Non-Foster Design for Antennas

Stavros Koulouridis
Microwave Communications Group, Electrical and Computer Engineering Department University of Patras Rio-Patras, Greece koulouridis@ece.upatras.gr

AbstractWe propose a simple but robust technique for NonFoster circuitry design for antennas. Proposed technique can match antennas of different type and, thus, it is able to move beyond the usual Non-Foster dipole matching applications. It consists of introducing an additional port inside the antenna and loading of the port with a negative network. As shown, the novelty of the technique lies in reproducing a selected part of an ideal impedance (using only series negative elements) that would perfectly match the antenna at the full length of the examined frequency window. Proposed method is demonstrated by transforming a narrow resonant patch to a very broadband and much smaller antenna. More specifically, its original 2% impedance bandwidth (-10dB) is increased to an impressive 100% bandwidth. Moreover final antenna is 2 times smaller (1/4th of its original size). Keywords-component; Negative broadband; novel design technique circuitry; minimization;

and increase stability are some options. We can also design and implement lossless matching networks for the tuning of the transistors [12] or we may need to alter the internal design of used amplifiers [13]. Still, what we may actually need is design methods to formally employ negative networks in antenna design. Carried out research follows rather a simplistic approach where either optimization or antenna input impedance characteristics define negative networks implementation. In other words, a formal approach does not exist and negative circuits either cancel reactive part of the antenna input impedance or are being implemented after an optimizer reveals a negative sign in the optimization circuit process. However, in that way, negative networks are not fully exploited since they are being used as passive circuits alternatives. In this paper we propose a novel Non-Foster design technique that can tune a large number of different antenna types. A port is appropriately added inside the antenna. Then negative circuits are appropriately introduced (negative capacitance and inductance) that, as shown, can greatly alter input impedance. Applied negative circuits permit to increase the very low radiation resistance values and significantly decrease the large antenna reactance that demonstrate at the low frequency regions (i.e. we put our focus on small antennas) Proposed approach can make Non-Foster implementation straightforward for several antennas. In the same time antennas decrease in size drastically and obtain large bandwidths. In the next section proposed technique is discusses followed by results for a common patch antenna. As shown, bandwidth increases from 2% to 100% and size decreases 2 times (i.e. final size is 1/4th of original antenna). II. METHOD

I.

INTRODUCTION

There is renewed interest in Non-Foster circuitry for antennas lately [1]-[10]. Further, Non-Foster antenna matching has been recently laboratory tested and demonstrated in [9], [10]. Interest has been driven by the continuous need to minimize radiating devices making them lighter and more compact. In addition technology in microwave integrated circuits and transistors seems to have advanced to the point that inherent losses and stability issues could be addressed. Negative circuits are attractive because they can easily overcome size restrictions in antenna design. That is negative matching elements can provide a way to construct very small efficient antennas. As it has been actually shown very few negative elements, are able to achieve great bandwidths [1]. Certainly, there are major issues to solve in order to introduce negative elements in antenna design [3]. Negative networks realization asks for efficient circuit design and overcoming of active elements problems (compensate for mismatch losses, deal with deviation from ideal behavior or extend the narrow bandwidth). While numerous Non-Foster circuits have been proposed very few have been actually tested. Further, it is equally important to implement design techniques that will allow improvements in active elements characteristics [7]. Use of FET cascading to decrease the unwanted paracitics of the active elements [11] or transistor biasing [9] to decrease noise

Proposed approach [2] (see Fig. 1) is based on simple circuit theory principles. First step consists of appropriately inserting a new port inside an antenna. In that sense antenna is transformed into a two port network that can be analyzed and we can obtain its S-parameters or ABCD matrix. From circuit theory is well known that applying a load at the second port we can actually control the input impedance seen at the first port (the input port of the antenna). Indeed, from ABCD matrix we can actually calculate a load ZL that will alter antenna input impedance at will. For example we can use a load ZL that would match the input impedance at that of the feeding

100

Initial Re{Zin} Initial Im{Zin}

(a) (a)

50

Z (Antenna)

Ohm

Z in =

A ZL + B C ZL + D

(b) (b) D Z in B ZL = C Z in A (c)

(c)

-50

-100

200

400

600

800 1000 1200 Frequency (MHz)

1400

1600

1800

2000

Figure 1. Applying a load at an added port inside the antenna can control antenna impedance at will; (a) Analyze the antenna and obtain S or ABCD matrix for the two port system; (b) ABCD matrix allows for the calculation of (c) a load ZL for the needed Zin .

Figure 3. Initial (before matching) input impedance of patch (see Fig.2).

0 -5 -10

network in all the frequencies of interest. Moreover, since we do not change its shape, material or volume, antenna directivity is not affected and we can thus alter its efficiency at will. Naturally, matching the antenna is not that simple since the "needed" ZL would not be realizable as one would expect. However, as it can be shown, by focusing only to the reactive part of the load ZL an area with negative slope can be identified. The use of a simple negative network to replicate that negative slope can alter the antenna input impedance enough in order to increase input resistance to values close to characteristic impedance of feeding network for a broad frequency region. In that case, we can then move to antenna input and apply a secondary matching network in order to tune antenna reactance. It has been seen though that even the first step is enough to create satisfactory results. The discussed method is explained with an example in the following section. III. NUMERICAL EXAMPLE AND DISCUSSION

dB

-15 -20 -25 -30 0

200

400

600

800

1000

1200

1400

1600

1800

2000

Figure 4. Orignal (before matching) and after matching return loss. Matched return loss is lower than -10dB from 400MHz up to 1600MHz.

A common patch antenna (see Fig. 2) is employed. It has 3.25x3.25cm2 aperture and is printed on 10.2 substrate. Its thickness is 5mm. Antenna calculated input impedance is given in Fig 3 and Return Loss in Fig. 4 (red line). As can be seen antenna has a narrow resonance at f=1.59GHz with 30 MHz 10dB impedance bandwidth. This is commonly expected. Naturally optimizing feeding positions, bandwidth could be somewhat enhanced, nevertheless this would not lead to significant changes to the general antenna performance. Further, if we tried to apply matching (either passive or negative) at the input we would not achieve much since radiation resistance becomes very low when we move away from the resonance. We can now select a second port inside the antenna to apply our technique. Position of second port can be seen in Fig. 2. As
Substrate: r =10.2

seen, port is placed at a symmetric point to the feeding in relation with the antenna center. Then we obtain ABCD matrix (from simulations). According to discussed method we can calculate a load ZL that, when applied at the second port, could alter input impedance at will. Indeed if we desired an input impedance of 50+j0 Ohm in the entire frequency window this would lead to the ZL load given in Fig. 5 (denoted with red color). As expected the needed ZL is not realizable. However we can easily identify a negative slope of the ZL reactance. Thisarea for example extends up to 1600MHz. It can be easily found that this slope can be curve fitted by a combination of series negative inductor and a series negative capacitor. Indeed the series negative circuit shown in Fig. 2 has the response denoted with green color in Fig. 5.
100

50

-50

Feeding

-100

400

800

1200

1600

2000

Figure 2. A 3.25x3.25m patch antenna with r= 10.2 is employed to demosntrate proposed technique. Antenna is matched at an internal introduced port with a series negative circuit.

Figure 5. Calculated ZL load (denoted with red) to create input impedance of 50+j0 Ohm in the entire frequency region; applied reactive load ZL (denoted with green) to match the antenna. Green line corresponds to series negative circuits of Fig. 2.

Applying the negative circuit of Fig. 2 at the second port we obtain a significantly improved input impedance. Input resistance obtains values close to 50 Ohm for a broad frequency region (up to 1000MHz) while input reactance is zeroed at a region from 600MHz to 1200MHz! Fig. 4 makes more clear the achieved antenna minimization and bandwidth enhancement. As can be seen (blue line) return loss drops below 10dB from 400MHz to 1200MHz. That is a 800MHz bandwidth (i.e 100% bandwidth for 800MHz central frequency). Further, antenna size has decreased two times (since initial 1600MHz operation has been driven down to 400MHz). As it can be shown above behavior is not unique for the patch antenna. It can be observed in various antennas types (dipoles, loops. etc.) either narrowband or broadband. In that way proposed technique can provide for a robust approach to Non-Foster antenna circuits design. IV. CONCLUSION

100

50

-50

-100

200

400

600

800

1000

1200

1400

1600

1800

2000

Figure 6. Final matched inpute impedance for the patch antenna when the negtive circuit of Fig. 2 is applied at the second port. [10] J. T. Aberle, Two-port representation of an antenna with application to non-Foster matching networks, IEEE Trans. on Antennas and Propag., vol. 56, pp. 12181222, May 2008. [11] I. J. Bahl, Fundamendals of RF and Microwave Transistor Amplifiers, John Wiley and Sons, 2009. [12] S E. Sussman-Fort, An NIC-based negative resistance circuit for microwave active filters, Inter. Journal of Microwave Millimeter-Wave Computer Aided Design, pp. 130-139, vol. 4, 1994. [13] A. Kaya, Wide-Band Compact Microwave Transistor Amplifier methodology and the analysis of its input matching mechanism using negative impedance converter, Microw Opt. Tech. Let., vol. 50, pp. 192-197, Jan. 2008.

In present work, a novel technique has been presented to obtain Non-Foster circuits for antennas. Proposed technique is robust and rather straightforward. It consists of introducing a second port inside the antenna and apropriatelly applying negative circuit matching. As it was shown, the 2% impedance bandwidth of a common patch antennas can be increased to 100% while its size can be by 200%. Finally, proposed approach is robust and it can be shown that it can be applied to various antenna types. REFERENCES
[1] S. Koulouridis and J. L. Volakis, Non-foster circuits for small broadband antennas, in Proceedings IEEE Antennas Propagation Society Inter. Sym., Charleston, SC, 2009 S. Koulouridis, Impedance matching for small antennas using passive and active circuits, in Small Antennas: Miniaturization Techniques & Applications, John Volakis, Chi-Chih Chen, Kyohei Fujimoto, Eds. New York: McGraw Hill, 2010, pp. 361-388, S. Koulouridis, Negative networks design for antennas, URSI North American Radio Science Meeting Digest, Toronto, Canada, July 11-16, 2010. K. Karlsson, J. Carlsson, "Non-Foster networks for improvement of radiation efficiency and effective diversity gain of a multi-port antenna,", in Proceedings European Conference on Antennas and Propagation (EuCAP), pp.1-4, 12-16 April 2010. Keum-Su Song, R.G. Rojas, "Non-Foster impedance matching of electrically small antennas," in Proceedings Antennas and Propagation Society International Symposium (APSURSI), pp.1-4, 11-17 July 2010. Keum-Su Song, R.G. Rojas, "Electrically small wire monopole antenna with Non-Foster impedance element," in Proceedings European Conference on Antennas and Propagation (EuCAP), pp.1-4, 12-16 April 2010. S.E. Sussman-Fort, "Non-Foster vs. active matching of an electricallysmall receive antenna in Proceedings Antennas and Propagation Society International Symposium (APSURSI), pp.1-4, 11-17 July 2010. Peng Jin, R.W. Ziolkowski, "Broadband, Efficient, Electrically Small Metamaterial-Inspired Antennas Facilitated by Active Near-Field Resonant Parasitic Elements," IEEE Transactions on Antennas and Propagation, vol.58, no.2, pp.318-327, Feb. 2010. S. E. Sussman-Fort, R. M. Rudish, Non-Foster Impedance Matching of Electrically-Small Antennas, IEEE Trans. Antennas and Propagation, pp.2230-2241, vol 57, Aug 2009.

[2]

[3]

[4]

[5]

[6]

[7]

[8]

[9]