Zip Cord Transmission Lines and Baluns An Analysis of a Low Cost Speaker Wire with Common Mode Chokes as a High Frequency Transmission Line by Dr. Carol F. Milazzo, KP4MD (posted 25 September 2010) E-mail: [email protected] SUMMARY This article describes an application of a low cost speaker wire zip cord with common mode current chokes as a high frequency transmission line. INTRODUCTION In Computer Assisted Low Profile Antenna Modeling II1, I described the design of a 40 meter full wave horizontal loop antenna using speaker wire as a balanced transmission line.  Despite its lower power capacity and higher attenuation per unit length than some coaxial cables, a short run of dual conductor speaker wire used as a parallel transmission line is lighter, less visible, more easily available and economical.  Speaker wire is also very easily wound on a ferrite toroid core to form a common mode choke or 1:1 current balun.  Hall2, Parmley3, and Wiesen4 have also discussed and characterized the use of zip cords as transmission lines.

CONTENTS Introduction Decreasing feed line loss Improving balun performance References

DECREASING FEED LINE LOSS - 04 December 2010 Originally, a 40 foot roll of 24AWG speaker wire (Figure 1) was used as the transmission line.  Its theoretical calculated DC resistance was compared to its measured resistance.  The loop antenna was made of CTI-20 gauge stranded wire with a resistance specification of 8.63 ohms per 1000 feet, and typical 24 gauge stranded copper wire (7/32) is listed as 23.3 ohms per 1000 feet. The calculated total DC resistance at the transmitter end of the 40 foot 24 gauge zip cord feed line attached to the loop antenna would be (140'/1000' x 8.63) + (80'/1000' x 23.3) or 3.1 ohms. The actual measured DC resistance was 7.7 ohms!  I confirmed this after taking down the 40 foot zip cord feed line, twisting the wires at the far end together and measuring 6.2 ohms for the entire 80 foot length of wire. This raised my concern for I2R losses. A 100 watt transmission to approximately 100 ohm impedance would produce radio frequency currents on the order of one ampere, and approximately 6 watts of the power would be wasted as heat in the feed line. For this reason, I replaced the 24 gauge speaker wire with 36 feet (11 m) of Pfanstiehl 18-gauge AS-18/50Z speaker wire (Figure 2).  An MFJ-202B RX noise bridge was used to measure this zip cord's characteristic impedance, velocity factor, and matched line attenuation listed in Table 1 and in the Transmission Lines Details program plots in Figures 3 through 5.5  The attenuation of the 36 foot feed line was then 1.5 dB at 7 MHz (comparable to RG-174/U at 1/3 of the cost) with a total measured antenna system DC resistance of 2.5 ohms with only 1 ohm due to feed line resistance. The speaker wire performed most satisfactorily with low standing wave ratios on frequencies 14 MHz and below. It exhibited increasing loss above 14 MHz and significant dielectric loss at high standing wave ratios.
	Figure 1.  Original Feed Line Uninex 24AWG Speaker Wire	Figure 2.  Replacement - Pfanstiehl 18AWG AS-18/50Z Speaker Wire

Figure 3.  Characteristic impedance Z0 vs. frequency	Figure 4.  Velocity factor vs. frequency	Figure 5.  Attenuation vs. frequency

IMPROVING BALUN PERFORMANCE A commercially available 4:1 voltage balun (Figure 6) was originally used for the transition between the 50 ohm coaxial cable and the balanced speaker wire transmission line. Lewallen6 and others have criticized the 4:1 Ruthroff (voltage) baluns7 commonly included in commercially produced antenna tuning units as being excessively lossy at high standing wave ratios and susceptible to common mode radio frequency noise. (Voltage baluns present these problems increasingly as an antenna is operated away from its resonant frequency and is in an asymmetrical environment that renders it an unbalanced load). After a 10 minute contact on 18 MHz CW running 100 watts output power into the 40 meter full wave loop, my 4:1 voltage balun felt quite hot to the touch.  In "Baluns and Tuners"8 Ehrenfried compared and documented the thermal power losses in various baluns (Figure 7).  Also, while using the voltage balun, shorting both sides of the balanced feed line together did not quiet the receiver as expected. A significant level of radio signals and noise were heard, representing common mode signal pickup on the feed line and conversely implying radiation from the feed line.
	Figure 6.  Original Balun Lossy 4:1 Ruthroff (voltage) balun	Figure 7.  Thermal loss in voltage balun (from "Baluns and Tuners," Ehrenfried)

In their articles Common-Mode Chokes⁹ and A Ham's Guide to RFI, Ferrites, Baluns, and Audio Interfacing¹⁰ Counselman and Brown discuss the use of common mode chokes to:

1. reduce undesired common mode radiation and noise pickup on transmission lines;
2. serve as a transition between balanced and unbalanced sources and loads if needed; and,
3. be configured to provide impedance transformation, if needed.

Figure 37 in reference 1 shows the feed point impedance variation between 80 and 350 ohms at the resonance frequencies of 7, 14, 21 and 28 MHz, presenting corresponding VSWRs of 1.3:1 to 3:1 to the 114 ohm zip cord.  4:1 impedance transformation at that point would increase the VSWR to 5.7:1 at 7 MHz; therefore no impedance transformation in needed at the feed point.  Neither is a balun needed as both the antenna and zip cord are balanced.  Therefore, a simple common mode choke is used there to reduce common mode radiation and noise.  Likewise, figure 54 in reference 1 shows the impedance at resonance frequencies measured at the transmitter end of the 51 foot transmission line varying between 45 and 120 ohms.  Therefore, a simple common mode choke would serve to reduce common mode noise and to transition the unbalanced antenna tuning unit (ATU) to the balanced zip cord, leaving the ATU to perform the impedance transformation to the 50 ohm radio.

Since common mode noise and feed line radiation were such a "pain in the derrière", I used an appropriate disposable 4" x 6" x 2" hinged plastic container to make a box (Fig. 8 and 12) for testing different baluns, with a dual binding post mounted on one end and an SO-239 female UHF receptacle on the other.  (Given the source of the box, I named it the "H Special"—"H" for the Henry unit of inductance, of course!)  The baluns under test would be attached to the connectors with alligator clips.  At first, I tested the 1:1 ferrite rod current balun made previously.  As superior common mode noise rejection was observed with the toroid core baluns, no further measurements with the ferrite rod balun were made. Then, at a local electronics dealer, I purchased some unmarked surplus ferrite toroid cores ($.85 each) that were identified as 43 ferrite mix with the use of an MFJ-202B noise bridge.  To make a 1:1 toroid current balun using coaxial cable, I stacked two FT114-43 toroid cores and wound 10 turns of RG-174/U 50 ohm cable through them (Figs. 10 and 11).  The inductance of this balun was approximately 200 µH as measured at series resonance (1642 kHz) with a 47 pF capacitor.  The actual common mode impedance of all constructed common mode chokes varied with frequency as shown in Table 2 and Figures 19 through 21.

Figure 8.  Homemade balun test box	Figure 9. 1:1 ferrite rod balun	Figure 10. 1:1 coaxial cable toroid choke balun. 10 turns of RG-174/U on 2 FT-114-43 cores	Figure 11. 1:1 toroid balun

Although I did not use a 4:1 current balun in the final configuration (due to its elsewhere documented inferior common mode rejection), figures 13 through 17 show the construction of a 4:1 Guanella balun similar to a design by Green¹².  This consisted of two 1:1 current baluns on FT140-43 toroid cores connected as shown in Figs. 13 through 17.  For these, I wound 36" lengths of the 18 gauge speaker wire on each of 2 FT140-43 toroid cores.  The inductance of these chokes was 100 µH when measured in series resonance (2320 kHz) with a 47 pF capacitor. I subsequently measured 105 µH inductance and 1.5 pF parasitic capacitance with a vector network analyzer.

Finally, I wound 11 turns (16 inches) of 18 gauge speaker wire on each of three FT114-43 toroid cores.  The inductance of these chokes was approximately 60 µH when measured in series resonance (3000 kHz) with a 47 pF capacitor.  On 23 January 2011 one of these chokes (shown in Fig. 18 and as T1 in Fig. 22) was wound on the feed line 3 inches below the antenna feed point and another one (T2 in Fig. 22) was wound where the feed line entered through the station window to help suppress common mode currents caused by asymmetry of the antenna environment.  These feed line chokes might adversely affect the use of the feed line and loop as a top loaded vertical antenna; however, this was not considered a significant loss due to the poor performance of the antenna in that mode. The third choke (T3 in Fig. 22) was used as a 1:1 current  balun for comparison with other baluns in the balun test box. 

Figure 12. The "H Special" balun box	Fig. 13. 4:1 "Guanella" current balun schematic diagram	Fig. 14. 4:1 current balun made of 18 gauge  speaker wire on two FT140-43 toroid cores

Fig. 15. Close-up: each core has 13 turns on each FT140-43 core with a cross-over at the 6th turn	Fig. 16. Toroid cores taped together	Fig. 17. 4:1 balun in box	Fig. 18.  1:1 choke at feed point 11 turns on FT114-43 core

The MFJ-202B noise bridge has a greater range than the antenna analyzer and was used for further impedance measurements.  The noise bridge was calibrated⁵ to increase the accuracy of measurements.  Table 2 lists the common mode impedance of each toroid choke over the frequency range 1.8 through 28 MHz.  Above 1.8 MHz all measured impedances exceeded 1000 ohms.

Figures 19 through 21 plot the common mode choke impedances graphically.

Fig. 19. Coaxial cable FT114-43 choke impedance vs. freq.	Fig. 20. Parallel line FT114-43 choke impedance vs. freq.	Fig. 21. Parallel line FT140-43 choke impedance vs. freq.

The MFJ-202B noise bridge was then used to measure impedance through either of the 1:1 baluns or 4:1 balun with the other end connected to the 37' feed line and loop antenna.  The standing wave ratios were calculated with reference to a 50 ohm source.  The results are listed in Table 3.

The impedance measurements varied in a complex manner as expected with changes in frequency and transmission line length. The feed line was ultimately lengthened to 51 feet (an electrical 1/2 wavelength at 7 MHz) to improve impedance matching on all desired frequencies (see 40 Meter Loop Antenna Transmission Line Optimization for 80 Meters and WARC Bands). None of the current baluns heated perceptibly with 100 watts transmitted power on any frequency. 

Figure 22 shows the present station antenna configuration.  With 1:1 common mode chokes at the antenna feed point and on the output of the antenna tuning unit and the total transmission line length adjusted to 51 feet, that is, 1/2 electrical wavelength at 7 MHz, the impedance presented was well within the range of the antenna tuning unit on all frequencies 3.5, 7, 10, 14, 18, 21, 24 and 28 MHz.

NOISE LEVELS

ACKNOWLEDGEMENTS Many thanks to Dan Maguire for the Transmission Line Details program. REFERENCES Computer Assisted Low Profile Antenna Modeling II, Milazzo, C KP4MD "Zip Cord Antennas - Do They Work?", Hall, J, K1TD, QST, March 1979, pp. 31-32. "Zip Cord Antennas and Feed Lines For Portable Applications", Parmley, W, KR8L, QST, March 2009, pp. 34-36. "Portable Antenna Notes", Wiesen, R, WD8PNL "Antenna System Measurements with the MFJ-202B RX Noise Bridge", Milazzo, C, KP4MD, March 2011. "Baluns: What They Do And How They Do It", Lewallen, R, W7EL, ARRL Antenna Compendium, Vol. 1, 1985, pp. 157-164. "Some Broadband Transformers", Ruthroff, C, Proceedings of the IRE, Vol 47, No. 8, August 1959, pp. 1337-1342. "Baluns and Tuners", Ehrenfried, M, G8JNJ "Common-Mode Chokes", Counselman C, W1HIS "A Ham's Guide to RFI, Ferrites, Baluns, and Audio Interfacing", Brown J, K9YC "New Method of Impedance Matching in Radio-Frequency Circuits", Guanella, G, Brown-Boverie Review, Vol 31, September 1944, pp. 327-329. "Build a Low Power 4:1 Balun", Greene, C, W1CG, QRP Homebrewer, Vol. 8, 2002.

Fig. 22. The station configuration updated in 2013.  The radio is connected through RG-58/U coaxial cable jumpers to an automatic antenna tuning unit (currently an LDG Z-11 ProII) and a 1:1 Guanella current balun (T3). A total of 4 feet (1.2 m) of the AS-18/50Z speaker wire are wound on the the toroid cores (T1, T2 and T3) for an approximate 51 feet (15.5 m) of total transmission line length of speaker wire to the full wave 40 meter horizontal loop antenna.