. I incorrectly argued for a lower Z with Dan earlier. By and large, moving away from the center raises the Z (principally R at resonances above the fundamental). 73's Richard Clark, KB7QHC |
nec simulation - unexpected result ??
Dot wrote:
On Fri, 24 Feb 2006 22:41:54 GMT, Cecil Moore wrote: Roy Lewallen wrote: Resonance has everything to do with impedance. Resonance is defined as any frequency at which the impedance is purely resistive; ... In the distant past, when I had a dinosaur for a pet, resonance was defined as the frequency at which the impedance is a purely low impedance. The frequency at which the impedance was a purely high resistance was known at the anti-resonant point, the exact opposite of resonance, and indeed, it was the exact other side of the SWR circle on a Smith Chart. These days, resonance is described as either: a) the point at which Inductive Reactance and Capacitive Reactance are equal or b) the point at which a load impedence is purely resistive. The two points are exactly the same. Looking at your definitions I would suggest that "resonance" is really the point at which the antenna mimics a series resonant circuit, exhibiting a low impedence and "anti-resonance" is the point at which it mimics a parallel resonant circuit, exhibiting a high impedence. The high-impedance full-wave resonant point (for a dipole; half-wave resonant point for a monopole) is sometimes called "anti-resonance", but not commonly, and mostly in older literature. It's a true point of resonance, that is, where the reactance is zero. I don't believe I've ever heard the term "anti-resonance" applied to other high-impedance resonant circuits, such as a tank circuit. It would then be reasonable for a given wire perpendicular to a good ground plane to exhibit "resonance" at odd multiples of a quarter wavelength and "anti-resonance" at even multiples of a quarter wavelength... Translating gives low impedence at odds and high impedence at evens, which is where I started out in this discussion.... If you choose to call the high-impedance resonant points "anti-resonance", that's true. But again, they're points where the reactance is zero, just like the points you're calling "resonant". The only difference is that the impedance is high and the antenna acts more like a parallel tuned circuit at nearby frequencies rather than a series tuned circuit. Your semantics is correct if you are looking to define an antenna as "a current fed device", but that's not always the case. There are end fed half waves out there... they are voltage fed, they are resonant and they do work. (Ask anyone who owns a "Ringo Ranger".) No, the definition of resonance has nothing to do with how an antenna is fed. The impedance of the antenna doesn't change with the feed method (assuming of course that it has a single feed point), and therefore its resonant frequencies don't change with the feed method. (You can, of course, alter the resonant frequencies of an antenna *system* by adding reactance at the feedpoint or elsewhere.) And an antenna doesn't have to be resonant (that is, have a non-reactive feedpoint impedance) to "work". Resonance is only an indication of the reactance of the input impedance, and has nothing to do with an antenna's gain, pattern, bandwidth, or other performance characteristics. Roy Lewallen, W7EL |
nec simulation - unexpected result ??
Roy Lewallen wrote:
The impedance of the antenna doesn't change with the feed method ... One feed method is center feed. Another feed method is off-center feed. The feedpoint impedance of the antenna changes with position since for 1/2WL, for instance, the net voltage is a sine wave referenced to the center, and the net current is a cosine wave referenced to the center. The feedpoint impedance is approximately sin(x)/cos(x)=tan(x) where 'x' is the number of degrees away from center. -- 73, Cecil http://www.qsl.net/w5dxp |
nec simulation - unexpected result ??
Geez, Cecil, I don't have an argument with either of you. I'm just
telling you that the people I work with qualify resonance with different terms than you do. You're welcome to use whatever terms you want. Cheers, Tom |
nec simulation - unexpected result ??
K7ITM wrote:
Geez, Cecil, I don't have an argument with either of you. The Devil made me do it. :-) -- 73, Cecil http://www.qsl.net/w5dxp |
| All times are GMT +1. The time now is 07:12 PM. |
Powered by vBulletin® Copyright ©2000 - 2025, Jelsoft Enterprises Ltd.
RadioBanter.com