Free ferrite from TV sets in BALUN use

Free ferrite from TV sets in BALUN use JK De Marco, PY2WM 18/jan/2006, revised on 2/April/2009 After an article by Ian White, G3SEK, in RadCom magazine, suggesting the use of ferrite removed from deflection coil ( yoke ) found in TV and PC monitor sets for balun construction, and having already used the ferrite core obtained from horizontal transformers ( fly-back ), I decided to measure some baluns made with these cores and compare them to one made over a conventional FT240-61 toroidal core. Fig. 1 Figure 1 shows the schematic drawing of a balun. It consists of a single winding of bifilar wire or coaxial cable. It is known as current balun or Guanella balun, one of the first to describe its functioning. Note that it is placed in series with the transmission line. Traditionally we ve known the voltage balun or Ruthroff s, connected in parallel like a conventional transformer. This has been shown to be inferior (a commercial example of voltage balun is the W2AU balun, not to be confused with W2DU's which is a current balun). Photo 1 As an image is worth a thousand words, let's look at some photos. In photo 1 we see two baluns wound with bifilar wire. On the left 14 bifilar turns of #16 enamelled wire (kept together with

pieces of heat-shrink tube). On the right 7 turns over a ferrite core from a TV deflection core, the halves were glued with cyanocrylate instant glue. Photo 2 For maximum bandwidth it is interesting to use coaxial cable for the winding as shown in photo 2 (beware, it is wired as a transformer). Here I used RG-58 with the outer jacket removed in order to make winding easier. It should be well protected from weather effects. Sevick does not recommend RG-8X cable since the insulation, foamed polyethylene, would not prevent migration of the inner conductor (and hence shorting) due to the small radius of curvature. I also included a balun using a fly-back core, photo 3. And now for the measuring results. My set-up consisted of a Yaesu FT840 MF-HF xcvr, a commercial/modified and calibrated SWR and power meter, a laboratory 50Ω DC-8 GHz Weinschel Corp M1426 load, an oscilloscope Tektronix 453, an antenna analyser Autek Research RF-1 and a homebuilt LC meter (described in my website).

1 - INDUCTANCE TABLE 1 - number of turns and inductance measured in LF (about 700 khz) with the LC meter FT240-61 Yoke, RG-58 Yoke, bifilar Fly-back, bifilar wire wire # turns 14 6 7 10 Inductance 31,6 uh 14 uh 18 uh 99 uh In table 1 we can see that the fly-back ferrite has much more permeability than the others. I correlated these cores with those found on Fair-Rite catalogue (they are the manufacturers of the ferrite cores sold by Amidon in North America). I compared models with similar transversal area. This is a valuable way to infer the actual permeability. The yoke core has an intermediate permeability between ferrite materials 43 and 61, exactly the most used for transmission line transformers in MF, HF and VHF, this is good news! The fly-back core looks similar to ferrite 77, this is a high permeability but potentially lossy material. 2 REACTANCE AND LOSS INTRODUCED WITH WINDING IN PARALLEL TO THE LOAD Fig 2 This measurement, shown in Figure 2 with results in table 2, simulates a balun in the worst situation, when asymmetries in the antenna/transmission line induce currents on the external part of the transmission line (in case of a coaxial cable). These currents are called common mode and travel on the outer surface of the coaxial cable sheath. The impedance presented by the balun has to be high enough to impede this current. In this measurement, if the impedance is low the indicated SWR will be high. All the models did well in this test. TABLE 2 SWR measuring with windings in parallel to the load FT240-61 Yoke, RG-58 Yoke, bifilar Fly-back, bifilar wire wire Frequency SWR SWR SWR SWR 1,8 1,05 1,15 <1,05 1 3,5 1 1,05 1 1 7 1 1 1 1 21 1 1 1 1 29 <1,05 1 <1,05 1,05

3 USE AS CONVENTIONAL VOLTAGE TRANSFORMER Fig 3 This measurement, shown in Figure 3 with results on table 3, evaluated the performance as a conventional transformer. The blank cells in the table mean no measurements were taken.

TABLE 3 - balun used as conventional voltage transformer FT240-61 Yoke, RG-58 Yoke, bifilar Fly-back, bifilar Frequency SWR SWR SWR SWR 1,8 1,15 1,4 1,2 <1,05 3,5 1,2 1,3 1,25 7 1,45 1,5 1,4 1,2 21 29 2,5 The yoke core is similar to the FT240-61, and the fly-back core seems to be better on the low frequencies but losses were not evaluated. The fly-back core heats up with 100W on frequencies above 14MHz. It is the only situation where heating occurred. Please note that as a current balun it will never be subjected to this much power as long as there is a suitably matched antenna. A conventional transformer like this is found in interstage coupling in transmitters and receivers, not in antenna and high power applications. The reason is loss in the core. For example a loss of 0.5 db may be insignificant on an HF receiver or transmitter signal chain but at the output of a kilowatt amplifier it equals 108 Watts, this is what the core will have to dissipate as heat!

4 CALCULATED REACTANCE AND MEASURED IMPEDANCE TABLE 4 - Calculated reactance FT240-61 Yoke, RG-58 Yoke, bifilar Fly-back, bifilar Frequency Ohms Ohms Ohms Ohms 1,8 357 158 203 1119 3,5 714 316 406 2238 7 1428 633 813 4476 21 4284 1899 2439 13428 29 5715 2532 3255 17900 TABLE 5 - Impedance, measured FT240-61 Yoke, RG-58 Yoke, bifilar Fly-back, bifilar Frequency Ohms Ohms Ohms Ohms 1,8 370 170 1330 3,5 880 346 1080 7 >2000 830 630 21 614 880 286 29 380 490 216 I calculated the reactance using the inductance value obtained with the LC meter, table 4. Table 5 shows the impedance measured with an antenna analyser. A simple rule of thumb says the reactance must be at least 4 or 5 times the impedance of the load/source. Therefore 200 or 250 ohms in a 50Ω system. The yoke core balun would need 1 or 2 more turns to operate on the 160m band. I also took this balun and made a frequency sweep with the RF-1 antenna analyser while observing the impedance and noted a very broad peak of 1440 ohms when the frequency was 13.3 MHz. This is where a resonance should occur with 10 pf of stray capacitance. The fly-back balun has too many turns. The impedance decreases with frequency (even though the inductive reactance increases) because we are past the resonance frequency and now the impedance is capacitive. Permeability decreases while losses increase with frequency, complicating characterization. The yoke core is a good substitute for a large toroid core, as used in baluns for antennas and antenna tuners/couplers. With permeability between that of material 43 and 61 it is a perfect choice. 5 POWER HANDLING According to measurements made by Sevick, different cores with similar permeability have their other characteristics also similar. So we can use this attribute to determine power handling capacity of unknown cores, comparing to a known type with the same cross sectional area. A good indicator of power handling capacity is heating. If a balun under test suffers a temperature rise then it is unsuitable. Fly-back core: this is a high permeability core with high expected dissipative loss. This is only a problem with high power. Heating was observed with 100W at 14MHz in conventional transformer

use. A current balun will be subjected to much less current through its winding provided a reasonably matched load exists. Losses increase with frequency. For conventional transformer to adapt impedances this is a good choice as it requires less turns for a given inductance. One example would be a Beverage antenna for 160m and/or 80m. Losses must be evaluated though. Another good use for this type of core is in common mode filters for switching power supplies. In such a case the loss may come as an asset. Yoke core: it is expected that a balun with such a core will handle the full legal power. It is recommended to wind with a suitable coaxial cable capable of this much power. For powers up to 100W a good candidate is RG-174. Stacking is always a good option to increase power handling and even the bandwidth as fewer turns are needed for the same inductance, spurious capacitance is also lower. The inductance is proportional to the number of cores and to the square of the number of turns. The W2DU balun consisting of a string of ferrite beads works on this principle, it is a single "turn" of coax cable through a long core. 6 SOME ADDITIONAL POINTERS What happens when the core breaks? One of the fly-back cores broke while I was dismantling it. I simply glued the parts together with cyanocrylate glue. I compared it to a similar, intact unit which measured 1.5 uh with a single pass of wire, the broken core measured 1 uh. This is the result of the gap introduced by the glue thickness. I also measured a core with the original gap as used in the flyback TV transformer, there is a very thin (around 0.2 to 0.4 mm) plastic piece between the two halves, it measured 0.3 uh. See photos 4, 5 and 6.

Photos 7, 8 and 9 show the destruction necessary to reach the core! I also used a Dremel tool to make some cuts in some places; The fly-back transformers are generally filled with a though epoxy material.

On a later date I managed to discover more information regarding ferrites employed in TV sets. An interesting aspect is the resistivity of the materials. High permeability is associated with metallic MnZn mixtures having low resistivity so the core acts as many paralleled resistively shorted turns. Fly-back core resistivity is only 3 ohm/m. Resistivity is strongly frequency-dependent and decreases rapidly with frequency increase. Resistivity is expressed in ohms per metre units or ohms per centimetre units, in table 6 I have converted all that I have collected into ohms/metre units to allow direct comparison. Table 6 Common ferrite materials, their permeability and resistivity Material Permeability Resistivity 43 800 1x10exp3 ohm/m 61 125 1x10exp6 ohm/m 77 2000 1 ohm/m Yoke 350 1x10exp5 ohm/m Fly-back 1800 3 ohm/m Sabin, W0IYH, reported his measurements of three W2DU current baluns, one with 100 beads of 77 material and inductance 100 uh, other with 100 beads of 43 material, inductance 82 uh and another with 50 larger beads of 43 material for an inductance of 105 uh. He concluded that the high permeability is good for low HF bands but the other two are better overall and tolerate more power. He indicates the need of a balun also at the equipment side as a way to reduce pick-up of household noise. 7 CONCLUSION The yoke core is well suited to HF broadband use in baluns and transformers. The fly-back core is a high permeability material better suited to low HF. These cores can be found for free in repair shops. There's much more to discover as one delves into this subject. For the time being I'm very happy to report these findings and I thank Ian, G3SEK, for his enlightening RadCom note. References: Below is a list of what I consider the most informative texts. Some older publications contain incorrect information based on postulates that were demonstrated to be wrong. "Transmission line transformers", Jerry Sevick, W2FMI, ARRL, 2001. "Some aspects of the balun problem", Walter Maxwell, W2DU, QST Mar 1983. "Some additional aspects of the balun problem", Roehm, W2OBJ, Antenna Compendium 2, ARRL, 1989. "Baluns: what they do and how they do it", Roy Lewallen, W7EL, Antenna Compendium 1, ARRL, 1985. "Exploring the 1:1 Current (Choke) Balun", Sabin, William E., W0IYH, QEX July 1997. "Designing wideband transformers for HF and VHF Power Amplifiers", Chris Trask, N7ZWY, QEX Mar/Apr 2005.

"The 1:1 Current Balun", Roy Lewallen, W7EL: http://eznec.com/misc/ibalun.txt "Balun Balance Quality Test", Tom Rauch, W8JI: http://www.w8ji.com/baluns/balun_test.htm Fair-Rite Technical Information: http://www.fair-rite.com/newfair/pdf/broadband.pdf http://www.fdk.co.jp/cyber-e/pi_fer_tv.htm http://www.magtek.com.cn/sfc.html Resistivity: http://www.earthsci.unimelb.edu.au/es304/modules/res/notes/resistivity.html