Embedded Files
Created: Feb 2025.
Introduction
Introduction
The PA0FRI preamp [1] uses fewer components and occupies half the PCB area compared to the LZ1AQ (fig. 1). The smaller PCB can be advantageously shoved inside the PVC pipe that is used as the loop support frame.
Fig. 1: The PA0FRI PCB (bottom) is half the size of LZ1AQ (top)
PA0FRI prolifically updated the circuit several times per year! The version that we are evaluating is "12-nov-2021". It differs from an earlier version, viz. 31-oct-2021, by changing the output transformer from 4:1 Faraday to 1:1 Guanella and reducing the input coupling caps from 100 nF to 1nF.
Like the other active loop designs, PA0FRI's performances are poorly documented. As knowledge of the preamp's performances is necessary for (a) checking homebrewed prototypes, (b) improving on the original design and (c) comparing against the competition, e.g. LZ1AQ [2], MA0AYF [3], and MLA-30 [4], etc., this article aims to bridge this knowledge gap. To this end, we performed simulation and measurements of the PA0FRI and then, collates the results here. Finally, we tabulate the most important results of the various preamps in order to facilitate comparison.
Material and method
Material and method
The evaluation samples deviate from the original design in the following aspects:
1. 2N5109 replaced with 2N2222A as the former is not available in our junk box.
2. changed the base bias resistors R1-2 from the original 27 kΩ to 6.8 kΩ in order to maintain 44 mA current after changing the aforementioned transistors.
3. removed the 4 input protection diodes
4. replaced the unspecified toroid (2P80 ??) with TDK 7-14-3.5, u=800. The number of turns are then adjusted to achieve the specified 22 uH.
Fig. 2: Circuit
The prototypes were assembled on a double-sided FR4 PCB measuring 50 x 25 x 1.5 mm. For ease of replication by beginners, all components are through hole.
Fig. 3: part-list and PCB drawing
Fig. 4: Photo of completed assembly and blank PCB
During gain measurement, the preamp input is connected to a dummy aerial (aerial simulation network) that replicates the impedances of a 1m diameter loop [5]. Other measurements omit the dummy aerial and are directly fed from 50Ω. The preamp is phantom powered via a biastee (fig. 5).
Fig. 5: Gain test setup
To predict the performances and to facilitate circuit tweaking in the future, we developed an equivalent circuit model for simulating RF performances (fig. 6). The 2N2222A Spice model originates from Motorola, and is dated 1993. The balun CMP1 is ideal and does not have provision for loss. Hence, R7, which is not a physical component, is added to model balun loss. The RLC component models do not include parasitics - hence, they will become less accurate at higher frequencies. Additionally, the PCB traces are not modelled as we think they are negligible at HF.
To minimize screen clutter, the simulation is arranged in two-levels At the top-level, the preamp's model is represented by the sub-circuit X1 (fig. 7). The preamp input is fed by a Dummy Aerial X2 and the output is connected to a short transmission line TL1 which accounts for the finite phase length of the combination of coax cable and biastee. The dummy aerial is modelled as an unbalanced L-network, but the actual dummy is a balanced L-network. The reason for modelling it as unbalanced is the convenience of bypassing the sub-circuit for evaluating the preamp's S11.
Fig. 6: Preamp's equivalent circuit model
Fig. 7: Equivalent circuit model of dummy aerial (top) and the top-level representation of the simulation (bottom)
Results
Results
Measured and simulated gain show good agreement. Modelled gain (dotted black in fig. 8) is slightly higher than experimental because the former doesn't account for various circuit losses. Over the usable frequency range (3-30 MHz), the model's maximum error is < 4 dB. Of note is the relatively flat response over 7-30 MHz.
Fig. 8: Simulated (dotted red) and measured gain (solid black) show modest agreement
The S11 measurement exhibit good consistency between the two samples; i.e. the black and red traces almost overlapped (fig. 9). However, the modelled result is inaccurate. The grossness of the error makes us suspect that as the 2N2222A is not intended for RF applications, its model was not been optimized for s-parameters.
Fig. 9: The measured S11 traces (solid) almost overlapped, pointing to the samples' good consistency. However, the model (dashed) is inaccurate.
The S22 is skewed to the Smith chart's right - hence, pointing to a higher than 50Ω impedance. We speculate that this impedance mismatch can be solved by choosing the 31-oct-2021 version which has a 4:1 output transformer.
Despite our suspicion of the 2N2222A model, the simulated S22 (dashed) is reasonably accurate (fig. 10). The overlapping measured traces point to good inter-sample consistency.
Fig. 10: Reasonable agreement between simulated (dashed black) and measured S22
The measured 1-dB gain compression (P1dB) shows a slight drop with frequency (fig. 11) - this is consistent with increasing component losses with frequency. Taking the worse sample, sn2, P1dB is better than 5.8 dBm. In contrast to measurement, the modelled P1dB has the opposite trend of rising with frequency. The maximum modelled error is 3.2 dB.
Fig. 11: Simulated (dotted black) and measured P1dB (solid blue & red). The model has a maximum error of 3.2 dB.
The measured output 3rd order intercept point (OIP3) also exhibits a gradual decrease with frequency (fig. 12). The worst case OIP3 is >= 18 dBm. The modelled OIP3 also shows the opposite trend of rising with frequency. We suspect the erroneous positive slope is caused by the balun model failing to account for stray inductance. The maximum modelled error is 2.5 dB.
Fig. 12: Simulated (dashed green) and measured OIP3. The maximum model error is 2.5 dB
The experimental common mode rejection ratio (CMRR) is > 15 dB over 1-30 MHz (fig 13). As PA0FRI uses an inductor instead of a resistor in the differential amplifier's "tail", this causes the CMRR to drop at low frequencies. Nevertheless, above 9 MHz, PA0FRI's CMRR is highest among the competition (fig. 14). As CMRR is highly sensitive to any inbalance between the physical circuit's two halves, no matter how minute, we did not simulate this parameter because the perfectly balanced model will predict an unrealistic number.
Fig. 13: Measured CMRR exceeds 15 dB over HF, and rises to > 50 dB above 9 MHz
The PA0FRI offers a mixed-bag of performances compared to its peers. It has the highest midband CMRR, but middling P1dB and OIP3 (fig. 14). PA0FRI's P1dB and OIP3 are significantly lower than the leader LZ1AQ because of the former's low operating current. Moreover, PA0FRI's collectors are supplied through 390R resistors which approximately halved the Vce.
Fig. 14: A survey of active loops shows the PA0FRI having the highest CMRR, but moderate P1dB and OIP3
Appendix
Appendix
As this PA0FRI version is distinguished by very small capacitance of 1nF in positions C1-2, this begs the question whether low frequency performance is compromised. Simulating with 4 nF and 7 nF show a modest 2 dB gain increase at frequencies below 7 MHz (fig. 15). There is virtually no further improvement above 4 nF.
Fig. 15: Simulation shows a slight gain increase when the input capacitors C1-2 are increased from 1 nF (red trace) to 4 nF (blue), but no further improvement thereafter.
The differences are huge when comparing models from different manufacturers! Siemens PZT2222A severely underestimates gain (dotted blue in fig. 16). The 8 MHz gain peaks in Motorola 2N2222A (dotted red) and Natsemi PN2222A (dotted green) can be dampened by accounting for the input balun loss. Nevertheless, there is still 1 MHz difference between the latter two.
Fig. 16: Transistor models from different manufacturers exhibit huge differences in gain simulation!
References
References
[1] 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
[2] C. Levkov, “Wideband active small magnetic loop antenna”, 2011. [Online] Available: http://www.lz1aq.signacor.com/docs/wsml/wideband-active-sm-loop-antenna.htm
[3] 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/
[4] MegaLoop MLA-30+. Can be bought from various online shopping platforms
[5] "Aerial simulation network / dummy aerial for active loops", [Online] Available: https://sites.google.com/site/randomwok/Home/electronic-projects/aerials/aerial-simulation-network-dummy-aerial-for-active-loops
Page updated
Google Sites
Report abuse