Useful Constructional Hints

Radio Record, Volume II, Issue 38, 5 April 1929, Page 28

Useful Constructional Hints

A Helpful Article Dealing with a Few Practical Difficulties

By

PENTODE

UDGING by the ever-increas-ing number of questions dealing with technical and practical difficulties sent in by correspondents to this journal, it surely indicates that there is S| an insufficiency of literature on the difficulties that the amateur builder and set owner is likely to encounter. Unlike most other hobbies there are countless pitholes for the inexperienced to fall into, and, sad to relate, even the most experienced trip over trivials now and then. Quite a large amount of space is devoted to these questions and their answers, and this is done because often 2 correspondent gives voice to trouble experienced by others, and the answer, although short, sometimes gives a wealth of information. This week several letters have come to hand from readers, and the queries are typical of a few of the most likely sources of trouble met by most amateur set builders. The first letter opened comes from Wellington. This listener, who uses a crystal set coupled to a two-valve amplifier, is troubled with motor-boat-ing when used in conjunction with a home-built battery eliminator. The only way in which to stop this is by diraming the valve filaments, so our correspondent -vrites. To begin with, two PM2_ valves should not be used for first and second stage unless a high ratio transformer is used with the primary in the anode of the first PM2 valve. The use of a semi-power valve in the first stage has no advantage in its favour. Here are a few hints on how to begin to cure the trouble. Use a general purpose valve in the first stage, biased about 2 or 8 volts on the grid and no more than 90 volts on the plate. The last

yalve should be run according to the specifications given by the maker. Negative grid bias must be given when using battery eliminators, and although the correspondent has not stated this it is assumed that he is using a O battery. Unless used with care, grid bias worked from the bottom end of the B supply is not recommended. If these simple remedies do not prove effective, then a choke and condenser must be used for the supply of the first valve. This is done by connecting a low frequency choke in series with the B battery lead to the first valve, and incorporating a 1 or 2 m.f.d. eondenser between the terminal B plus on the first audio transformer and the A minus terminal. This prevents all back coupling, which is the cause of motor-boating. A.M.K., of Wellington, has sent in a letter for advice on the power transformer of a battery eliminator he is building. The specifications are taken from a previous edition of the "Radio Record" on the construction of a battery eliminator. Having built the transformer he has applied a few tests and desires to know if everything is OK before building into the instrument. The primary, wound for the 280volt supply, is centre tapped for the 110 mains. The secondary is centre tapped for use as a double wave rectifier. The primary was connected to the 110 mains and a 230 v. 60 watt lamp put across one side of the secondary winding to the centre tapping of the secondary. ‘To our reader’s way of thinking, this lamp did not burn brightly enough, and when left for some time the transformer became warm. Here is the information our reader desires. When designing transformers,

the most important feature to bear in mind is the current that is expected to be delivered. The more current or amperage, the thicker will be the wire in order to carry this current. When a transformer is designed for a small drain, such as that needed by a battery eliminator, the guage of the wire on both primary and secondary can be kept down fairly small. If, however, a greater load is imposed on it, the transformer will heat up. An eliminator usually needs a current of 20 to 40 milliamperes, or about 1-33 of an ampere, and when the rectifier is of the full-wave variety, only half of this current will be drawn from each half of the centre tap. Now, a 2380-volt 60-watt lamp takes 60-230 eq. } amp. approximately. The windings cannot be expected to stand this for long, and this accounts for the heating experienced. Also the resistance of the fine wire will be in series with the lamp, and with this current flowing will cut down the voltage considerably. So this shows that a lamp is not a reliable way to test a transformer. The only way is to use a high resistance A.C. voltmeter. Failing this, a rough indication can be obtained by determining the ratio of the primary and secondary windings, and calculating mathematically. If a transformer with no load on the secondary heats up when connecting to the mains, the trouble is usually rather serious, and it will probably need rewinding. It usually indicates that the insulation has broken down, and that part of the primary or secondary is short circuiting. Eiven one short circuited turn is fatal, and results in a burnt-out transformer. Learning by Experience. WHILE on the subject of transformers, the writer had a peculiar experience, and the information learned may be helpful to others who are building large transformers. The one under discussion was wound to deliver 650 volts on each side of a centre tapping, and the secondary was wound with No. 80 DCC wire. Wach layer was separated with paper, and liberal applications of shellac varnish applied. A. particularly neat job was made, which seemed to be its only good feature, for when connected to the mains —

the coils were soon smoking. This waa with no load at all. To the tune of much unparliamentary -language the transformer was dismantled and the coils unwound. Hach moment it was expected to come across an insulation breakdown, but right to the last turn the windings were intact, with no sign of charring. More wire was bought, and the job started again. Thicker paper between the layers antl f larger applications of shellac were used, with the result that the transformers didn’t possess the good looks it had before. Again it was connected to the power, and once again smoke began to show signs of appearing. Not being possessed with unlimited patience, the only revenge that could be thought of was to let the thing do its worst. So, with a fire extinguisher handy, the onlookers stood well back to watch events. It seemed that we were unlucky, because after about 20 minutes of patient waiting, the smoke ceased, und it apparently was cooling off. Then it dawned on the whole party that the whole trouble was due to moisture in the windings, and that a large leakage was taking place from layer. to layer, until all the moisture had been driven off. Apparently the leakage was sufficient to cause the windings to heat up, and the first winding need never © have been undone. When building a transformer always let each layer of shellac dry off before starting the next, and when completed bake the whole in the oven until all moisture has been driven off. The Neutralising Problem. A READER has built a neutrodyne receiver, and has taken special precautions to thoroughly screen each stage separately, and yet the high frequency valves refuse to neutralise. How many constructors have been unable to neutralise th’'r sets and spend hours unwinding and rewinding their coils? It is comparatively easy to neutralise a valve, providing the feed back is only taking place through the valve electrodes. ‘This is the reason that sereening is used, to avoid the, coupling between the different coils. ,) There are other points that have to be watched to prevent the feed back and the average amateur is apt to overlook these. First consider the different B battery leads to each compartment in a screened set. Here is one of the

most common causes of an unstable receiver, and is sure to be the case if the whole becomes unstable as the B buttery runs down. The reason is obvious as the resistance of the battery increases with age and use, and the more or less powerful electrical’ impulses from the second stage are thrown back to the first R.F. valve. Now for the remedy. Insert a 1 mfd. condenser .in each screened compartment, one end being to the B + lead and the other end to A-filiment terminal. Yet another source of trouble can be traced to a feedback occurring through the filament rheostat, especially if one rheostat is used to control both RF valves. As this resistance is usually wired into the negative lead so fo obtain a small amount of bias, it Miso provides a resistance for the HY currents travelling to the filament, and actually throws back energy to the first stage through the screening. Asa remedy, connect the moving vanes of the variable condensers, and the fila- » ment end of the grid coil, direct to the I’-terminal. If this is not convenient through the metal end plates coming in contact with the screening, the rheostat would be better if inserted in the positive filament lead of the valve. Also, it is inadvisable to rely upon the metal end plate of the condenser for a connection to the valve, but to carry a length of wire from the moving yanes to the coil, and to the valve filament. Where no screening is used a cure to persistent oscillation can sometimes be effected by reversing the primary connec- . tigns on one of the RF transformers. Points About Valves. C T.M., of the King Country, has sent * in a letter saying that he has a 3-valve receiver, and uses 201A type valves, which are two years old. The trouble is a persistent rattle in all the speakers he has tried. If a few more details had been given regarding the conditions under which the valves were being used, it would have been easier to give more specific reasons for this jarring. Assuming that correct B voltages are applied to each stage, that the C battery is of the correct, value, and that the various transformers are all in order, the only things to check over are the valves. As our correspondent has had these over two years, they will no doubt have passed their most useful days. It was asked whether a set of Mfllard valves would be in order. A sbggested set would be PMD or PM65x for detector. PM65 for first audio and PM256 for second stage. The de-. tector valve would need 22 to 45 volts B. The first audio, about 90 volts B, with 8 volts negative bias, and the final stage 90 to 150 volts B, with grid bias strictly in accordance with the

maker’s data. It is not very often that a severe rattle develops in a transformer, and listeners experiencing a rattle in their speaker should look to the last valve and remember that a power valve will handle very little more volume than a general purpose valve, unless supplied with sufficient B and © yoltages. The Sereened Grid. FEW weeks ago Pentode wrote an article on the application of a screened grid valve to the ordinary broadcast receiver. A very interesting criticism has come to hand from an Auckland enthusiast, who has made a number of experiments with one of these valves in his Browning-Drake. An extract from this letter is given: ‘Dealing with ‘Pentode’s’ suggestion of a 1 to 1 ratio transformer, with the primary interwound with the secondary, this idea seems to be wrong. As suggested, I made up the coil, but in my case this gave reduced volume and selectivity, and I suggest the following reasons: (1) With any coil or RIF transformer, there is an optimum coupling, tapping or turn ratio, which cannot be exceeded if good volume and selectivity are desired. This is due to the damping effect of the preceding valve upon the tuning of the coil, In fact, a plain tuned anode circuit gives the effect of a coil having twice the losses that it actually does have. The 1 to 1 transformer is virtually the same circuit as the tuned anode as regards losses, tuning, etc.; (2) Although the SG valve is very nearly perfect, there is still a smaller amount of self-capac-ity, due to the fact that it is impossible to have a perfect screening action. There are also stray capacities in the leads, etc. As the primary is so closely coupled to the secondary, and having the same number of turns as the secondary, the feed back is increased through these stray and residual capacities, and selectivity is again decreased; (3) The step-up effect of the transformer is also lost." "Pentode" is pleased with this criticism, as it shows that more than a passing interest is being taken in the more technical side of radio. Without desiring in any way to pull our friend’s letter to pieces, the writer will indulge in a little criticism of his letter in return. Dealing with the statements in the order in which they are set down the first refers to the best number of turns for use in the anode of a screened erid valve. Concerning any amplifying valve except the last sta$e, maximum amplification can only be obtained when the anode impedance is at least twice that of the valve preceding it. This holds good whenever a valve is used for voltage amplification alone, which is the case in all but the final stage.

In each case the impedance of the valve is given, but of the coil data very little information is available for the set builder, and so he has to either copy reliable specifications or experiment with a number of coils until best results are obtained. Actual measurements of tuned coils show that the average coil on a 3in. diameter former and wound with space wound single wire has an impedance of 100,000 to 120,000 ohms. A. coil wound with Litz wire and tuned cun have an impedance of 200,000 ohms at any particular frecuency. This resistance is only at the special frequency at which the coil may be tuned to and falls to about nothing on either side of this point. A screened grid valve has an impedance of 100,000 ohms and, according to our statement regarding suitable anode impedances, it will need an impedance of at least 200,000 ohms in its anode to obtain maximum efficiency. But this is only obtained with a tuned coil wound with Litz wire. So it is evide.t that a tuned coil will have to be inserted in the anode of ~ sereened grid valve. The tuned anode method is here indi- eated. If two coils are very tightly ‘coupled and one is tuned the other behaves as though it, too, was tuned. A one to one ratio transformer with the primary in the anode of the screen grid valve and the tuned secondary as the grid coil of the next valve would then give, as far as the R.F. valve was c -.cerned, greater efficieney. This is dealing with actual amplification and the correspondent is evidently trying to combine this with selectivity. A screened grid valve has no claims for selectivity except when used under the very best conditions. These best conditions need coils to be wound with Litz wire und all condensers, etc., to be of the best quality. The correspondent mentions the effect of the damping of the valve upon the anode coil, which has the effect of decreasing the selectivity. This is true to a certain extent, but small bias batteries incorporated tend to overcome this, and this negative grid bias is recommended in any set using sereened grid valves.

Regarding No. 2, "Pentode" believes our criticiser is slightly mistaken in his knowledge of how a feed-back occurs. Certainly there is a very slight amount of capacity inside the valve which cannot be entirelv got rid of, but this small amount is insuflicient to produce a feed-back through the valve electrodes. The coupling between the anode and grid coils of the first and second valve has no effect upon the feed-back. The place to prevent any intercoil coupling is between the anode and grid coils of the same valve. The fact that a step up is lost through using a 1-1 transformer is unavoidable, but all the amplification is obtained from the valve, which more than compensates for the loss of the 8 or 4 to 1 step up of the R.F. transformer.