Audio-Frequency Choke Coils

Radio Record, Volume II, Issue 48, 14 June 1929, Page 28

Audio-Frequency Choke Coils

Their Construction and Use

(By

MEGOHM

(Concluded from Last Issue.) DFORE proceeding it is necessary to rectify two errors which appeared in last week’s R article. At the top of the yy last colunin on page 28, second \ iline, the words "without gap" (>| were omitted after the word "chokes," so that it appeared that constructors were not recommeénded to make full-size eliminator ehokes, whereas they were to be recommended not to make them "without a gap." The second error was near the end of the article, the width of window in the output choke appearing as 1 inch, whereas it should be $ inch.

The resistance, 700 ohms, .of the output choke may appear to be high, and it certainly would be if used for eliminator smoothing, as 50 mils. passing through this resistance would drop 35 volts, and in two chokes double that amount. In an output filter, when a small power-valve is in the last stage, from 6 to 10 mils. plate current will be passing, and the’ drop of volts in this case will be from 4 to 7 only, and with

a larger power-valve passing 20 mils. the drop will be 14 volts. For plate currents of 20 to 30 mils. an effective choke of low d.c. resistance ean be made of 6,200 turns of 80’s on either a lin. core or the special stampings. With a resistance of only 250 ohms, the drop will only be from 5 to 74 volts. As this choke will carry 90 mils., the gap could be made 0.4, and the inductance increased over 20 henrys, still carrying 30 mils. with a good margin. A less bulky choke would result by using the 4700 turns of 32’s on #in. core, as a few volts dropped from 300 or more is not usually serious. ; If an output choke is of insufficient inductance there will be a tendency towards loss of deep notes. Some coils sold for output use are really designed for use as plate impedances in a choke-coupled amplifier, and as such the inductance and d.c. resistance will be too high to givé best results on the output, for which 20 to 30 henrys is sufficient. "B" Eliminator Smoothing Chokes.

E now come to a very important use to which audio chokes are put, and as chokes for this purpose usually have to carry a fair amount of current, the relation between core and windings has to be carefully considered, in order that the chokes may be of reasonable size without allowing too near an approach to magnetic saturation of the core. It may be assumed that the average five or six-valve receiver with a good

power-valve in the last stage must have about 30 miltiamperes available from the eliminator, and besides this the eliminator should have a good resrve. Fifty to eighty mils, may have to be earried by the elminator chokes, of which there should be two in’ series. As previously mentioned; chokes for this purpose must be provided with gaps, and a very useful choke is one that has already been specified-6250

turns of 32’s on a 1 x 1 core, with a D.C. resistance of 273 ohms, and capable of carrying 85 mils. with two gaps of 3/128ths each-this is the half of 3/64ths or .05. It should also be noted that .08 is just under 1/32nd. Constructors should get rid of any idea that eliminator chokes are to be small components compared to the transformer. A good choke will be very little smaller than the _ transformer, and for the home constructor there is nothing to be gained in operation by paring the size-in fact, the reverse is the case. Plenty of core material is essential in a good choke. This‘size is very suitable for a 200-volt eliminator supplying the average set. For larger output, where there is a 7% or 10 watt valve in the last stage with 300 to 400 volts on the plate, chokes of larger dimensions with lin. core are recommended, with 5800 turns of 380’s, which will keep the D.C. resistance low. With this size the full inductance value can be used, and though the lin. core previously speci:

fied with 32’s could be used, it would be safer to increase the gap about 50 per cent. without increasing the turns. It will be seen in the table that by decreasing the gap, fewer turns can be put on, the inductance being kept the same, but fewer mils. «re carried. Any of the reduced gaps may be utilised so long as the maximum mils, given is well over the amount of current to be actually carried. The larger the gauge of wire used, the more space it will occupy, and so a great number of the turns are farther from the core than would be the case with ner wire. This difference, actually great, and need not bejjmren into account for the gauges of wire likely to be used. For small half-wave eliminators such as are used for two-valve crystal amplifiers, it is usual to provide one choke only for smoothing, and if the plate current of the last valve does not exceed 9 or 10 mils., 980 turns of 36's on a lin. core without gap will give 20 henrys. But a better choke is made with a gap, 6200 turns of 36’s on a three-quarter core with .03 gap, window 2x tin. For this same purpose the core of an old andio transformer may be used if the central portion of the core is not under #in., ‘but better if or % A spool is made and sufficient turns of 38’s or 40’s put on to give about 20 henrys, judging from figures given in the table. The laminatio are all assembled the same way round, and it may be necessary with some patterns such as H’s to place built-up strips of stalloy across the open end to form the gaps. The gauge of wire should suit the mils. to be passed. From 6000 to 800 turns will probably be required. Constructors should avoid using heavy gauges of wire on smoothing chokes on account of the large amount of space occupied. Any gauge larger than 30’s should be avoided except in special cases where its use is warranted. Stalloy Stampings for Cores. LASst week a diagram was given showing the dimensions of the useful stalloy stampings now stocked by Messrs. Thos. Ballinger and Co., Ltd., Wellington. The use of these stampings does away with the cutting and flattening of stalloy strips, and they may be assembled either alternately with no gap, or all one way, giving three gaps which for ordinaxy purposes may be filled with three strips

{of visiting card thickness, or if the current to be carried is comparatively large, two thicknesses may be used. The combined gaps should, if possible, be computed to equal the amount given in the table. The dimension of the window in these stampings is 2 5-16 by fin., and allowing for spool ends, centre, and taping, a space measuring 2 by 7 is available for: the winding, and this will be taken as 1isq.in. This space will accommodate turns of enamelled

wire according. to the gauge as _ follows :-

Or d.c.e, wire aS Lolows:-

meme These stampings are what is known as "shell," as the core is divided, half. passing round each side of the ‘coil, forming two windows. As compared with the "core" pattern with only one window, the "shell’ is slightly better for a transformer, but the difference is small. The stampings are sold in lots’ of half a gross of each, T’s and U’s, and this quantity allows of the core. being built to 1 1-8in., so that if may then be used for choke purposes aS a one-inch square core. For the.pro-

cess of winding, a wooden former should be made 2}in. long and 1 1-8 by a shade over 15-16in. to allow the centre strips to slip in easily. On this former a spool is built with 1-8in. ends, and 2in, wide inside. The spool ends will be 2tin. square, and should be of strong material, as in choke winding, not in layers, there is a continually increasing outward pressure as the wire is wound on, and the heavier the gauge, the greater is this outward pressure, which will curve outwards any but the strongest material. It is therefore necessary when winding any

chokes to support the outer ends of the spool in some way that will withstand the pressure of the turns. A square of wood with a hole to slip over the spindle. and small nails in the latter will answer the purpose. ‘Upon referring to the table of gap chokes, we find that. 6250 turns of

80’s can be put upon this core to make a 20 henry smoothing choke with onethird of .05 in each gap, which is 1-64th in each. D.C. resistance, 250 ohms. When using these stampings it is best to see that the window was fairly well filled if possible and in some — ee er

cases.a suitable gauge of wire can be used with this object, so that ,previous remarks regarding gauges. heavier than 30’s need not apply. here. _ The method of using wooden clamps on these stampings requires some mention. «The clamps are best put on parallel. to the centre core, the laminations of which are packed tight by means of narrow strips cut off spare ‘ Jaminations and forced in on the out_side of both sides of the core. Subsequent strips must have the enter- . ing corners sloped off for safety and must: be pushed in between the first: extra strip and the core. Clamps are din. thick by din. wide, if the winding is a full one, and will be 44in. long, drilled 34in. centres. Threaded brass rods, 2$in. each. ' Where a low d.c resistance 20-henry choke is desired, 4700 turns of 28's enamelled may be used, having a resistance of 132 ohms. Gap .03in. _ Some of the. good makes of eliminators are now provided with 30-henry smoothing chokes, and some construc- _ tors might like the idea of putting in one if not’ both, of this value. "If one only, it should be the one nearest to the rectifier. A suitable choke can be selected from the table. If the stamp-: ings are to be used, 7400 turns of 32’s or 5700 turns of 30’s can be put on, with d.c. resistance 385 and 227 ohms. respectively. The average turn of ‘a full winding is 0.6 feet, and this multi-

plied by the number of turns.gives: the feet of wire used, from which the resistance can easily be calculated, by reference to a wire table. giving ohms per 1000 feet. It may be mentioned here that Some factory-made chokes sold for "B" eliminator work have a d.c. resistance as high as 500 ohms, which looks rather a high figure, though the drop will.only amount to a half volt per mil. Choke-Coupled Amplifiers. NSTEAD of the popular transformer or resistance amplifier coupling, choke coupling may be employed. It really consists in replacing the plate resistance of a resistance coupling by an impedance or choke of about 100 henrys, or sometimes ‘as much as 150. This system gives good reproduction, and avoids the great drop in plate voltage which is inevitable with resis-tance-coupling. .In. .dual-impedance amplifiers the grid-leak of the resistance-coupling is replaced by a choke of 200 to 250 henrys, and this is known as the grid impedance. The merits of the system will not be discussed here. In practice, a 100-henry choke gives good results in a plate impedance, and may be wound with 38’s or 40’s wire, 13,400 turns on a = core, gap .03, or 10,400 turns on a lin. core, as shown in the table. Home constructors will find it very convenient to use the secondary of any

old audio transformer as a grid impedance. A transformer witha brokendown ‘primary will answer. well. Chokes for Plate Leads. CHOKE of 15 to 20 henrys answers well in plate leads to prevent "motor Loating.". Very often the. core of an old audio transformer may be turned to good account for this purpose. In the detector plate lead, high d.c. resistance is of little consequence, so that the wire may be of very thin gauge, 38’s or 40s. Low Voltage Chokes. EXPERIMENTERS may require chokes suitable for "A’:eliminators or other low-voltage smoothing, so specifications are included. Values from 0.5 to 0.1 henry are used for this purpose. To pass 2 amperes continuously not less than 20's wire should be used; but 18’s is better, as the d.c. resistance must he kept as low as possible: for low Voltages, as appreciable drop is not permissible.’ Utilising the stalloy stampings, 400 turns of 18’s d.c.c. would give a value of over 0.1 henry, with a .05 gap. To pass not more than one ampere, 800 turns of .22’s d.c.c. will. give.0.5 henry with a gap of 1-64in. Other heavy duty chokes are given in a table. The heavy duty. chokes are wound with enamelled wire, preferably in layers. with paper between, and of gauge to carry the current-20’s for 1.5 amp., 22’s for .7, 24’s for .4, 26’s for .3, 28’s for .2. The chokes with the larger gaps are to be preferred.

Volts Dropped: in Chokes. | T"?. voltage of ‘direct current passing through a choke is always reduced. by the resistance’ of the wire. The thinner the wire and the greater the number of turns, so does the resistance increase. To find the voltage that is to be dropped we must-first find the d.c. resistance of the wire comprising the chuke. This is obtained by first ‘measuring the "average turn,"’ which is actually the length of the centre turns of the coil, reckoning from core to last Jayer. This measurement multiplied by the number of turns gives the total length of wire, and by finding the resistance of 1000 feet in a table, such as appeurs in the "Listeners’ Guide" the resistance of the coil can be calculated. Multiply the mils. passed (expressed as amperes) by the resistance, and the result zives the volts dropped. Note particularly that the more current passed the greater the loss of voltage. , Every mil. passing through 1000 ohms drops, one volt, so 50 mils. drops 50 volts. Jévery mil. passing through 500 ohms drops }-volt, so 50 mils. will drop 25 volts; 50 mils. through 200 ohms drops 10 volts; 30 mils. through 300 ohms drops 9 volts; 100 mils. passing through 150 ohms drops 15 volts, and through 250 ohms drops 25 volts. The foregoing figures give a good . idea of the voltage drop across the average choke, which in "B" eliminators with two chokes, will be doubled. General Points.

[fF there is hum in an eliminator. it may be due to saturation of the chokes if not insufficient capacity of smoothing condensers. In such a case it will pay to increase the gap in the chokes and note the effect. The given dimensions of gaps cannot be more than approximate, as several other factors affect the general performance. "The chokes given in the tables are worked out on a basis of a flux density

of not. ‘more than 35,000 lines per inch, which gives a good, margin to meet the case of rectified alternating. current -with little. or no smoothing being applied to the choke. peal THE action of a choke coil takes place between the turns of wire and the magnetic lines of force -of ‘the field of the coil. The interscetions between the lines of forte and the turns produce within ‘the coil'a "back electromotive forge," so called because it always opposes the alternating current applied to the coil ,and which is pro~ducing the "back e.m.f." _The presence of an iron core increases the back e.m.f., which therefore causes the coil to offer greater resistance to-any chang (alternation) of current, ‘so causi} ® ‘the maximum "choking" ‘effect to ‘ta . place. That is to say, the choke will freely. pass direct current, but offers 4 very high impedance to alternating current. , Always make sure that laminations of ‘cores are packed tightly, and free from ‘any chance of rattle, which if present in an .ontput filter choke will Spoil quality. . All gauges of wire given in ‘this article are s.w.g. . Machine-cut Stalloy makes. a more compact core than the average handcut and flattened material. All choke specifications here given are calculated upon:a flux density of 35,000 lines to the inch, which. gives a very ‘liberal margin below: saturation point. Stalloy permits of working up -to 55,000 lines ‘without saturation, .so that chokes ‘given in the tables. may be — made to carry a reasonable oyerload if necessary, though it ‘is not recommended.. . ‘ ‘

When a choke of high inductance is required to carry heavy current, this can only be done by a large increase in both the number of turns and the width of the gap, thé latter sometimes being as much. as } inch in-a 100-henry choke on a 1 x1 core, which is thus enabled to carry 50 mils. with 18,000 turns of 36’s wire, but with a drop of 114 volts. : ’ In the table given for ordinary chokes, the gaps have not been increased for the higher inductances, because as a rule these are not required for purposes using heavy current. The heavy duty chokes will carry as much current as the average constructor or experimenter is likely to require. — When the length of window exceeds 1tin., the larger gap given should be used. It has been decided that for the meantime the particulars of chokes without gaps will be held over, as data on hand appears to be insufficient to enable a useful table to be. compiled. In any case, the use of such. chokes is. somewhat limited, owing to the comparatively small current allowable, but where this factor is not of consequence, a good reduction in bulk is often possible, on account of the shortened winding required when no gap is used. This article has been written in an endeavour to clear up the difficulties confronting constructors owing to the general lack of information on the subject, and the writer trusts that his efforts will fulfil the desired object.

Heavy Duty Chokes-Stalloy Cores] Core Uxl WxIY 2x2 [kenry] Turns Igaplanp] Turns IgaplAmA Turns baplAmp 5 z70/05] 7 860] 15 750 "12 I"5 7500| 251 "4 3400| "081 3 30007 1*4 10 3300/o5| 2 2050| [05] 3 15 5800 2 B900 3 2600- "05 7225 6700 2 7900| 2 3 4200| 3 20 4900 "05/20i 970 3 4 2950/05| 2 50 Iz500| 4 3 9500| "2 "225 4800/05| '12 floo 13200 "2 2 6600/a5|

Ghoke Coils with Gaps_Stalloy Cores Core Y xlz % X % X i9 xIY4 Henrys Turns_ IGap| 'Milst Turns Igap Mils] Turns [Gapl Milsl Turns [gae] Mils] 7m ioo00 051 dN: 55 6900 '05 80 5200 "05 IIO 4200 05 I40 5 8000 [o3 42 5300 "03 55 400 [03 85 3450 03 IIO 12000 05 48 7900 "05 70 6250 "05 90 4800 "05 I20 20 9300 [03 37 6200 03 48 4700 "03 70 3750 '03 90 14600 05 37 ioooo: 05 57 7400 '05 75 5800 '05 Ioo 30 1iso0 o3 30 7500 [03 45 5700 '03 50 4500 [03 70 I9000 ro5 30 12700 "05 45 9500 "05 60 7500 (05 75 50 I5000 03 23 9800 [03 35 7300 '03 45 5900 (oa 58 21000 '03 16 17500 "05 30 13400 '05 42 10200 "05 55 100 13400 (03 25 10400 "03 37 8200 "03 40 Garcying Capacity of Wire (swo mils 200 130 1Oo 80 50 32 20 mils Gauge 28 30 32 34 36 38 40 Gauge

\ S.W.G. Turns. 8.W.G. Turns. 26 3,100 34 18,000 28 = 000 36 19,000 30 6,300 38 34,500 32. 9,400 40 54,000

S.W.G. Turns. §8.W.G. Turns. 18 420 26 1,600 20 550 28 2,160 22 840 30 2,500 24 1,190