FEBRUARY 1937 57

A UNIVERSAL TESTING SET FOR RADIO VALVES

By D. ERINGA.

Summary. A testing set is described which combines the functions of a receiving-valvetest unit and a universal measuring unit.

All switching operations are performed automatically by the closing of a contactbridge with 14,0 contacts initiated after inserting a selected code card into the bridge.The efficiency of a receiving valve is indicated by the deflection of the pointer into eitherthe red or blue area of the scale of the milliammeter. Measurements of voltages, currents,resistances and capacities and the output voltage of wireless receivers can also be carriedout with this apparatus. Resistances are measured on direct current and condensers with50-cycle alternating current.

Introduetion

For a long time the need has been felt for atest set for radio valves, which was easy to mani-pulate and with which the efficiency of radio valvesin service could be tested without extensive lahora-tory equipment. An apparatus of this type is requiredwhere a radio dealer wishes to demonstrate to acustomer the efficient or defective functioning of aparticular valve being used by the latter. The testunits hitherto evolved for this purpose were eithertoo complicated for unskilled users or did not afforda satisfactory and adequate test.

The testing set designed by us offers a verypractical solution of the problem, since it fulfilsthe following requirements:

1) It is suitable for all standard types of valves;

2) Manipulation has been made extremely simple;

3) Mistakes in adjustment cannot damage eitherthe valves or the testing circuits;

4) Schedules giving the limits of efficiency and ser-viceability for different currents and voltagescan be dispensed with;

5) Incorrect manipulation of the apparatus doesnot make the valves appear more or less ef-ficient than they are actually.

General Design and Manipulation of the Apparatus

The general appearance of the apparatus may begathered from fig. 1. It contains the followingcomponents:

1) Twelve different valve holders to take almostall standard commercial types of valve bases(European, British and American).

2) A milliammeter with a scale divided in twoportions red and blue. If the pointer reads inthe red scale the valve is defective, while if itgives a reading on the blue scale the valve isin good condition.

3) A neon lamp for detecting short circuits betweenthe individual electrodes.

Fig. 1. General view of complete testing set.

4) An electric glowlamp for detecting brokenfilaments.

5) A switch with eight pushes for testing the elec-trodes for shorts and continuity; this unit is sodesigned that only one push can be depressedat a time, hence not more than one switchingoperation can be performed at a time.

6) A potentiometer for adjusting the mamsvoltage to the correct value.

7) Two connector plugs for a test cord permittingmeasurements of voltages, currents, resistances,capacities or the output voltage of a radiorecerver.

8) A slot for inserting the code cards in the socalled"contact bridge". A II switching operations requiredfor testing a specific type of valve are performed

58 PHILlPS TECHNICAL REVIEW Vol. 2, No. 2

automatically when the corresponding code cardis inserted in 'the contact bridge. One or morecards are provided for each type of valve.

A view of the contact bridge is shown in jig. 2.One half of the bridge (on the right) consists of a

Fig. 2. View of contact bridge,

stationary plate with 140 contact pins. The otherhalf can be displaced by means of the handle onthe right hand side-wall of the apparatus and has10 contact bars. When the code card is insertedand the bridge is closed, the contact pins in frontof the perforations in the code card make contactwith the bars. The card is made of an insulatingmaterial, so that where there is no perforation thecontact pin is insulated from the correspondingcontact bar behind the card.

Fig. 3 gives as an example the code card for theAL 4 receiving valve. At the top of the card thecircuit of the valve under test is shown diagram-matically, and .indicates with which connectorsockets the various electredes are connected.The distinguishing numbers of the electredescorrespond to the numbers on the valve holdersand on the eight-way switch.The connections set up when the contact bridge

is closed after inserting a code card serve the fol-lowing pmposes:

a) Connectiontotherequisite filament voltage.

b) Choice of an 0 d e vol tag e, and if necessaryadjustment of the te s t vol tag e for conden-sers or resistances.

c) Adjustment of the sceen grid voltage, ifnecessary (with rectifying valves) selection ofalternating voltage applied to the anodes.

d) Adjustment of negative grid bias.

e) Fixing of correct loa din g res i sta n ces formeasurements on rectifying valves.

f) Connecting par alle 1 and shun t res i s t-anc e s to the measuring circuit, so thatalways the same red-blue scale can always beused for various types of valves.

g) Connection to the correct valve holder. In thisway each electrode of the valve holder receivesthe correct voltage through the perforationsin the card.

By employing the special circuit under g) abovethe use of a number of valve holders of the sametype is superfluous.A test is carried out as follows:The valve is fixed in the holder, and the card

inserted in the contact bridge. The glowlamp andthe neon lamp Ls, which indicate the results of thetest made, are not connected up through the con-tact bridge, so that for the time being the lattercan remain open. The bridge is only closed afterthe tests with the open bridge have indicated nodefect in the valve.

20615

Fig. 3. Code card for an AL 'J, receiving valve.

1) Immediately after inserting the valve, testthe continuity of the filament;

2) Depress the pushes of the eight-way switchin succession to test for shorts between theelectrodes.

-.

FEBRUARY 1937 TESTING SET FOR RADIO VALVES

z,Fig.,4,. Circuit diagram for testing set.

If test 1 and 2 give satisfactory results, thecontact bridge is -c los e d, and the following testsare dien carried out:

5) .A test for electrode disconnections or possible./ bad insulation.

3) A test for anode or auxiliary grid current;

4) A test for slope;

After _closing the bridge, manipulation of theapparatus is again limited to pressing in successi~n - ,the pushes on' the eight-way switch.

.60 PHILIPS TECHNICAL REVIEW Vol. 2, No. 2

Explanation of Circuits

The circuit arrangements of the testing set areshown in fig. 4; the principal 'circuits are describedbelow. In the circuit diagram the permanentcircuits are indicated by dots, while connectionswhich are set up by means of the contact. pinsare indicated by a circle. .

Filament voltage. The fifament voltage is fur-nished by eight windings marked Sa, Sa', Sa" in thecircuit diagram. These windings are able to furnishall filament voltages from 1 to 56 volts in stages of1/2volt, and partly ill stages Ofl/4 volt. The filamentvoltage is taken from the first two ba~s on the left.

Anode Voltage. The third and fourth bars fromthe left serve for setting up the requisite circuits.The rectifying valve Ll to whose anodes differentaltemating voltages can be applied according torequirements, 'furnishes the anode voltage. The'anode-voltage unit has been so designed that up toabout 30 mA the voltage is independent of thecurrent 'tapped.

Auxiliary Grid Voltage. The auxiliary grid voltageis tapped from a potentiometer which is connectedto the fifth and sixth bars from the left.

The potentiometer. resistance also acts as aload resistance for the aforementioned anodevoltage unit. The potentiometer is composed of theresistances Rn to R17' The duty of the neon,lamp.L4 isto maintain the auxiliary grid voltages constantàt 60, 80 and 100 volts.

Loading Resistmices for Rectifiers. For testingrectifying valves, resistances Rn to R17 are used asload resistances. In this case the A.C. supplyfrom the third mid fourth bars is not appliedto the anodes of the rectifying valve Ll' but tothe anodes of the rectifying valve under test.

: Negative .Grid Bias., The negative grid bi~s' istapped from the potentiometer in the usual way,the latter being fed from a separate rectifyingvalve L2• The seventh and eighth bars from the'left are provided for selecting ~he correct tappings., To measure the slope of the characteristic, thepotentiometer resistance can he adjusted so thatthat the negative grid bias is increased by 2 voltson pressing the, push M. Thus from the differencein the reading on the milliammeter the differencein anode current for a change of 2 volts in gridvoltage can be measured, this measurement being ,

, sufficient for most purposes.

, Milliammeter with Different Shunts. The two lastbars serve for connecting up a number of shuntsin para~lel With the milliammeter. The combinations.

possible with the different contacts are so numerous(about 70) that the testing ranges "bad" and"good" as marked on the meter scale of the testingset can be retained for all valves and measurements.

Protecting of Testing Set

Since the testing set is intended for the use ofunskilled operators, it is naturally possible that a,valve may be tested which already has a short-circuit,for instance between the grid and the cathode, orthat during test a short of this type develops inthe valve. It beèomes imperative, therefore, toprotect the milliammeter from damage. Since themethod adopted by us is not generally employed, itwill be described in detail below.The circuit employed is shown in fig. 5. A metal

,------{ M }----,

20607 ,Ra

I Fig. 5. Circuit diagram for instrument' with oxide cell.

rectifier Kl is connected in parallel with the am-meter as well as a correction resistance 'Ra. A"blocking layer" rectifier - (cuprous oxide -) cellhas the property that its internal resistance dependson the voltage across its terminals. Fig. 6 shows

112000

,

~~:\II -,I

I <;II

I --I---I-

1000

oo 0,1 0.2 0,4 0.5 0,6V20608

Fig. 6. Characteristics of "blocking-layer" cell.

the, resistance in ohms plotted as a function of theapplied voltage ~ volts.' If the voltage betweenPand Q is only of the order of 0.05 volt, the resist-ance of the cell is about 1000 ohms, and the bulkof the current applied externally flows throughthe measuring instrument whose resistance is

FEBRUARY 1937 TESTING SET FOR RADIO VALVES 61

200 ohms. If, however, the applied current rrsesconsiderably the voltage across Pand Q does notincrease in the same ratio, since as the terminalvoltage at the cell increases its resistance diminishesconsiderably. If the terminal voltage is 0.5 volt,the resistance is only a few ohms. A higher currentapplied to the terminal P thus does not overloadthe instrument, for the greater part of the currentflows through the cell shunt, whose resistancethen has a reduced value.

This solution of the problem introduces, however,a very undesirable secondary factor. The milliam-meter is not exclusively used for testing receivingvalves, but is also employed for measurements onrectifying valves in which the measuringcurrent isa rectified alternating current, i.e. a pulsatingdirect current.As a rule the circuits of rectifying valves are

so arranged that the anode current exists onlya small part of the cycle. The pulsating directcurrent is then approximately of the type showninfig. 7. In this diagram the broken line represents

Fig. 7. Current curve for rectifying valves. The broken linerepresents the mean value indicated by a moving-coil am-meter.

the mean value as indicated by a moving-coil am-meter. The instantaneous value of the currentduring part of a period is, however, about 10 timesgreater, and during this interval the voltage alsorises between the points Pand Q. It followsfrom the characteristic of the cell under discussionthat these peak currents will select the path throughthe cell instead of through the coil of the measuringinstrument with its comparatively greater self-induction. Unless suitable precautions are takenthe reading obtained on the instrument whentesting rectifying valves will be inaccurate. To elimi-nate this source of error a small choke of severalhenries (S9 in fig.5) is connected behind the cellKl' This gives the circuit Kl-S9 so great a reactancethat the resultant error is not too high. It is unfor-tunate that the coil S9 is necessary since it hasa certain non-reactive resistance, of about 20 ohms.This consequently reduces to some effect theshunting effect of the cell.

Testing a aeceiving Valve

A simplified circuit reproduced in fig. 8 illus-trates how any receiving valve, e.g; one with three

206/0

Fig. 8. Circuit for testing a pentode.

grids, has the various voltages supplied to it onclosing the bridge. The feed circuit of each electrodepasses through a single-pole change-over switch.The seven single-pole switches are the seven pushesof the eight-way switch shown in fig. 9. The testsmade by means of these switches are togetherwith the slope and anode currents of the greatestimportance.If a push is not depressed the corresponding

electrode is connected to its current supply. Ondepressing the push, the feed circuit of the electrodeis opened and the electrode is connected to the com-munal bar situated under the pushes, this bar havinga negative voltage of approximately 200 volts. Inseries with the communal bar is the neon lampL6 with the resistance R45 as shunt.

Fig. 9. Eight-way push-button switch.

When no pushes are depressed, the requisitevoltages are applied to all of the elëctrodes and themeasuring instrument M is in the anode or auxiliarygrid circuit so that the current to either of these

62.

PHILIPS TECHNICAL REVIEW Vol. 2, No. 2

electrodes may be tested. What happens when'the various pushes are depressed?

Assume that, ..according to fig. 8, the filament isconnected to the contacts 2 and 3, the cathode tocontact 1, the control grid to 'contact 4, the screengrid to contact 5, the suppressor grid to 6 and theanode to 8, and that all pushes are out of circuit.With a good valve the instrument Mwill then givea certain reading. If the pointer lies over the bluepart ofthe scale the anode current of the valve is suffi-cient, The anode current is then above the specifiedlimit and which is the end of the red measuringrange. If now the push marked M is depressedthe resistance R46 in the potentiometer is added tothe negative grid bias circuit, and the bias isincreased by 2 volts. The reading of the milli-ammeter will then be a few divisions lower, andthe half number of these divisions for a specificvalve and with a specific code card will be a measureof the slope of that valve at the testing or workingpoint.If .push 1 is pressed, the cathode feed circuit is

opened and the' cathode connected through 'theneon lamp 'L6 to 200 volts negative with respect to, the common neutral point of the whole circuit. With~ hot cathode the neon lamp L6 will burn withits maximum brightness. This test is very importantsince by its means it can be established whetheror no~ the cathode is emitting,_ ill~o when a dis-connection in the' anode circuit causes no anodecurrent reading to be obtained. On pressing push 4(push 1 will then automatically spring back intothe off position), a negative voltage of 2.00 voltsis applied to the control grid 4. If during heatingof the valve a short has developed between thisgrid, and the cathode, the neon lamp will burnbrightly. If however the grid insulation is good andthe grid connexion is not broken, the pointer ofthe measuring instrument will return to zero. Onthe, other hand, if the grid lead is broken in theinterior of the valve, there would be no changein the meter-reading on pressing the push,' Thebreak in the circuit may therefore he directly est-ablished. Since in this operation the neon lamp L6does not light and is therefore non conducting,metallic contact between the grid and the200 voltsterminal has been made for by connecting theresistance R45' in piuallel with the lamp. Similarcontact is made when pressing pushes 5 and 6.When push 8 is pressed the meter drops back tozero, since the circuit containing the instrument isopened, while thè anode is connected to the 200 voltsterminal and a test thus made for insulation.

Finally, if, push P is pressed, then with certain

types of valves, according to the arrangement 'ofthe perforations in the corresponding code card,the cathode heater insulation may be tested on200 V with L6 in circuit. If the insulation isinadequate lamp L6 will light.

Detection of Broken Filament

The lamp' L5 is in parallel (cf. fig. 4) with thefilament pins 2 a:n:d 3 and is fed from winding S2of the transformer through the resistance Rw Thevoltage drop in RIO is of such magnitude that thevoltage across' lamp L5 is just correct. On insertinga valve in the holder the filament is connectedin parallel with lamp L5, as a result of which thevoltage drop across R10 increases to such an extentthat the brightness of the lamp is considerablyreduced. Immediatelyon inserting the valve to betested. the lamp L5 will indicate whether or notthe filament is broken.

~Universal Measuring Apparatus

It is evident that an apparatus as comprehensiveas that described above can also be made to .carry out other measurements in addition to valvetesting. In designing the 'apparatus this possibilitywas given due consideration, Leads for connectingup with current sources which it is desirable tomeasure can be inserted in the pushes 1 and 4(fig. 4). To enable alternating currents also to bemeasured with the instrument incorporated in thetest set, provision has been made for connectinga rectifying bridge K2 in series.By inserting suitable code cards the testing set

can also be, employed for the following:

a) Voltage measurements: Alternating current a~ddirect current up to a ma~imum of 500 volts.

b) Current measurements: Alternating current and-direct current up to a maximum of 1 A.

c) Measurements of output voltages of radioreceivers.

d) Resistance measurements from 1 ohm to 5 meg-ohms.

e) Capacity measurements from 1000 fLfLF to200 f1:F.Fig. 10 illustrates a code card for resistance

measurements "in the range. from 1000 to 100000ohms.The apparatus can also detect short-circuits.

If the contact bridge is opened and one of thepushes 1 or 4 is pressed, a short in a circuit connectedto the terminals 1 and 4 will be indicated by theneon lamp L6 (see fig. 8). This is a very sensitive

FEBRUARY 1937 TESTING SET FOR RADIO VALVES 63

method, for ahe rapid discovery of shorts in areceiver. It is not necessary to use a special codecard for this purpose.

20615

Fig. 10. Code card for resistance measurements fro.n 1000to 100000 ohms.

Constructional Details

Contact pins. One of the principal factors inensuring the efficient operation of the testing setdescribed here is naturally the design of the 140contact pins. A cross-section through a contactpin is shown in fig. 11. The point of the pin is madeof silvered brass. If when being inserted exactlyvertically, the point comes in contact with one ofthe flat bars below it, it is quite possible that goodcontact will not be made should the contact surface

and the pin be separated by a microscopically-thin layer of oxide or dirt. The greatest danger ofthis occurring is in the case of contacts which carry

206/1

Fig. 11. Section through one of the 140 contact pins.

no current, such as the leads to the control gridof a receiving valve. To guard against this a conicalcontra-contact of solid silver is mounted on the baropposite to the conical apex of the contact pin.On closing the contact bridge these two conesslide over each other during part of the rotation,so that the contact surfaces are self-cleaning.

Contact Bridge. The contact bridge is constructedon such lines that the movable plate can be removedin a few minutes by undoing four nuts and with-drawing a spindle. The contacts are thereforereadily accessible, which considerably facilitatesmaintenance.

Mechanical Lock. Under the contact bridge isa mechanical contact. The main transformerof the measuring instrument is only switched onafter a card has been inserted in the contact bridge.This contact is so designed that the apparatuscannot be switched on if a card is inadvertentlyinserted incorrectly into the contact bridge.