Yet another SStran AMT3000 review

Well, I hope Daniel is OK with letting us drift away from his great kit review, but Daniel, you apparently said all that needed to be said in your review so here we go off topic.

Just a question regarding the near field vs. far field. It has been stated that in the near field the energy is stored and the field is reactive. If this is similar to a tank circuit, then the energy has to go somewhere from storage, such as back to the antenna or circuit or is dissipated. What happens to this stored energy?

Also, the radiated E and H are at right angles in space and in phase in time and are perpendicular to the direction of radiation (Poynting's vector). Does the time or space phase change in the near field?

Thanks for any help on this as I try to picture in my mind how this works.

Neil

Kraus's definition of the boundary between the near and far field makes good physical sense. I have the first edition of Kraus (1950), which, unfortunately, does not have the information contained in the figure in Kraus 3.

A person I sometimes refer to in this Forum is Prof. Valentin Trainotti, who, I think, is the leading authority on short medium-wave AM antennas still extant. In "Short Low-and Medium-Frequency Antenna Performance" in the October, 2005 issue of IEEE Antennas and Propagation Magazine, Trainotti defines a "Radiating Hemisphere" a half wavelength in radius. He says that, "The antenna consists not only of the conductive wires, but a hemispherical free space wave generator a half wavelength in radius. The area of the Earth under this hemisphere is...very important, because all of the conductive currents flowing through it are part of the antenna's circuit..." Trainotti considers the circle under the hemisphere, which is a half wavelength in radius, to be the boundary of the near field. The energy exiting the radiating hemisphere is almost entirely radiated. Inside the hemispere, some of the energy does not propagate, but returns to the short antenna to form a standing wave.

I calculated the vertical electric field for a distance of lambda/2 from a differential current element and found that the real part of the field exceeds the imaginary part by a factor of only 2.824. So, even at a distance of a half wavelength, some of the near field still exists. Trainotti makes the important point, however, that the area a half wavelength from the short antenna forms part of the transmitter and antenna circuit, and the earth in this area should be highly conductive for good efficiency.

The Poynting vector does not apply to the near field, because very little of the energy of the fields near the antenna flows away from the antenna. Instead, the displacement current generated by the electric field mostly returns to the antenna. A very short antenna is basically a capacitor.

Ermi,

Thanks for the great answer to what I asked. Then it appears that the energy can go back to the antenna/transmitter. This must have an interesting effect on the antenna Z. I do recall my study of the Poynting vector had to do with the propagated wave and it was easily calculated as a vector cross product, but do not recall that it applies to the near (as I was taught the induced) field. Seems to be the case.

I admit that I am not well read on the literature but what you posted is helpful.

Thanks again,

Neil

"Well, I hope Daniel is OK with letting us drift away from his great kit review, but Daniel, you apparently said all that needed to be said in your review so here we go off topic."

More than ok with it, Neil. There's only so much that can be said that's meaningful when reviewing a kit. I have some projects in mind now that the kit is built, but those would go in a new topic anyway for sake of clarity.

On the other hand the direction the topic has "drifted" is fascinating and gives a better picture for visualizing what's actually going on when a part15 transmitter is operating. I may not jump in real often when the tech-savvy folks here like Rich and Ermi are discussing the theory and formulas, but that doesn't mean I'm not reading and also plugging in some values to try the equations. This type of discussion is one of the things I like best about part15.us and I see little (if any) of it elsewhere when part 15 at the hobby level is being discussed.

Now, you likening the near field effect to being in some respects like a tank circuit cleared up a lot of what "stored" may mean for me. Thanks, Neil. And Ermi's explanation of the "radiating hemisphere" and it being part of the transmitter/antenna circuit also fills in pieces of the picture. Taken in light of the quote Rich gave from Kraus,

"The situation is like that inside a resonator with high-density pulsating energy accompanied by leakage which is radiated."

..at least an approximate idea of the principles involved forms which gives a better understanding of what is happening and why the usable signal from a small transmitter seems to drop off so rapidly after a certain distance.

So I'm more than "Ok", I'm very much enjoying the direction the topic has gone in.

Daniel

I think that the discussion so far has been mostly on-topic. Rattan reported excellent results with his completed SSTRAN kit while transmitting to AM receivers inside his house. This is a common experience. Good signals can be obtained inside the house even if the antenna is much shorter than three meters, and there is no ground connection. This is why Part 15 AM devices were originally used as "phono oscillators," which allow a phonograph to make use of the audio power amplifier in an AM receiver. The fact that the near field of a short antenna has a minimum radius not limited by the short antenna length helps explain why signals inside a residence of typical size are so strong.

There was some difference in opinion about how far the near field extended, and so much of the discussion was to determine the extent of the near field.

The other discussions about noise in signals beyond the near field (including comparison between noise in AM and FM systems) were, I think, also on-topic because they relate to performance the SSTRAN might have when used outside with an antenna using a loading/tuning coil and a ground.

Part 15 AM systems are small, but they are not therefore simple. The environments they are used in can make them even more complex, in principle, than commercial AM systems. Perhaps the antenna of a commercial transmitter can (for example) be located more than a half wavelength from the nearest obstruction, as is recommended, but that is not going to happen if a Part 15 AM transmitter is operating in a typical back yard. So, if a discussion in the Forum appears to get "technical," this is justified because of the difficulties and complexities under which the Part 15 operator must labor.

Here's another digression: Part 15 AM can be used across all of the broadcast band, and even slightly below. Part 15 AM operators stick to the upper end of the band because of the much higher antenna efficiency there. However, the near field extends a lot further at the lower end of the range because of the longer wavelengths there. Maybe Part 15 AM is not hopelessly bad at the low frequencies.

Poynting's theorem applies to the transmission of energy by electromagnetic fields from one place to another. Since the waves in the near field are mostly stationary, there is little transmission of energy there. It is in the far field that energy is transmitted, and that is where the Poynting vector is applicable.

Poynting's theorem applies not only to antennas, but to transmission lines. It even applies to DC circuits. The transmission of energy from a battery to a resistor along wires can be calculated by determining the electric and magnetic fields along the transmissin path, and using the Poynting vector. I would not recommend such a complicated method for solving such a simple problem, however.

Ermi, Neil, Daniel, et al

Below are addditional snips on this topic that may add to the general understanding.

1. From F. E. Terman's RADIO ENGINEERS' HANDBOOK, 1st edition, pp 771-772:

Fields in the Vicinity of an Antenna - Induction Fields

The electric induction field becomes proportionately stronger than the magnetic induction field as the distance to the antenna becomes less in the case of a doublet antenna. With a loop antenna, the magnetic induction field is increasingly predominant as the distance becomes less.

2. From Kraus' ANTENNAS, 3rd edition, page 40:

For a 1/2-wave dipole antenna, the energy is stored at one instant of time in the electric field, mainly near the ends of the antenna near maximum charge regions, while 1/2 period later the energy is stored in the magnetic field mainly near the center of the antenna or maximum current region.
____

Also in response to Neil's comments, "Then it appears that the energy can go back to the antenna/transmitter. This must have an interesting effect on the antenna Z..." From Kraus and other authors, and also by NEC analysis, the base impedance of a series-fed, 1/4-wave vertical monopole above a perfect ground plane is exactly 1/2 that of a 1/2-wave dipole in free space (36.5 vs 73 ohms).

Rich

Actually, Ermi,

"Here's another digression: Part 15 AM can be used across all of the broadcast band, and even slightly below. Part 15 AM operators stick to the upper end of the band because of the much higher antenna efficiency there. However, the near field extends a lot further at the lower end of the range because of the longer wavelengths there. Maybe Part 15 AM is not hopelessly bad at the low frequencies."

...happens to be something I was already thinking on and why the current direction of the discussion has been of particular interest. One lambda at 1700 khz is about 176 meters.. while one lambda at say 530 khz is about 565 meters.

Now using the formula Rich quotes from Balanis of lambda/2*pi, that'd give the outer boundary for the near field as approx 92 ft for 1700 khz and approx 295 ft at 530 khz.

Using Ermi's calculation of 0983634*(lambda), it'd be approx 57 ft for 1700 khz and about 182 ft for 530 khz.

Either way, it's a difference of over 300% in the radius of the near field, which is where (as Ermi Roos noted) it is not unusual for good signals to be obtained. Also, if I recall correctly, 300 ft (very close to Rich's formula when applied to 530 khz) is what the "Talking House" claims as it's minumum distance. Or if you take the 295 ft that Rich's formula yields and the 183 ft that Ermi Roos's formula gives (at 530 khz) and "split the difference", you'd be at about the 250 ft the FCC estimates for a part 15 AM.

Now, as I understand what has been said about the near field, it would be where we'd expect reasonable reception with the infamous "cheap portable radio". A good car receiver is considerably more sensitive and could pick up the signal at a greater distance, BUT...as it moves progressively further from the near field and out beyond the "radiating hemisphere" that Trainotti places at lambda/2, the signal (if it can be received at all) will be primarilly groundwave. And since the part 15 AM radiator is extremely short, any propagation much beyond the near field "bubble" would be more dependent on the ground of the part15 station than on the length of the very short radiator?

Am I understanding the basics correctly here, or am I way off base?

I may have an anecdotal point from the testing of the kit which would seem to support the idea of near field increasing as frequency is lowered (and wavelength increases). When I first powered up the transmitter I tried the high frequency end of the AM band. Reception on a very poor clock radio on the other end of the house was so-so, but that may have been due to the radio picking up strong local stations all the way to the top of the band (it's not very selective, it's a GPX clock radio that has seen better days). In any case, I was picking up hum and other stations on more than one receiver in the house, not due to the SStran, since the hum was present even when the transmitter was off. As I tuned down the band, the hum also decreased and the "bleed" from the local stations all became negligible at about 1 mhz. All the local stations are at 1230 khz and up, so I started thinking at least in my area the mid to lower end of the band might be better in terms of having less chance of interfering with people trying to listen to the local commercial stations. But the cheap clock radio definitely seemed to receive better at the lower frequency as well. Might have been due to other factors, but with this discussion I'm thinking it might be because it would be (by Ermi's formula) close to the edge of the distance for near field at the high end of the band, but well inside the near field at 1 mhz. (1 mhz near field = approx 97 ft by Ermi's formula)

In any case, definitely a lot of good food for thought in this topic.

Daniel

Daniel wrote: Either way, it's [the difference between 530 and 1600 kHz] a difference of over 300% in the radius of the near field, which is where (as Ermi Roos noted) it is not unusual for good signals to be obtained.

Other thoughts about this -- assuming that the goal is to be a "community broadcaster," then the most important consideration is to maximize the far field. In that respect the upper frequencies on Part 15 AM have a definite advantage, as antenna system radiation efficiencies up there are higher. Here are some numbers to illustrate.

For a 3-meter, series-fed, base-loaded, vertical monopole with its base at earth level, 30 ohms total coil and r-f ground loss, and 50 mW of applied power:

At 540 kHz, ERP = 0.0202 mW, and the inverse distance groundwave field at 1 km is 0.043 mV/m.

At 1650 kHz, ERP = 0.1878 mW, and the field at 1 km is 0.130 mV/m.

The distance of 1 km was chosen to get well into the far field, yet be well within the coverage radius one might hope to achieve.

Ground conductivity begins to reduce the inverse distance field value of the groundwave at 1 km, and those losses increase with frequency. But for most such short-path situations that would not be enough to favor the 540 kHz system. A groundwave propagation path long enough for conductivity losses to favor the 540 kHz signal would mean that the amount of 540 kHz signal left would be too low to be useful, anyway.

So while neither frequency in the above comparison provides the robust signals probably wished for, it is clear that the system on 1650 kHz likely has the better performance for the purposes of Part 15 AM.

Another factor here is that many home AM receivers have better r-f sensitivity at high frequencies than low ones.

Something to ponder...

Rich

If the goal is to be a community broadcaster, the goal appears to me to be an elusive one for Part 15 AM. The estimates of radiated power in Rich's post seem high. Based on discussions on another recent thread, the estimate of about 20 uW is about right for a fairly good installation at the upper end of the band, but not at 540 kHz. About all 540 kHz has going for it is the larger near field radius. One of the things that would reduce radiated power at 540 kHz, in addition to the smaller radiation resistance, is appreciably increased loading coil loss resistance.

What is lacking is good data. Field strength readings are needed for knowing what the actual radiated power of a Part 15 AM installation is. About all we have now is an FCC field strength reading in an enforcement action for a transmiiter and antenna combination for which the input power to the final transmitter stage was measured. The cost of a field strengt meter that can be used for FCC qualification testing would be prohibitive. However, WEAK AM's blog has a link to plans for a field strength meter that might be used for troubleshooting a transmitter and antenna system. The author's suggestion of using an AM station of known power with a quarter wave antenna as the calibrating signal source seems a little crude, and has already been commented on by Rich. However, this might be good enough for using the field strength meter for rough performance testing.

Ermi writes: The estimates of radiated power in Rich's post seem high. Based on discussions on another recent thread, the estimate of about 20 uW is about right for a fairly good installation at the upper end of the band, but not at 540 kHz.

Yet this is the way the numbers fall for the conditions I gave.

In the thread where ~ 20 µW ERP was discussed, the r-f output power from the tx applied to the antenna system input terminals is an unknown -- even though the relevant FCC NOUO gave a measured tx PA DC input power of 1200 mW in that setup. But of necessity by the measured field it produced, the power accepted by the antenna system would have been very low, probably indicating a poor impedance match between the tx output and the input terminals of the antenna system (including whatever matching coil was included). In that case most of the output power of the tx was reflected by its mis-matched load impedance, and dissipated as heat in the output stage of the transmitter.

OTOH, my calculations used 50 mW of applied (tx output) power to each antenna system and each antenna system had a perfect impedance match to the transmitter, so as to compare their far-field coverage potentials on 540 kHz and 1650 kHz with other things equal.

What I have gathered from several FCC NOUOs, and the Part 15 AM tx certification data I've read is that the radiated power from these tx+antenna systems with 90-100 mW of DC input power to the final r-f amplifier shows some combination of suprisingly low DC-rf conversion efficiency in the tx PA, a poor impedance match to the antenna system, and high r-f ground loss.

One of the things that would reduce radiated power at 540 kHz, in addition to the smaller radiation resistance, is appreciably increased loading coil loss resistance.

I'll be glad to re-run the numbers using whatever coil loss and r-f ground loss anyone wants to consider.

What is lacking is good data.

True for measured data, but probably the real performance of Part 15 AM setups can be calculated with higher accuracy than it can be measured.

Of course that calls for accurate numbers for the calculation entries, and they need to be based on measured data and/or experience. Due to the skills and instrumentation needed and the low powers involved in Part 15, none of this is easy to get for the average hobbyist, unfortunately.
//

I agree that the results follow from the assumpions. It is the assumptions that I think are wrong. The 50% efficiency assumption is already known to be wrong. Nobody seems to have a transmitter with efficiency that high. Transmiter efficiency can be in the single digits, or less, with poor impedance matching to the antenna. Ground resistance is totally unknown, but it is likely to be high because it is hard to avoid a poor ground plane in a Part 15 AM installation. It could easily be a lot more than 100 ohms. The loading coil resistance can be unknown. This, at least, can be found if Q can be measured. 20 ohms is not bad at the high end of the band. It could be over 100 ohms at the low end, since a good loading coil is difficult to make at 540 kHz.

This is why I think it is important to get field strength measurements from installations. I have a couple of breadboards containing Signetics NE 602s, which I used in homodyne receivers I tested years ago. I'll try to make a crude field strength meter to get some real data about the efficiencies in actual installations.

Ermi writes: I agree that the results follow from the assumptions. It is the assumptions that I think are wrong. The 50% efficiency assumption is already known to be wrong.

The comparison I made shows the signal strength differences due to the choice of two frequencies and the resulting change in the radiation resistance of the 3-m monopole, other things equal. That was the purpose of my comparison.

Of course the coil, matching and ground loss also vary with frequency, and to get the complete story they would have to be included, as you point out.

The loading coil resistance can be unknown. This, at least, can be found if Q can be measured. 20 ohms is not bad at the high end of the band. It could be over 100 ohms at the low end, since a good loading coil is difficult to make at 540 kHz.

Your coil resistance values may be a bit too pessimistic. The link below leads to a rather comprehensive, on-line calculator for air-core coils. Here are the two possible results from it, for the reactances needed to resonate a 3-m, 1/2" OD monopole on 1650 and 540 kHz.

For 1650 kHz:

Coil OD = 125 mm
Coil length = 175 mm
Wire OD = 2 mm (hard-drawn Cu)
No. of turns = 66
Reactance =~ 3,000 ohms
AC resistance = < 4 ohms

For 540 kHz:

Coil OD = 150 mm
Coil length = 300 mm
Wire OD = 1 mm (hard-drawn Cu)
No. of turns = 210
Reactance =~ 9,000 ohms
AC resistance = < 13 ohms

The coil form used to wind these coils on would have some effect, which is not included in these calculations. Also other parameter combinations may have even lower resistance, but I didn't spend a lot of time looking for them.

Here is a new comparison of these two Part 15 AM systems using the above values of coil resistance, a matched transmitter output power of 5 mW, and your 100 ohms of r-f ground resistance:

At 540 kHz, ERP = 0.0005 mW, and the inverse distance groundwave field at 1 km is 0.007 mV/m.

At 1650 kHz, ERP = 0.0054 mW, and the field at 1 km is 0.022 mV/m.

Below are the values repeated from my original post, using my original assumptions (30 ohms total of coil and ground loss for each):

At 540 kHz, ERP = 0.0202 mW, and the inverse distance groundwave field at 1 km is 0.043 mV/m.

At 1650 kHz, ERP = 0.1878 mW, and the field at 1 km is 0.130 mV/m.

Interesting that the 1650 kHz signal at 1 km is ~3X higher than the 540 kHz signal in both of these scenarios.

So what to think? I guess one conclusion could be that The Truth lies somewhere between your approach and mine.

Link to the coil calculator: http://hamwaves.com/antennas/inductance.html

Rich

I just wanted to say thank you for the bits you posted from Terman and Kraus a few posts ago. They *did* help a lot in getting a clearer idea of what was meant by "stored" and how the fields are alternately present on different parts of the antenna. Great stuff.

It did get me to wondering about the feasibility of small loop antennas for part15 AM though. Loop antennas are usually low impedance comparatively and might they present a better "match" to a part15 AM transmitter's output than a loaded vertical? I used to tinker with VLF listening and I seem to recall reading a mention that "eccentric loops" (meaning loop antennas that are taller than they are wide) approach the characteristics of a standard vertical as they become more eccentric. Basically that a loop antenna acts more like a loop if it's square or circular or some regular polygonal shape between, but as it's made narrower (like a tall rectangle) the characteristics including impedance become more like a vertical. I wonder if it might be possible to use something like that to get a better natural impedance match to a part15 transmitter so that the efficiency in terms of how much gets from the rf finals to a radiating surface could be improved somewhat?

Daniel

Daniel wrote: Loop antennas are usually low impedance comparatively and might they present a better "match" to a part15 AM transmitter's output than a loaded vertical?

But a single turn loop with a circumference of 3 meters (to meet Part 15.219) will have an extremely small radiation resistance in the AM broadcast band, much smaller than the r-f loss resistance of the conductor itself. This results in very poor radiation efficiency -- Kraus shows about 0.009% for a 1 MHz loop of about this size. For comparison, a 3-m monopole system at the high end of the broadcast band with coil and ground losses totaling 100 ohms has an efficiency of around 0.1%.

So even with perfect impedance matching to a transmitter in each case, for a given power the 3-m monopole will have much better performance.

A small loop also has pattern nulls perpendicular to the plane of the loop, so for vertical polarization it does not have uniform radiation in the horizontal plane.

I used to tinker with VLF listening and I seem to recall reading a mention that "eccentric loops" (meaning loop antennas that are taller than they are wide) approach the characteristics of a standard vertical as they become more eccentric. Basically that a loop antenna acts more like a loop if it's square or circular or some regular polygonal shape between, but as it's made narrower (like a tall rectangle) the characteristics including impedance become more like a vertical.

From what I've read the performance of a loop is about the same for loops of any shape having the same area.

I wonder if it might be possible to use something like that to get a better natural impedance match to a part15 transmitter so that the efficiency in terms of how much gets from the rf finals to a radiating surface could be improved somewhat?

Impedance matching in the Part 15 world seems only to consider the reactive component of the load impedance. A coil with suitable characteristics will remove the effect of the reactive component of the 3-m antenna, but the pure resistance term that is left also needs to be matched to whatever load resistance the transmitter is designed to drive.

For example, a transmitter designed for a load impedance of 50 ohms will "see" a 2:1 SWR when connected to a pure, resistive load of 100 ohms (no reactance). Such mismatches reduce the power radiated by the antenna, and stress the components in the transmitter output stage.

Rich

Rich,

The coil losses you reported are much too low. In practice, it is very difficult to get a loading coil Q higher than about 350, or so. The highest Q I have ever measured on a coil of any kind is about 600. I have measured thousands of coils.

First of all, it is practically impossible to calculate coil Q. The major causes of coil loss are the skin effect and the proximity effect. Skin effect is relatively easy to deal with, but no accurate formulas presently exist for the proximity effect. Terman's handbook has formulas by Butterwoth, but these formulas are inaccurate. I have found from actual measurements that the Butterworth formulas greatly underestimate the proximity effect.

I am presently studying loading coils, and I will be posting in the future on Radio Joe's loading coil thread.