Servicing old radios: useful tips
We have just found an old radio and are ready to power it, to see if it works. Let us have patience before plugging it, or probably we will just see smoke coming out from the cabinet. A radio that was stored for years on a dusty shelf must be awakened gently to return to its reliable operation and provide us with warm and pleasant sounds for years to come. Here are few generic notes on how a radio can be easily returned to life. Of course, more details can be found looking at each specific model.
Never attempt to hastily connect the radio to the mains just to see if it works. Always follow a soft-start procedure as indicated in these notes.
Part 1 - Tools, useful and less useful tools and instrumentation
In addition to common tools, as screwdrivers, pliers, soldering iron and some brushes, we need few basic instruments. Some of the types, commonly in use in the past, are no longer manufactured. Some of the instruments still in production today can be more or less unsuitable for vacuum tube sets, where voltages could well be in the order of several hundreds volts. Anyway in many cases, old instruments still operating can be easily find at reasonable prices. Here an overview of instruments usually found in old service shops and their usefulness.
1.1 - Variable AC Voltage Source – *** Essential
A step-down transformer, a Variac or even an incandescent lamp, to reduce the mains voltage during the early power-on, is the most useful tool. The step-down transformer is to be preferred when servicing radio sets with chassis directly connected to the AC mains. When a transformer is used, a 2 to 1 voltage ratio can be selected. We will start at half voltage and switch to the full voltage after the reforming of filter capacitors. A soft start-up procedure is advisable even for radios already serviced in the past, if left inoperative for several months or years.
Of course the combination of an insulation transformer followed by a Variac is the most versatile and safe solution.
1.2 - Voltmeters and/or Multimeters – *** Essential
Most of the digital multimeters, today available, have high input impedance, in the order of 10 Megaohms. and are suitable to measure voltage and resistance inside our radio. Unfortunately the voltage range of modern multimeters is limited to few hundreds volts, unsuitable for some old electronic sets. I burned some digital multimeters, attempting to calibrate the HV supply in old Tektronix oscilloscopes. For this reason, I prefer old fashioned moving coil multimeters but, in this case, an additional high input impedance VTVM is required for accurate measurements on grid circuits.
Here are a couple of Avometer 8 and two of the most known VTVMs, the RCA Senior VoltOhmist WV-98C and the HP 410B. Note that the Avometer on the left, a model 8 MK IV, has separate inputs for AC/DC ranges up to 2500 volts, useful to service some old lab instruments. HV inputs are no longer available on newer releases of the same model. I love Avometers because of their smooth taut-band movement, their performances, stable through the years, their effective protections against overloads and their incredible ruggedness, that make them indestructible.
1.3 - Insulation Meter – ** Useful
In contrast to common believe, the insulation meter is not essential to find leaky coupling capacitors. DC leakage currents in a capacitor are always present to some extent, when a DC voltage is applied to its armatures. It is important to evaluate how these currents affect the normal operation of involved circuits. The insulation meter returns at a given test voltage a resistance value, which can be represented in parallel to the capacitor itself, to calculate the DC shift of involved nodes. But the actual shift can be directly measured with a voltmeter, provided that its impedance is high enough to prevent appreciable influence on the circuit under test. A VTVM, with its typical 10 Mohms input impedance, will just add a predictable 10% error, when measuring the voltage across a 1 Mohm grid resistor, and even less for lower resistance values. An insulation meter can give unpredictable errors, if the test is run at any voltage higher or lower than the actual operating one.
Rather the insulation meter it is useful to locate leakage paths that could impair operation or safety of the set being serviced. It is not uncommon to find leakage paths, as burned paper surfaces in wound components, power or output transformers. Other critical paths can be found in phenolic terminal strips and sockets, in AC power switches and even in leaky paper capacitors from AC mains to the chassis. The insulation meter can also be used to evaluate in advance the suitability of NOS capacitors, before replacing old leaky ones.
1.4 - Capacitance Meter – ** Useful
A capacitance meter can be used to check old electrolytic filter capacitors. It is not uncommon to find even NOS electrolytic capacitors partially dried, with capacitance readings well under the nominal values printed on their cans. In the power supply section a dried capacitor may be the cause of considerable ripple and, consequently, of a loud hum in the speaker. If we have an old analog ohmeter, with a little experience, we can evaluate the capacitance of filter capacitors by ballistic reading, comparing the pointer deflection, caused by the unknown capacitor, against the deflection for a good one of the same nominal value.
The capacitance meter can also be used to match coupling capacitors in push-pull audio amplifiers. Low capacitance ranges, under one nanofarad, are of some usefulness to find open styrene or mica capacitors in RF/IF resonating circuits, when we are not able to tune them to the specified frequency. Most useful ranges should cover from about 1 nF to about 1000 μF. Of course, AC impedance bridges can even measure other parameters of limited usefulness for service purposes.
Here the 1650-A Gen-Rad impedance bridge and an inexpensive digital LC meter. Capacitors must be carefully discharged before measuring their value, to prevent damages to the meter.
1.5 - RF Signal Generator - ** Useful for alignment, but…
A RF signal generator can be very useful when the RF or IF stages of a radio must be aligned. Unfortunately it is impossible to find a single generator capable of covering all the needs arising during the alignment of radio sets. Calibration of AM radios can be somewhat simpler, but FM sets ask for quite special generators. In the past the alignment of RF/IF stages in FM radio sets was usually carried out by few service centers, equipped with dedicated instrumentation.
The alignment of AM receivers can be performed with an AM signal generator, usually covering from about 100KHz to some 30MHz. Unfortunately the accuracy of most of these generators is quite poor, when we need to adjust the IF to 467 KHz, rather than 465 KHz. A crystal calibrator or even a frequency meter are recommended, in addition to the generator itself, to set its frequency close to the desired value. A good output attenuator will help us to properly drive the various stages of the radio, adjusting the signal amplitude within the dynamic range of AGC. Here an URM-25D military signal generator.
The alignment of the tuned stages in a radio could be quite difficult for a beginner, even asking for special non-metallic tools. Anyway, we will carefully follow the calibration procedures given for each radio model or family.
1.6 - Tube Tester - * No need at all, but impresses friends and visitors.
The tube tester was, and still today is, one of the most expensive and useless instruments in the radio repair shop. In the past, a simple emission tester could cost as much as the equivalent of 500 to 1000 new receiving tubes and over than 2000 tubes pulled from surplus sets.
The problem is that, when a vacuum tube fails, the tube tester can only say that probably the tube might be defective. Under no circumstances, the tester can be used to restore the operation of the radio. So we realize to have spent time and a lot of money just to know that we need a replacement tube. But, if we decided to enjoy the sound of a tube radio, in any case we must have at least one full set of spare tubes. Then we can simply replace the doubtful tube with a new one and see if the radio works again.
Even the use of the tube tester to match a couple of tubes for a push-pull stage is quite questionable, since we should assume to have many and many tubes of the same type, from which select the matched pair. This TV-7D/U was one of the most versatile military tube tester for in-field service.
1.7 - Oscilloscope - * Some usefulness, with a warning
A small oscilloscope can be useful mainly to evaluate the ripple on DC power supplies. Unfortunately modern instruments and attenuated probes do not withstand the voltages commonly found in vacuum tube sets. For this reason, should an oscilloscope be used to see the ripple on B+ rectified voltage, it is advisable to insert a 10:1 or 20:1 voltage divider, or even a suitable DC blocking capacitor, between the circuit under test and the oscilloscope itself. Here a vacuum tube Tek 515A service oscilloscope:
I was asked recently for a DAC90 mains lead for "testing" by a dealer. I advised him that a DAC90 is worth more untested unless professionally restored/repaired. Anyone that knows how to properly "test" an old set doesn't need a mains lead for test purposes as they can easily temporally attach a figure 8 socket to the mains switch and perhaps use an isolating transformer on a "live chassis" TV or Radio (not limited to Valve AC/DC sets at all, e.g. TX9 chassis and all SMPSU driver side PCBs have live chassis or "common rail" no matter which way round the mains is due to a bridge rectifier).
Many parts are not easily replaced, or not replaceable by original. Even if the radio doesn't smoke, the Dropper / Mains transformer, output transformer, output valve or filter resistor can be damaged such that it subsequently fails when the expert person has replaced faulty anode/grid coupling capacitors or short circuit cathode capacitor.
Cheap DMM are actually only 1M input impedance on all ranges and on the 750V range are only safe to about 240V!
A 20 K ohm per volt Analogue meter is a misleading description. What it really means is that full scale is 50uA on ANY range. A modern DMM is usually the same impedance on every voltage range.
|3||60,000 = 60K|
|100||2000K = 2M|
Resistance vs Range on a 20kOhm/volt analogue Meter
So an AVO 30KOhms/V or cheap Eagle 30KOhms/V analogue meter loads the circuit less than a cheap DMM at 100V range or better. A 10M Ohm fed Screen grid from a 90V HT can be measured on 300V range as you will see over 1/2 of correct value, yet cheap DMM will read about 1/10th! Thus capacitor leakage can be estimated. Such checks can be done with all the tubes unplugged.
So I would emphasis what Emilio says
Never attempt to hastily connect the radio to the mains just to see if it works.
Especially if you are just reselling a radio you bought or found and have no experience.
Count of Thanks: 42
I would rather try out a possible radio replacement tube in a rugged tube tester than in a radio that might not be too tolerant of a grid short or a gassy "NOS" tube in my inventory. If a replacement tube has lost its vacuum, the excessive filament current during its death throws might forever wipe out the filament windings of an irreplaceable mains transformer or damage other tubes in a series-filament string. While the cost of a TV-7D or an AVO tester is off-putting even at eBay prices, posession of an economical tester like those manufactured by Eico can save a radio from a killer tube or at least instill confidence that my tube inventory is more than a collection of pretty glass.
I think a good tube tester is more use for a new design outside of realm of data sheet. With a €60 Bench PSU you can test the filament. With filament powered you can use a capacitor/insulation tester to look for electrode shorts etc.
Similarly a 'scope or spectrum analyser can be used, but both more use for design. A 'scope is hardly of value repairing radio sets at all, but can be invaluable on TV repair.
A "basic" tube tester for radio repair isn't much better than a PSU. For a TV with the large number of tubes a tester may be of more value. In any case I think a tester has to be a fairly high end model to be generally useful. A Go/No tester isn't of much value. But not many can afford a VCM163.
So I'd put a tube tester at the bottom of the list. A Noise Generator + Spectrum analyser, or Wobbulator + Scope or Tracking Generator + Spectrum Analyser is more use.
If you have 2,000 + Old Stock tubes of unknown "goodness" then a Tube Tester may be worthwhile.
So nice to have, but you can live without it!
of course I am just talking of my experience. Through the years I serviced several hundreds of old radio sets, but even many and many communication receivers and test instruments. Still today I have retained for example about 20-25 Tek vacuum tube oscilloscopes, about 80 tubes each, and some Collins, Hammarlund and other military receivers, 25 average tubes each, all working and calibrated. Well I never used a tube tester to service them. In the years I bought several tube testers, including an AVO, just to select some thousands of old mixed tubes and I never found the time to do the selection.
A tube tester was useful in the past in the acceptance of lots of tubes from suppliers and today it can be useful for vacuum tube dealers, in order to check their tubes before shipping them to customers. No usefulness at all in the service of an old radio set. My article was planned to include other chapters with tips on how use available instruments, and even our senses, to service a radio. A gassy amplifier can be sometimes identified by a simple visual inspection or anyway by a simple check of the grid bias. A smooth power-on will anyway prevent any damage to critical parts, as transformers.
But I will talk of this in the nest days.
Cleaning and fixing of preliminary problems encountered in the visual inspection
Usually our old radio must be first cleaned from dust inside, carefully removing the chassis from its cabinet. Brushes and a small vacuum cleaner will be useful to remove dust. Solvents, alcohol or water will be used with the greatest precautions, since they can remove writings or even rub off the same tuning scale: inks used in a radio sets, even those on vacuum tubes, were often water-soluble.
A slightly damp, lint-free cloth will be used to clean the chassis, the IF cans and the tube shields. We will use a dry cloth to clean vacuum tubes. A cloth with few drops of oil will be useful to remove light rust stains from transformers and other iron parts.
Now we must check that commands, tuning, volume, band selector and similar, move freely and smoothly. Turpentine-based solvents will help us to clean bushings and shafts, before we add one or two drops of oil to lubricate them. If we find locked shafts, we will never force them. Hardened grease, that remains when volatile fractions were evaporated, is very similar to tar. It is hard at room temperature, but it can be easily melted by heat. We will press the hot tip of a solder iron against the bushing and let it warm-up until the shaft starts to move. Now, while rotating the shaft back and forth, we slowly inject with a syringe droplets of turpentine-based solvent between the bushings and the shaft, until we feel that shaft rotates free of friction. Few drops of oil will complete this job.
When all commands move free, we will use a contact cleaner spray to clean contact surfaces. We add cleaning fluid in the contacts of band switches and inside the potentiometers of volume and tone control, while moving back and forth the control knobs or levers.
In this early servicing, a visual inspection will evidence damaged parts that must be replaced. We can easily identify and fix now many of the problems that could otherwise cause heavy damages later, at the power-on:
|- Wires or power cords with frayed or hardened insulation||Replace|
|- Tuning dial cord worn out or broken||Replace|
|- Vacuum tubes with gas inside, whitish or iridescent getter||Replace|
|- Tubes not properly seated in their sockets||Push them in place|
|- Tubes with loose base or cap||Use cement to lock loose parts|
|- Paper capacitors with cracked body and/or filling wax or tar dripped out||Replace with suitable types|
|- Electrolytic capacitors showing traces of dried electrolyte out of the vent hole or of the rubber sealing||Replace|
|- Burned resistors||Replace, but look before for leaky or short capacitors and other components that caused the overload|
In this step, we will add parts that are missing and repair or replace other parts visibly damaged. We will refer to the documentation available for the specific model, to decide the most suitable parts. Old carbon resistors can be replaced by metal-oxide power types. To replace paper capacitors, we will prefer polyester film types in DC circuits and film/foil or ceramic X2 types were remarkable AC components are present, as on the primary winding of the output transformer or from the AC mains to the chassis.Before unsoldering components with multiple terminals, to service or replace them, we will note the connections, the wire colors and even the wire seating. In case we need to replace several parts, we will try to end every single operation, before moving to the next.
If we have a capacitance meter on hand, we can also check for dried electrolytic capacitors in the supply section. If readings are under some 80% of their nominal value, capacitors are dried and must be replaced. If the capacitance meter is not available and we have on hand an analog ohmmeter, we can roughly evaluate the efficiency of filter capacitors by ballistic reading, comparing the pointer deflection against a good capacitor of similar value.
We use an ohmmeter to check the continuity of fuses. When a fuse is open, we look at its appearance. A broken wire inside can be the result of a simple oxidation process through the years but, when glass looks burned, we must investigate for heavy overloads or shorts somewhere. The selenium rectifier, if any, or a shorted filter capacitor are the most common causes of failure, especially if the radio was powered-on with no precautions by the previous owner. We must always replace the open fuse with a new one of the same type and value.
Using the ohmmeter, we will check the power-on switch for proper operation. Silver-alloy contacts inside are self-wiping but, when heavily oxidized, surfaces act like insulators, preventing the closing of the power circuit. A contact-cleaner spray can be used to remove the oxide layer. Often switches are fully encased inside small Bakelite bodies, attached to the volume control or to the keyboard selector, and contacts are not accessible. In this case, we will gently drill a small 1.5 mm diameter hole in the body of the switch. We will care to use a sleeve on the bit, to leave it unprotected just for two or three millimeters and avoid penetrating deep into the switch. Then we can spray a contact cleaner into the switch body through this hole, while moving the switch on and off. The hole could be sealed at the end with red silicone sealant.
To be continued...
Voltages inside a vacuum tube radio are dangerous to life. To prevent shock hazard when we must make service checks inside the set, we will always operate with power cord unplugged and waiting for capacitors inside to discharge. To check voltages inside, we will make permanent connections from the meter(s) to the set, using insulated alligator clips. We avoid touching parts inside the cabinet when our set is powered.
Now we are ready to perform a smooth power-on procedure. This procedure should always be performed for every electronic set containing electrolytic capacitors, after one year or more shelf storage. In absence of electric field and under the effect of the basic solution inside the capacitors, the dielectric aluminum oxide layer slowly dissolves. Huge leakage currents, well exceeding the rated values, can circulate when power is first applied after a prolonged storage. We know that power is the product of circulating current by voltage. At the high operating voltages of vacuum tube sets, these currents can cause heat spots, which result in further current increases and eventually in the fast destruction of the insulating layer. A pre-conditioning at reduced voltage for a short while is advisable, in order to prevent excessive thermal build up during the reforming of the oxide layer. By the way, a pre-conditioning is specified by manufacturers as Philips, even for new capacitors, after prolonged storage or storage at temperature over than 40 degrees.
A quite similar behavior can be observed in selenium rectifiers. Here, when no AC power was applied for a long while, the reverse current of rectifying contact layers increases considerably. Even in this case, the temperature increase, due to the reverse current, could trigger a regenerative process, that rapidly ends in the melting of the rectifying surfaces. However, even in this case, a pre-conditioning at reduced AC voltage will reform the rectifying layers, bringing back reverse currents to their normal values. (*3a)
To perform the soft start procedure, it is advisable to operate with the chassis outside of its cabinet, so to have easy access to components and wirings. The speaker(s) must be connected. A DC voltmeter, V, will be connected by insulated alligator clips in the power supply, across the input filter capacitor, as shown the figure below. The radio will be plugged to a 2:1 step-down transformer or to a Variac.
As safety precaution, we will connect a discharge resistor Rd across the first filter capacitor. This resistor will provide a fast discharge path for the capacitor, when we remove power. A couple of 220 kohms, 2.5 watts or more, metal oxide resistors in parallel could be fine.
Now we can power the radio. If we use a Variac, we slowly raise its output voltage up to about 50% of the line voltage. We will monitor the voltmeter across the filter capacitor C1, while paying attention to signs of overloads or overheating. Depending upon the type of the rectifier used, solid-state or vacuum, the voltmeter reading could peak to about two thirds of the nominal value and then stabilize at about one half of the same in about a minute, the time required for vacuum tubes to start emission at reduced heater voltage.
If nothing happens when power is first applied, then we should investigate with an ohmmeter around power cord, plug, fuse or voltage selector. If the pilot lamps and the tube heaters glow and the B+ voltage is zero, we will check the rectifier and the filter capacitors, as described later in section 4.
We will leave the radio operate at reduced line voltage for about 15 to 20 minutes, always monitoring for faulty conditions, as a sudden drops in the voltmeter reading, smoke or smell of overheating. This procedure will allow the safe reforming of the aluminum oxide layer in the electrolytic capacitors and of the rectifying barrier in the selenium rectifiers, if any. Even if leakage currents start quite high, they remain into safe limits and rapidly drop to their specified values. The power transformer and the selenium rectifier, if any, must remain cool.
As next step, we will raise the supply voltage at its nominal value, always monitoring the value of B+ voltage. Tolerances up to 10% on its nominal value can be acceptable but, when the reading is too low, probably there is a problem somewhere, not necessarily in the power supply section. We must investigate quickly, looking at any sign of overheating, either in the power supply itself and in other directions, the audio power amplifier being a probable cause of overload.
If nobody before had tried to power our radio and if the power-on procedure went smooth, now probably we will have the pleasure to see our radio working, with a more or less pleasant sound. Anyway, even if the radio looks to work fine, we shall immediately check the proper operating conditions of the power amplifier and of other stages, as described later. Failure to perform these checks could rapidly cause severe damages to expensive vacuum tubes and to other hard to find components.
The recommended checks to perform, in addition to those required for fixing specific malfunctions of our radio, are the following:
- AF power amplifier: check of the DC bias of the control grid(s). Refer to chapter 5.
- RF/IF stages: check of the screen grid voltages. Refer to chapter 6.
(*3a) Henney and Fahnestock, Electron Tubes in Industry, McGraw-Hill
To be continued...
Here are the most common basic circuits commonly found in the power supply section of radio sets.
Fig. 4.1 - The half-wave rectifier circuit, 4.1-A, can be found in inexpensive sets: its drawback is the need for high value filter capacitors, to keep the ripple low enough. Therefore this circuit easily gives unacceptable ripple, when electrolytic capacitors are somewhat dry. The other two simplified diagrams refer to full-wave rectifiers.
Usually the plate supply of AF power amplifier is directly derived from B+, while other tubes are fed through an additional filtering section. The smoothing filter includes then one more electrolytic capacitor, or sometimes a second section in the same can, and a wirewound resistor or, in radio sets made up to the early ‘950s, the field coil of the electro-dynamic speaker, with a DC resistance in the order of 750 ohms. Here is a simplified diagram of the complete power supply section of a typical radio set.
Fig. 4.2 – Simplified diagram of a typical power supply. In old radio with electro-dynamic speakers, the field coil was used in place of the resistor Rf.
As we can see, there are just a few components and therefore the fault finding is quite simple. Electrolytic capacitors and selenium rectifiers, if any, are the most common cause of failure, often as consequence of hasty power-on attempts by inexperienced people. The table below lists the common troubles we could encounter in the high voltage section of the power supply of fig. 4.2.
|Symptoms||Possible cause and actions|
|No B+ voltage at all, no shorts||
|No B+ voltage at all. The fuse blows, even at reduced voltage, or the dial lamps glow very very dim. If the set uses a selenium bridge rectifier, temporarily disconnect wires from its AC terminals.||
| No B+ voltage at all. The fuse blows, even at reduced voltage, or the dial lamps glow very very dim. The selenium rectifier, if any, is good.
Temporarily disconnect the wire from low voltage secondary winding to vacuum tube heaters and connect an AC voltmeter to the same winding.
|No voltage at all, low resistance across filter capacitor C1||
|Low voltage across filter capacitor C1, often with audible hum in the speaker, as result of a remarkable ripple.||
|Voltage OK on C1, low on the second filter capacitor, C2||
|Voltage OK on C1, no voltage on the second filter capacitor||
Sometimes, even because of the poor availability of some spare tubes, silicon diodes can be used, directly soldered under the socket of the rectifier. The old vacuum rectifier can be left in its socket, unless shorted. Silicon diodes are connected in parallel to the vacuum diodes and, since their direct voltage drop is under 1 volt, the vacuum rectifier is completely bypassed.
Typical forward voltage drop in a vacuum rectifier is around 20, 25 volts, at the mean current drained by a radio set: actual drop can be derived from datasheets of each tube. Then we can expect some 10% DC voltage increase when we replace vacuum rectifiers with silicon diodes. To prevent overheating and stresses, which should always be avoided in an old radio, we can add a small voltage drop resistor in series to the rectifier. A 220 or 270 ohms resistor, 4 W or more, will be fine for a load current of 70 milliamps.
Fig 4.3 – Connection of silicon diodes to bypass a vacuum full-wave rectifier. Rdrop can be omitted, or even replaced by a power zener diode of about 20 volts.
Selenium rectifiers can be replaced by silicon diodes, when damaged or even when just an occasional use of the radio is expected. In this case we have to isolate or remove selenium coated plates, to prevent flow of high reverse currents.
Fig. 4.4 – In A we see a shorted selenium rectifier. In B, the same rectifier disassembled. In C we see a couple of the fiber washers used to insulate the elements, replacing one of the original metal washer in each diode section. In D small silicon diodes have been soldered to the terminal lugs of the rectifier, that has been assembled with one fiber washer in each diode stack. Any conduction path inside the selenium rectifier is now open and the old rectifier is just used as a terminal strip, to hold the silicon diodes. A touch of red nail polish complete the job.
Few words now on the negative voltage supply section found in some radio sets, where the control grid biasing voltage of the audio power amplifier is supplied by a separate circuit. In fig. 4.5 there are the two basic circuits usually found.
Fig. 4.5 – The circuit A uses a separate rectifier to generate considerably high negative grid bias for the audio power amplifier. This circuit was used mainly in radio sets with triode power amplifiers, operating with grid bias in the order of –50 volts. Tetrodes and pentodes require lower biasing voltages, in the order of –5 to –10 volts: the bias then can be derived from the B+ voltage, without appreciable drop, as in the circuit B.
The negative voltage is essential for the proper operation of the audio output tube(s). Should this voltage drop to zero, expensive power tubes could be damaged in a few minutes. If our radio has one of these circuits inside, we will carefully check this section. The separate vacuum rectifier and the capacitor Cbias are the most critical components. Should we need to replace the capacitor, usually a low-value and quite low-voltage type, we will prefer long-life electrolytic types, or even polyester film ones. If our radio uses a vacuum rectifier in the grid bias supply, as in fig. 4.5-A, we could add a protective backup silicon diode, wired as shown by dashed lines. A fuse added in the B+ line would be another simple precaution to protect expensive tubes against the risk of considerable damages.
To be continued...
The AF amplifier contains the power amplifier, which is the most critical stage of our radio, since expensive output power tubes could be easily and rapidly damaged by overloads, often caused by improper grid biasing conditions. Usually the sound is more or less distorted in this case, but it can be difficult to determine, when the radio is powered for the first time, where any distortion comes from. Even other malfunctions of RF and IF stages, as poor sensitivity, weak signals or frequency drifts, might give more or less distorted sound. And even poor sensitivity or frequency shift may sometimes be related to improper bias of the output amplifier. The output power tubes may drain heavy currents causing an appreciable B+ voltage decrease, often combined with a remarkable ripple increase. So it happens that, while we are looking for a bad capacitor which could be the origin of a loud hum, the plate of the power output amplifier begins to color of a deadly reddish light in the nearby.
For the above reasons, we will always check the bias of the power amplifier, just after the soft power-on described in chapter 3 and before moving to other circuits.
Control grids of AF power tubes are always biased to a negative voltage with respect to cathodes, on order to operate at the proper quiescent DC plate current. The bias voltage is set by design, basically depending upon the available anode voltage, the class of operation and the output power. We can find the same tube differently biased in different sets or in different applications. In this datasheet of the EL84 we find examples of different grid biases in several operating conditions. The proper bias voltage should be read in the documentation of the specific set, when available. Else, for a common radio set, we can try to refer to the typical value usually found in the data of the specific power tube or in the documentation of a similar set. We can find either the negative grid voltage or the value of the cathode resistor: in the second case, we will multiply this figure, in ohms, by the total cathode current (plate plus screen grid), in amperes, to know the grid bias voltage. Philips, for example, suggests for the EL84 a 135 ohms cathode resistor, corresponding to about –7,3 volts G1 bias at no-signal anode current of 48 milliamps at 250 volts.
As general rule, the more positive is the anode supply voltage, the more negative must be the bias voltage on the control grid. Triode power tubes require more negative bias values than pentodes or beam tetrodes.
Fig. 5.1 - We usually find two common circuits for the negative biasing of the control grid: the first one is based upon the voltage drop across a cathode resistor on the same power tube, the second one uses a drop resistor, from the HV secondary winding of the power transformer to the chassis. Both the circuits take advantage of DC feedback: the greatest the anode current of the power amplifier, the highest the voltage drop across the resistor and then the grid bias, until an equilibrium is reached. A third way, when more negative bias voltages are needed, is the use of a separate rectifier to generate a fixed bias, as shown in the previous chapter. This case can be regarded as a variant of the second one. The dotted resistors Rleak, in both diagrams, indicate the leakage resistance of coupling capacitors Cc in the actual operating conditions.
We can measure the actual bias voltage across the source, the drop resistor Rk or Rb, and compare the reading with the value given in the documentation of our set, or with typical values found in the datasheet. When the absolute voltage value is too low, either the capacitor, Ck or Cb, is shorted or probably the emission of the output amplifier is poor and the tube must be replaced. If we find abnormally high values, then we must look for an excessive current drain, frequently derived by improper grid bias of the output amplifier.
Check of the control grid bias
Since no current should flow in normal conditions through the grid when biased for negative values, no voltage drop could take place across the grid resistor Rg. Then we should find, with a high impedance voltmeter, approximately the same values both across the bias generator, Rk or Rb, and on the control grid pin. If readings differ, we must look for the cause.
The simplest yet reliable way to check the G1 bias and to identify leaky coupling capacitors in a power amplifier is to temporarily remove the power amplifier tube from its socket. Of course, once removed the power tube and then its current drain, the voltage we could expect from the bias source generator, Rk or Rb, is quite different. However, any drop across Rg, when the tube is pulled from the socket, is due to the leakage current flowing through Cc, or its leakage path, referred to as Rleak.
In the circuit of figure 5.1A, since one side of Rg is connected to the chassis, we must measure zero even on the other side, on the grid pin. Just very few tens of millivolts can be tolerated, due to normal DC leakage through the coupling capacitor Cc. But, if we find more than some 100 mV across Rg, we can assume that the capacitor is too leaky and must be replaced. We will use a low-leakage polyester film type.
The same applies to the circuit of fig. 5.1B, but in this case, to measure the voltage across Rg, we must either use a battery-operated high-impedance multimeter or a differential voltmeter. Or even, if we have an old VTVM, we will measure at different times the voltage from each end of the resistor Rg to chassis and then calculate the difference.
When the voltage drop across Rg is within acceptable limits, we will note its value. Now we can insert the tube back in its socket and measure again the voltage across Rg. Should the drop on it increase, or in other words should grid voltage move toward less negative values, the tube itself is gassy and must be replaced. In this case in fact, the further drop can only be generated by positive ions impinging the grid.
Just few words about the measurement of leakage currents of the capacitor Cc. We must remember that the DC conduction across a capacitor is due to discharge phenomena along tiny punctures, irregularities or impurities in the dielectric layer. The value of Rleak decreases rapidly, and the leakage current increases, with the increase of the applied voltage or even of the temperature. Rleak cannot be measured with a normal ohmmeter, since little or no conduction occurs at low voltage. Even the value returned by an insulation meter could be deceptive, if its static test conditions, as the test voltage or the temperature, are considerably different from the actual in-circuit operating conditions.
To be continued...
We will check now the proper operating conditions of RF/IF stages. The diagram of a typical RF or IF amplifier is given in the figure below. Apart from the local oscillator section, we can extend these considerations even to the frequency converter.
Fig. 6.1 – Simplified diagram showing DC paths of a tuned RF or IF amplifier. Typical values are: Rg in the order of 0.5 to 1 Megaohms, Rp from 1 kohms to 10 kohms, Rs from 100 kohms to 220 kohms. Rp and Cp could be missing, the plate resonant circuit being directly connected to B+ supply.
Usually tubes used in RF/IF amplifiers are variable transconductance pentodes, their control grid bias coming from the AGC circuit. The AGC voltage from the dedicated detector may be sometimes added to a fixed negative bias, derived from the bias of the power amplifier tube, through a suitable voltage divider network. Anyway, we should check wether the voltage values in our radio are more or less within the limits given in the documentation of the specific set, or at least are similar to typical values given in the datasheet of the specific tubes. We will also check the operating voltage on the screen grid and of course on the anode of each amplifier section.
Control grid bias
We must read, using a high-impedance voltmeter, a little negative voltage, that goes more negative when tuning a station, its absolute value being larger as signals become stronger. To know the theoretical limits of the G1 voltage range, we can refer to the datasheet of the tube.. Below we find data of the 6SK7, a sharp-cutoff pentode widely used for some 20 years in RF/IF stages of radio sets.
We see that the highest value of transconductance, or the greatest gain, occurs when G1 is biased at –3 volts (row A on the sheet). Conduction and hence gain fall to zero approximately when G1 is at –8 volts (row B). Then we should expect to read grid voltage values within the above limits in our radio, when properly operating.
The most likely cause of failure in this circuit is the resistor Rg, sometimes found open. If we note poor operation of the AGC, we should also investigate the AGC diode and the fixed bias generator, when used. Anyway, in this case, we should first align IF/RF tuned circuits, since AGC operation is influenced by signal strength.
Here the possible malfunction is due to excessive leakage currents in the capacitor Cs. In this case, due to the quite high value of Rs, in the order of some hundreds kohms, we will find on G2 a voltage considerably lower than the specified one. We can make a countercheck, removing the tube from its socket and measuring the voltage drop across Rs with a high-impedance multimeter. If the drop is appreciable, in the order of 1 volt or more, the capacitor Cs must be replaced.
Here we will just check for presence of the filtered supply voltage. The decoupling network, Rp and Cp, often is missing and anyway the value of Rp is quite low, in the order of very few kohms. Even if Cp is quite leaky, the increase of voltage drop across the resistor is usually negligible and the radio works fine. Should leakage currents become too high, the resistor, usually 1/4 W, blows as a fuse. For the said reason, we will replace Cp just when we find the corresponding resistor open.
Other capacitors in AGC or in tuned circuits
In our early visual inspection, we already replaced those capacitors and other components visibly damaged. No other capacitor must be replaced, unless of malfunction of some circuits, as result of defective units. In some cases the capacitance value may be out of specs. In paper capacitors this could occasionally take place, because of variations of dielectric constant of the paper due to absorbed moisture. But not always a small change in the capacitance results in a remarkable loss of performance.
No voltage, no leakage! When the DC voltage across a capacitor is very low, as in the case of the capacitor used in the AVC circuits, even leakage currents are negligible, regardless of the reading returned by an insulation meter.
We will never attempt to unsolder and even to move from their original position those capacitors, ceramic, mica or styrene film types, used in RF/IF tuned circuits, or we destroy the alignment and even the same performances of the radio. Some tuning capacitors could have controlled temperature coefficients, factory selected to balance the opposite coefficient of the associated coils. And styrene film capacitors can be permanently damaged by even slight overheating of the leads, around 80 degrees C, during soldering or desoldering. Then we will just replace those capacitors which proved to be really faulty.
The radio is now ready to operate. Should an alignment of IF/RF tuned circuits be required, we will refer to the service notes of the actual model.