Loop Antenna Sensitivity Dallas Lankford, April 1991 How can one estimate the sensitivity of a loop antenna? And how sensitive should a loop antenna be for s tate of the art performance in the MW band. The purpose of this note is to discuss these two questions. Ideally, the best way to estimate the sensitivity sensitivity of a loop antenna is to measure it. ut that requires a shielded room and calibrated test equipment which are beyond the means of most hobbyists. !ortunately, in his article "!erroma#netic $oop Aerials !or %ilometric Waves,& Wireless Engineer , !eb. '()), pa#es *'+*, -. . elrose derived some formulas which can be used to ma/e quite accurate estimates of loop antenna sensitivity. He showed that the si#nal to noise ratio of a tuned loop is #iven by
where 0 is the number of turns of the loop coil, A is the area in square meters enclosed by one turn of the loop coil, 1rod is the rod permeability permeability 2see elrose3s article for a #raph for convertin# the initial permeability permeability 1 to rod permeability 1rod4, 1rod4, 5f is the bandwidth in Hert6 Hert6 seen at the detector detector of the receiver, receiver, $ is the inductance in Henrys Henrys of the loop coil, and 7 is the field stren#th in volts per meter of the received si#nal. !or an air core loop, 1rod 8 '. In elrose3s article it was pointed out that the noise which limits the sensitivity of a loop antenna is thermal noise due to the resistive component of the loop antenna impedance. We will call this noise loop coil noise. The noise floor of a loop antenna is defined as the volta#e equal to a field stren#th which produces a si#nal to noise ratio of '. The noise floor of a loop antenna is the volta#e equal to the loop coil noise. olvin# the previous equation with i#nal90oise 8 ' we #et the followin#.
The variables in the above equation are dependent on each other to some e:tent, but for the sa/e of discussion let us assume that $ 8 ')* 1H 2so that a ; p! capacitor tunes the loop to );; /H64, that 5f 8 < /H6 2which is about the minimum usable bandwidth4, that = 8 ';; 2a not unreasonable value for a loaded loop coil4, and that f 8 ' MH6. !or these assumptions we #et the followin#.
With the assumptions above, a < foot square air core loop has '* turns, a ' foot square air core loop has << turns, a inch square air core loop has > turns, and a pace Ma#net rod 2'< lon# by (9' diameter, 1 8 *;;4 has >; turns. The air core loops have 1rod 8 ', while the pace Ma#net rod has 1rod 8 ');. We can also convert to other bandwidths by multiplyin# multiplyin# by the square root of one half the bandwidth bandwidth in /H6. Thus we #et the followin#. followin#.
'
o what do these numbers mean? @an you hear anythin# on a ' foot air core loop that you can3t hear on a pace Ma#net? @an you hear anythin# on a < foot air core loop that you can3t hear on a ' foot air core loop? It all depends on the /ind of si#nal you want to hear and the minimum man+made and power line noise at your location. As I have said in previous articles, if you live in a lar#e urban area, I doubt that you will ever hear anythin# on a ' or < foot air core loop antenna which you cannot hear equally well on a pace Ma#net or similar ferrite rod loop antenna. Incidentally, I have wound coils on all /inds of ferrite rods, bundled, not overlapped, overlapped, cose wound, spaced over the entire len#th of the rod, you name it ... I3ve tried them all. As lon# as you use at least * rods of 1 8 '<) bundled and overlapped, or at least B rods bundled and not overlapped, the noise floors are all virtually identical. Also, based on listenin# tests, the noise floor of a inch loop is virtually identical with the noise floor of a pace Ma#net, which a#rees with the numbers in the table above. Calph currently uses '; ferrite rods made by #luin# to#ether '; ferrite cylinders of 1 8 '<) which are ' lon# by '9< diameter. The noise floor of his ferrite rod loops is identical to the pace Ma#net based on my listenin# tests 2this is the Dreat $ittle $oop sold by Cadio West4. o my followin# remar/s about the pace Ma#net apply to any well desi#ned ferrite rod loop and to the inch air core loop. As I have remar/ed in previous articles, I am fortunate to live in a small town where ambient man+made and power line noise occasionally fall to very low levels. En those occasions 2and only on those occasions4 I can hear a very definite difference between the pace Ma#net and a ' foot air core loop. En a few wea/ daytime si#nals the ' foot air core loop 2and of course the < foot air core loop4 will produce clear audio when the pace Ma#net produces no audio at all. The amps of my loops have been equali6ed so that the output si#nal levels of all of my loops are virtually identical. @onsequently, I have concluded that this difference between the ' foot air core loop and the pace Ma#net is due to the lower noise floor of the ' foot loop. About once a year the daytime noise levels at my location 2Custon, $A4 drop to super low levels. En these occasions I can often hear WEAI an Antonio on '<;; /H6 fadin# in and out of the noise while the C+>(;A meter sits s olidly on ;. Festerday was such a day, and provided me with a rare opportunity to try to hear a difference between the ' foot air core loop and the < foot air core loop. 2My ferrite rod loops and inch air core loop were producin# no audio at all from WEAI, Gust thermal noise from the 'oop coils.4 Maybe I wanted to hear a difference, but it did seem li/e I could follow WEAI deeper into the ambient noise with the < foot loop than with the ' foot loop. And it did seem that WEAI was clearer on the < foot loop than on the ' foot loop when WEAI was in the clear above the bac/#round noise. However, there was no dramatic difference between the ' foot and < foot loop li/e there is between the ' foot 2or < foot4 loop and the pace Ma#net. When si#nal levels are much hi#her, such as when in# domestic channels or forei#n splits at ni#ht, there is no difference between what you can hear with the various loops. However, there are some situations where a ' foot air core loop mi#ht produce clear audio when a ferrite rod loop does not. Jnfortunately,.I am not in a position to compare loops for most of those situations. A few times a year I can hear a/ar ene#al on B) at local sunset when noise levels are low and adGacent channel si#nal levels are low. I have always used a < foot air core loop for those receptions, but I e:pect a ' foot air core loop would be equally #ood, while a ferrite rod loop would not. 7@0A ers often report TA reception Gust before and at local sunset. I e:pect that a ' or < foot air core loop would be better than a ferrite rod loop in that situation if ambient noise is low. W@0A ers often report TK reception at and Gust after sunrise. A#ain, I e:pect that a ' or < foot air core loop would be better if ambient noise is low. !or these reasons, I have developed a ' foot air core loop and companion two !7T amp which is not much lar#er than a pace Ma#net or Dreat $ittle $oop, but which has a demonstrably better noise floor. This loop will be described in a future article. Added 9/2/06, rev. 9/12/06: The followin# noise floor formula for one turn air core broadband loops is #otten by settin# 0 8 ', 1rod 8 ', and = 8 '.
!or f 8 );; /H6 we #et the followin#. The '; d 90 ratios are #otten by multiplyin# the noise floors by >.'. The si6es are similar to well /nown broadband loops. The noise floors and '; d 90 sensitivities do not ta/e into <
account amplifier noise, which may contribute si#nificant noise in the cases of first twloops and si#nificantly increase the noise floors and '; d 90 for those two loops. A 2area4
$
0oise !loors < /H6 W
/H6 W
'; d 90 < /H6 W
/H6 W
' m2
1H
<.* 1L
*.' 1L
B. 1L
'>.< 1L
* m2
'< 1H
'.< 1L
<.' 1L
>. 1L
. 1L
<) m2
<* 1H
;.< 1L
;.*) 1L
;.< 1L
'.* 1L
<<) m2
';; 1H
;.; 1L
;.'; 1L
;.'( 1L
;.; 1L
The ' m2 loop probably does not have adequate sensitivity for the lower MW band in locations with low levels of man made noise, especially when used in a phased array. In any case a hi#h #ain ampwould be needed, with its potential noise and intermod problems. The * m2 loop mi#ht be adequate for the lower MW band, but it would depend on the amplifier used and on ambient man made noise. I have built and tested the <) m 2 loop usin# a ';. d #ain push+pull 0orton amp with hi#h intercepts and low noise fi#ure. Its sensitivity seems acceptable, thou#h perhaps not entirely adequate for some phasin# situations without additional amplification. The nominal impedance of the <) m 2 loop is about B) ohms at );; /H6 so I did not impedance match the loop to the amp. I used the amp input transformer as a 'N' balanced to balanced transformer so that the loop was balanced and to use twin lead. Two <) m 2 amplified loops spaced ); meters apart were tested as a phased arrayN #roundwave 2daytime4 nulls were fine, but s/ywave 2ni#httime4 nulls were not as deep or as stable as a pair of ') meter noise reducin# verticals spaced ); meters apart. The <<) m 2 loop has been described in The allas !iles at www./on#sfGord.no . It does not require an amplifier. I have used it mainly at L$! with a nearby *) foot noise reducin# vertical for null steerin# with my modified Mise/ phaser. It also wor/s well for null steerin# in the MW band, but I prefer phased verticals for MW because they have deeper and more stable s/ywave 2ni#httime4 nulls. espite what some have claimed, I have never observed that loop antennas have inherently better si#nal to man made noise ratios than other antennas provided the comparison is fair. Man made noise often has a stron#er component which is directional, and a wea/er component which is more or less omnidirectional. In such cases I have observed many times that the amount of improvement in si#nal to man made noise ratio which can be obtained by rotatin# a loop antenna to minimi6e the stron#er component is at least as much as the amount of improvement in si#nal to man made noise ratio which can be obtained by phasin# two spaced noise reducin# verticals or noise reducin# inverted $3s. I have also observed many times that if the stron#er directional man made noise component is not reduced by partial or complete nullin#, then the si#nal to man made noise ratios of noise reducin# verticals, noise reducin# inverted $3s, and loops are virtually identical. @laims that a loop antennas are less sensitive to man made noise are often due to 2'4 the loop havin# accidentally or deliberately been oriented so that the stron#er man made noise component has been partly or completely nulled, and9or 2<4 a small active loop antenna havin# accidentally or deliberately placed in a locali6ed re#ion of lower man made noise. In the latter case, I have observed that a small active whip antenna will provide Gust as much man made noise reduction as the loop.
>