Aerial simulation network / dummy aerial for active loops

Created: Sep. 2024. Modified: Nov. 2024.

-------------------------------

Introduction

Active loop aerials enable reception in the HF range (3~30 MHz) when space is limited. However, active loop aerials are difficult to test in the lab because they lacked input connectors for connection to signal sources. Instead, the test signals are customarily radiated to the DUT loop through a reference loop. This method of feeding loop-equipped receivers is described in the IEC315-1 standard [1], but it has limited utility because the specified reference loop is limited to <= 2.5 MHz . Moreover, the IEC method is inconvenient to non-specialists because it requires tedious field strength calibration and a shield room to prevent the DUT loop from picking up on-air interferences. Alternatively, a second method connects a signal source directly to a small section of the loop aerial, but there is no detail of the implementation [2], and presumably, it will also require a shield room to prevent reception of on-air signals (fig. 1). Either test method also requires the loops to be kept away from conducting objects such as test equipment; i.e. these test setups require large areas. Furthermore, the loop's manufacturing tolerance can contribute to measurement uncertainty.

Fig. 1: A prior art feeds the active loop by direct connection to a small section of the loop. The balun converts the network analyser's single ended output to balanced output to avoid unbalancing the loop. Unfortunately, no dimensional detail is given. Diagram redrawn from Geerligs [2].

When testing the active loops' electronics, specifically, the pre-amplifiers (pre-amp), eliminating the loop as the input device can save space and obviates the shield room requirement. Moreover, substituting loops of different origins with a standardized aerial simulation network / dummy aerial can reduce measurement variability. To replace the loop as the pre-amp's input device, we develop a dummy aerial capable of emulating the loop behaviour over the HF range. This article describes the dummy aerial's topology selection, construction details, modeled performances, and finally, its application in comparing the performances of 4 different loop pre-amps.

Material & method

The task of developing the dummy aerial's circuit is fortunately simplified by the fact that HF active loops almost invariably have a 1m diameter. This size is dictated by the need to keep the self-resonance at ~30 MHz, hence, the loop cannot be made larger or else the upper frequency limit will be curtailed. Conversely, the loop cannot be made smaller without seriously degrading its sensitivity. The implication of the constant size is that the equivalent circuit model is also relatively invariant.

Our task is further facilitated by two prior arts: a Norton equivalent circuit model for the archetypal 1m loop [3] and measurement of the loop impedance Z [4]. The loop measurement reveals that Z is very low at 1 MHz (marker 1 in fig. 2), but rises with frequency until reaching a maximum at 30 MHz (marker 5) due to resonance. Over most of the evaluated frequency range, the trace stays at the edge of the Smith chart, only to veer inward at 30 MHz - the latter behaviour is likely due to increased losses at resonance.

Fig. 2: Previously reported 1m loop's impedance showing its value changing from near zero at 1 MHz (marker 1) to almost infinity at 30 MHz (marker 5). Image used with permission [4]

At 1 MHz, the measured inductance (3.0768 uH) is reasonably close to the modeled value (fig. 3). To account for the 30 MHz resonance, a capacitance of ~9.1 pF will be required by calculation.

Fig. 3: The 1m loop's modelled and measured equivalent circuit values showing good agreement

Using the reported values of the equivalent circuit model, we investigated several candidate topologies (fig 4). Using circuit simulation, the candidates were assessed for their resemblance to the measured data. In all cases, the left terminals are terminated with 50Ω. Of the 4 candidates evaluated, only the top right one can replicate the loop's impedance - henceforth, all subsequent discussion is referring to this topology.

Fig. 4: Of the 4 candidate topologies evaluated through simulation, only the top right one can reasonably replicate the loop's measured data

The chosen L-network topology has an unbalanced output, but the pre-amp input is balanced. Hence, a 1:1 balun is required for the interface. The balun can be placed either before or after the L-network (fig. 5). However, each choice has its attendant pro’s and con’s. Positioning the balun at the output simplifies the L-network because an unbalanced version can be used, but the balun ‘sees’ a varying differential mode impedance which can cause the test source to be momentarily un-levelled during frequency sweep. Conversely, positioning the balun at the input has the advantage of the differential mode impedance being fixed at 50 Ω, but has the disadvantage of requiring a slightly more complicated balanced / symmetrical L network. In the end, we selected the second option because the first one was plagued by unlevelled sweeps.

Fig. 5: Two possible positions for the 1:1 balun, each with its attendant pro’s and con’s. In the bottom arrangement, the L-section has to be converted to a balanced version

The practical implementation of the dummy aerial is as shown in the circuit of fig. 6. A BNC female connector J1 serves as its input. Resistor R1 provides a good match to the signal source. The loop impedance simulation L-network is represented by the combination C1-2 and L1. Instead of the modelled value of 9.4 pF, the nearest standard value of 10 pF was picked for for the capacitor combination; i.e. C1 and C2 in series. L2, a common mode choke / Guanella balun, enables the unbalanced input to interface with the balanced L-network. Both L1 and L2 are wound on ferrite rings to achieve broadband operation.

Fig. 6: Part list & circuit diagram of proposed dummy aerial

The prototype was assembled using point-to-point wiring and enclosed in a compact 50x35x20 mm plastic box (fig. 7). Using crocodile clips for the output connection J2 enable connection to different types of terminals at the pre-amp (fig. 8).

Fig. 7: Photo of dummy aerial's internal. The dummy aerial's 50x35x20 mm size is significantly smaller than the 1m diameter loop that it replaces

Fig. 8: Photo of dummy aerial connected to an MLA-30+'s wing-nuts via crocodile clips

Subsequently, we used the prototype dummy aerial to compare the frequency response of 4 popular active loops: LZ1AQ [5], M0AYF [6], MLA-30+ [7], and PA0FRI [8]. With the exception of the commercially manufactured MLA-30+, the other pre-amps are home constructed. However, our version of the M0AYF uses 2N2222A BJT instead of the original BC547. For the PA0FRI, we also use 2N2222A instead of the original 2N5109. The reason for the transistor substitution is using devices that were available at hand.

Results

The dummy aerial prototype works as predicted. Both simulated and measured impedances agree well over 1-30 MHz (fig. 9). The model exhibits a slight divergence from reality at 30 MHz, because it cannot account for the real circuit's increased loss at resonance.

Fig. 9: The dummy aerial's simulated and measured impedances show good agreement over 1-30 MHz

The experimental dummy aerial is able to replicate the loop aerial's impedance. The numbered markers 1-5, which represent the same frequencies, are at nearly identical positions on both charts (fig. 10). The markers 5 on both charts curl inward due to higher losses at resonance. .

Fig. 10: Measured impedances of loop aerial (left) vs. dummy aerial (right). The two charts agree reasonably well over 1~30 MHz. The numbered markers 1-5 represent the same frequencies on either chart

The prototype dummy aerial is capable of teasing out the frequency response differences among the 4 evaluated pre-amp designs. The measured gains drop dramatically at the bottom end of the frequency range (1~6 MHz) because the dummy aerial is able to emulate the real loop's decreasing efficiency with falling frequency (fig. 11). LZ1AQ exhibits a peak at ~ 29 MHz because it incorporates an input low-pass-filter is intended to resonate with the real loop at the HF range's top end. The other 3 pre-amps, which do not have filters, exhibit peaks in the 9~10 MHz. The gain peaks occur due to dummy antenna resonating with the pre-amps' input capacitances of ~80 pF. MLA-30+ has the highest gain of all, hands-down, because it utilizes a highly integrated op-amp / video amp, TL592B [9] as opposed to the others' single / dual transistor stages. To our knowledge, this is the first time the performances of these 4 designs are measured under similar conditions and then, compared in one graph.

Nevertheless, it must be admitted that the measurement may not accurately replicate a particular active loop's frequency response, especially when the simple dummy circuit has been designed for impedance and not for frequency response. However, the dummy is still useful for comparing different pre-amps' relative gain at a particular frequency because it can mimic the actual loop impedance, e.g. if pre-amp X's gain is higher than pre-amp Y at 5 MHz when measured with the dummy, then there is no reason why the real-world gain performance will be dissimilar. In the future when a particular active loop’s frequency response is made available, then it is possible to correct any deviation from reality by adding a amplitude correction network - similar to the amplitude equalizers inside CATV amps. Alternatively, it can be simply implemented using a correction table for the signal source, inside the test software.

Fig 11: Measured gains of 4 active loops, LZ1AQ, M0AYF, MLA-30+ and PA0FRI, prove that the proposed dummy aerial is capable of teasing out the gain differences between them

Conclusion

The inconveniences associated with current test methods can be avoided by substituting the loop with a dummy aerial. The proposed dummy aerial can approximate the loop impedance over most of the HF range. The dummy aerial has shown great promise in elucidating the gain response differences of 4 popular pre-amps. In the future, we plan to use the dummy aerial to evaluate more active loops in addition the 4 designs that we have already evaluated.

References

[1] Methods of measurement on radio receivers for various classes of emission, IEC315-1:1988, 2nd ed., section 22.

[2] F.H.V. Geerligs, PA0FRI, "Active loop antenna for reception", 25 Jul. 2023. [Online] Available: https://pa0fri.home.xs4all.nl/Ant/Actieve%20ontvangst%20antenne/Active%20loop%20antenna%20for%20reception.htm

[3] C. Levkov, LZ1AQ, Wideband active small magnetic loop antenna, v1.1, Jun. 2011. [Online] Available: http://www.lz1aq.signacor.com/docs/wsml/wideband-active-sm-loop-antenna.htm

[4] M. Ehrenfried, G8JNJ, "Broadband loops", section: Input Low Pass filter networks. [Online] Available: https://web.archive.org/web/20230327034702/https://www.g8jnj.net/loop-inductance

[5] C. Levkov, “Wideband active small magnetic loop antenna”, 2011. [Online] Available: http://www.lz1aq.signacor.com/docs/wsml/wideband-active-sm-loop-antenna.htm

[6] Des, M0AYF, 'Active loop antenna for HF', transcribed by M0LMK. [Online] Available: http://www.m0lmk.co.uk/2015/02/14/active-loop-antenna-for-hf/

[7] MegaLoop MLA-30+. Can be bought from various online shopping platforms.

[8] F. H. V. Geerligs, " PA0FRI's active loop antenna", ver. 12 Nov 2021, [Online] Available: https://www.pa0fri.com/

[9] M. Ehrenfried, G8JNJ, "Active antennas overview", [Online] Available: web.archive.org/web/20210306185317/https://g8jnj.net/activeantennas.htm

Acknowledgement

The author thanks M. Ehrenfried, G8JNJ, for the loop measurement and discussion.

Page updated

Google Sites

Report abuse