The Five Way Valve

Radio Record, Volume IV, Issue 35, 13 March 1931, Page 16

The Five Way Valve

The Pentode in Action-II

by

W. M.

DAWSON

AMIRE MIW.T:

HE penthode has yet another considerable advantage over _ hormal power valves in that to produce. the same volume as a three electrode power yalve the penthode requires j only a fraction of the a.c. = ' grid voltage that would be required for the triode. If the penthode is fully loaded it gives a larger output of better quality with smaller input signal voltage, and of course a smaller negative grid biasSing voltage is required. Eliminating One Audio Siage. "AS a result of the penthode advan- _ tages discussed above, it is possible and practicable to eliminate the normal first stage of a.f. amplification with its valve and coupling transformer, _ In addition to the reduction of expense thus involved there is a definite gain in quality, as even the finest components are not 100 per cent. efficient, and every additional stage of af. amplifidation must add something to the overall distortion, The amplification possible with the penthode is of special appeal to the home set builder on account of the savings shown above. ‘The set may also be smaller and more compact, and requires less power for operation. As less preceding amplifiication is required, this gain can often be used to advantage to increase the selectivity of the set by making the radio-fre-quency stage couplings looser without reducing the signal strength too much. By virtue of the good high note amplification; the selectivity of the set may be much greater, than with ,a triode power valve, without seriously impairing the quality of the reproducon. j

Pentheode Can be Used in Every Set. "THE only external distinguishing feature of the penthode is an extra connection for the screen grid. The penthode can thus be readily adapted to existing sets by simply making an extra connection from the power pack (or B battery). Many enthusiasts are now using the penthode with excellent results. In gome eases, however, users have been unable to achieve success, and whenever these cases have been investigated the writer has invariably found that the operating conditions have been quite unsuited to the valve, considering the associated equipment, and it would seem that the penthode is UT. +

still a very much misunderstood valve with the average radio enthusiast. No valve’ can work to best advantage under adverse conditions, and the penthode is as docile and easy to handle as a triode once its little peculiarities are understood. No apology is therefore required for going into practical considerations in some detail, This, in conjunction with the release of new data, will, it

is hoped, enable readers.to use this remarkable valve intelligently and effectiyely. . Plate Impedance. HH matter of penthode plate impedance is a common stumbling block with experimenters, who _ frequently have yisions of attempting to "match" the valve. impedance to the speaker’ by using a high ratio stepdown transformer. It has been shown above that an essential fundamental of the penthode circuit is the very considerable valve impedance compared . with that of the speaker. What is unknown to most .people, however, is that recent research reveals a very Gefinite optimum load impedance for a given penthode valve, and this optimum value does not hold good for other types. The correct load impedances for the following well-known types are:C243 ok ce eee 4500 ohms D243 and H448 . 8000 ,, B443 ceveceese 6250 ,,- C443 ce eeeeeee 9000. ,, 4438 N & F443 . 10000 ,, C6438 ok eee ee ee 9500 _ ,, With a triode valve, a considerable amount of mismatching of the load impedance is possible before audible distortion takes place, though the undistorted output of the valve is reduced. In the case of the penthode, the correct load impedance is more important, and when deviated from, produces that shrill reproduction, with high note emphasis that many people wrongly consider to be a definite charactertistic of penthode reproduction. The penthode can and will give quality at least equal to the triode-repro-duction with well-defined bass frequencies and excellent balance.

The Moving-Coil Speaker. ET us consider first the case of the moving-coil type of loudspeaker, where the impedance of the voice coil is supposed to remain constant at all frequencies in the musical range. In practice we must qualify this by saying comparatively constant, because the impedance does, of course, rise at higher frequencies. Through a constant impedance load we must pass a uniform current, if we are to obtain uniform power expenditure, and as this "constant current" principle is the main feature of the penthode, it follows that the penthode power valve, plus moving-coil speaker, very closely approaches the ideal in respect to absence of frequency discrimination. We qualified, and said "very closely approaches the ideal," not because of any shortcomings of the penthode,

but because even the morying-coll speaker is not perfect in that its mechanical and electrical resonances introduce distortion of their own, and a fidelity curve of the acousfie output plotted against wattage input will soon disillusion any who fancy that the modern speaker closely approaches perfection, : In fact, if the speaker inductance, capacity and resistance are known, it is comparatively simple to proportion the values of the shunt capacity and resistance so that the effective impedsf ance of the speaker-filter. network 1 mains substantially constant over the useful .part of the .audio-frequency range. . The "plate-load" of the penthode thus remains constant, as in the case of the moving coil speaker. The effective impedance of the speaker circuit at, say, middle © (256 hertz), is reduced by the effect of the shunt circuit, and it may then be desirable to use an impedance adjusting’ device to have the effect of raising the speaker impedance to the optimum value indicated previously.. This can readily be done with an output choke or. output transformer, . and is amenable to ‘mathematical analysis precisely as shown for. the moving coil case. However, as hardly any readers have at their disposal means of measuring their effective..speaker im--pedance or other. characteristics, the" output choke filter ‘system will be adopted here as it is edmparatively easy to build one with the necessary high inductance, and tappings can be readily brought off, and the correct one determined by trial. — The circuit will then be as in Fig. 10. Alternatively the filter C2R may be shunted round the choke in which ease -- (a) The breakdown voltage across C2 will-be greater. (b) C2 requires to be smaller for the same effect, ‘ (c) R requires to be larg er for the same effect, A suitable choke for any of the penthodes mentioned in this’ article could "ae

_ — fae " Sa di finen be constructed by winding on a /din, square core some 9500 turns of 36 ‘S.W.g. enamelled wire, -~ The core window would require to be approximately 1tin; x Jin. in view of the desirability of layer insulation, and particularly of adequate end turn insulation from core .(the voltages developed across it being several times the plate voltage.) ‘The airgap in the core would be in the vicinity of 1.32nd inch. ~ mat . An alternative design, using fin. ' material for the core limbs, would require 6300 turns of 36 s.w.g, enamelled wire, the window being approximately jin. x fin, and the airgap about 8/64th in, mS Suitable tapping positions from the commencement of the coil (say the top r HT end) are given in the following able with the impedance ratios thus ayailable-they will cover all ratios possibly required.

If the whole choke: is common to both circuits, the ratio is of course 1 to 1. The taps must-be well insulated where they are brought out, It is interesting to note in passing ‘that’ the average magnetic speaker cone material is made unduly: hard to accentuate the high notes, to make up for the triode power valves deficiencies previously discussed. A softer paper Nn. be used to advantage with the enthode valve. Using the corrective methods outlined above, the penthode user will be delighted with the qudlity of repro- _ duction obtainable on eyen a magnetic speaker, Adjustment Hints.

‘poe negative grid bias should first be set at the value given by the valve maker for the plate voltage used. This can then be varied slightly under operating conditions until a milliammeter inserted in the Bf lead shows a minimum of "flicking." Do not worry unduly about whether the needle flicks upwards or downwards, as this is often governed by another factor which does not seem to be at all well known. This is the. correctness of the plate logd, and in general it may be stated if with correct bias and "B" yolt&ge a strong signal causes the needle to flick upwards, second: harmonic distortion is indicated, and a higher: load impedance is desirable. This is obtained by using a higher step-down ratio in the output choke or transformer. . If; on the other hand, the milliammeter needle kicks downward, the principal distortion is due to the presence of a third harmonic component, and the load impedance may with advantage be reduced (by lowering the step-down choke or transformer ratio.) Careful attention to these points will prove well worth while, Caution. " [NDER certain conditions, e.g., with very high impedance plate load, the penthode is capable of developing enormous voltages which may rise to & value many times that of the plate voltage, and this voltage can easily wreck the valve itself, or break down

the choke, transformer, condenser, or speaker insulation, As an open circuit is an infinite im--pedance, enormous voltages are developed if any part of the plate circuit ig opened with the valve in operation. It is therefore absolutely imperative that the speaker is not diseonnected, while the penthode grid. is being excited (i.e., while the implifler is functioning.) The above article is by no means intended to be an exhaustive treatment of the penthode, infact many alluring possibilities, such as its use as a detector, or in other positions have not even been hinted at, but it is hoped ‘that the notes on its handling as a last stage "power amplifier" will at least assist in drawing aside the veil, and making the penthode better understood and appreciated, .as it is undoubtedly destined to be.-used much more freely than at present, Fortunately, the human ear is :yery tolerant of imperfections in reproduc-tion-particularly in the matter of relative sound intensities, A loudspeaker is not necessarily of high excellence because it is a moying coil type, but a geod moving’ coil speaker represents the highest attainment in this field at the present moment, and as explained above, it is peculiarly suited to work out of a penthode power valve. Match the Moving Coil to the Valve. Most manufacturers state the average value of the impedance of the voice coils of their moving coil speakers, while in many cases the necessary transformer is included with the loudspeaker, Let us. take some representative figures, and work out the transformer ratio required to give the ideal impedance match, Case 1: H443N valve with speaker impedance of 50 ohms, then :- Impedance Ratio =

Arne impedance or a _ transtormer winding at a given frequency is proportional to the square of the numberof turns of winding, consequently the impedance relationship between two windings of a transformer is proportional to the square of the turns ratio, e.g., 2 10 to 1 turns ratio gives a 10 x 10 or 100 to 1 impedance ratio, so that the required turns ratio for our matching transformer becomes Y200, approximately 14 to 1. Case 2:-C443 with 10 ohm coil, 9000 Impedance Ratio = --- 10 9000 And turns ratio = Y--10 = V 900 = 30 And a 80 to 1 ratio step-down transformer is indicated. If you have tried a penthode-dy-namie combination without full success, just check up your operating conditions as above, and make the necessary correction in your transformer ratio. Magnetic Speakers. (MAGNETIC speakers without exception have a rapidly rising impedance at the higher musical frequencies, in fact the impedance of a socalled "high impedance’ magnetic speaker is commonly of the order of 2000 ohms at 256 cycles (middle C), rising somewhat slowly at first, with increasing frequency, but later increas-

ing more rapidly until a value of around 20,000-40,000 ohms is reachedat 5000 hertz. (The hertz is now the adopted" unit indicating cycles per second.) Think in terms of triode theory for @ moment, and. remember the glibness with which speaker impedance "matching" is spoken. The so-called "match" is made at a comparatively low frequency, say 256 hertz, but as the impedance at 5000 hertz is many times the value at 256, it follows that for high frequencies the speaker is very much mismatched. With, triodes, this results in a drastie reduction of energy expended in the speaker at high frequencies, and the consequent lack of proper proportion of these in the reproduction. Mechanical high frequency resonances cannot properly make up this loss, and the treble is unnatural, With the penthode, this increased impedance at high frequencies acts somewhat in the reverse manner, and the upper frequencies are accentuated. Further, if the listener is critieal, he will notice that distortion first occurs on the higher notes. Fortunately, this effect can be overcome without much trouble by paralleling the speaker circuit with another whose characteristics are exactly the reverse of the speaker winding, The speaker winding increases in impedance with rising frequency because it is inductive (the impedance being proportional to 2 xX 38,1416 fL where f is the frequency in hertz and L the inductance ‘in henries). The impedance of a condenser, on the other hand, deereases with frequency increase (being proportional to 1+ 2 X 8.1416 fC, where © is the capacity in farads). A condenser shunted round the speaker windings then has a marked effect in stabilising the plate load impedance at different frequencies, and this is really the main action of a "tonefilter" used with a magnetic speaker on a penthode output. An auxiliary effect is, of course, the greater "bypassing" of the upper frequencies, but the major effect is ob-

tained through impedance stabilisation. Such a simple capacity filter helps considerably, and is sufficient in certain cases, but a slight elaboration of this idea is an improvement, and ¢onsists of adding a fixed resistance in series as indicated in Fig. 9.

Impedance Ratio. Turns Ratio. Tap at 16 tol 4to1 1/4 "12tol ' Fto2 2/7 9tol © '. Btol 1/3 6tol_ 5to2 2/5 4tol -2tol 1/2 3 tol 1.75 tol 4/7 2tol 14 tol 5/7 1.5 tol 1,25 tol 4/5