Chapter 4   Troubleshooting Techniques.

Replace indicator and dial lights. If the power indicator is burned out the operator may assume that the equipment did not come on when the power switch was activated. You should replace lights even if you have not been asked to do so. Equipment operators appreciate it and will have good feelings about the equipment and the people who service it. This is more important than you might think. If the operator believes that the service personnel are doing a good job, there will be fewer complaints and better evaluations for the service people.

Be tactful. If you find something so simple that the operator could have corrected it, do not be insulting to the operator. There is nothing to be gained and a great deal to be lost. If questioned by the operator, be as diplomatic as you can. Good relations between operating and service personnel are as important as the technical operation of the equipment itself.

Test for the presence of AC voltage. You may want to use a test lamp because it is quicker and easier than a voltmeter.

Test for the presence of DC voltage. Make sure that the power supplies are delivering the voltages and that they are getting to all of the circuit-boards in the unit.

Make sure the power supply voltages are correct and free of ripple. A sizable enough percentage of equipment failures are caused by the power supply to warrant always checking it first. As a friend Mike McCarty puts it, "The power supply is mama. If mama ain't happy, ain't nobody happy."

If a circuit used to work, it is safe to assume that it came from the manufacturer with all component values correct for proper operation. As a service technician your job is to find the defective component and replace it. Modifying the circuit to correct the fault will probably result in performance not as good as when the equipment was new. Modifications also make the job of any technicians who come after you almost impossible. Don't start changing component values and rearranging circuitry; find the defective part and replace it.

As with any rule there are exceptions. One is obvious, if you are the owner of the equipment and you want to add a feature that was not part of the original design. Even so you should print or write clearly information about the modification on a gummed mailing label and stick it inside the chassis or cabinet. You may think you will own the equipment as long as you live but you aren't going to live forever. Eventually this beloved device will wind up in the hands of a stranger who will be helped if such a note is included. Making notations in the manual is not good enough. After you have gone on to the big ham shack in the sky chances are the manual will be tossed and the equipment sold at an estate sale.

Another exception is if the owner specifically asks for a modification. He or she may ask if you could add a phono jack so the RCA 45 could be played through the radio. Or someone may ask if the radio, phonograph, or TV set, could be equipped with a headphone jack for private listening.

When the equipment doesn't belong to you and the owner just wants it to work again, modifications are right out.

Don't start turning internal adjustments. If the equipment is just a little bit off from its specifications, recalibration may be what is needed. If the equipment just doesn't work, changing calibration settings isn't going to fix it. If you have turned every adjustment you can find, it will be difficult to know when you have replaced the defective part or parts. You can never be sure if the symptoms are caused by a defective part or the fact that every adjustment has been tinkered with. Resist the urge to tinker. Adjustable resistors and capacitors don't turn themselves.

I personally know of a case where this happened. A student worker was troubleshooting a Wavetech function generator. There must have been half a dozen trimmer pots on the circuit board. The first thing he did was to turn every screw. Then he started randomly replacing components. He finally gave up saying it couldn't be fixed. When it landed on my bench the first thing I did was to perform the adjustment procedure in the manual. It came into alignment and worked just fine. Somewhere along the line he had replaced the defective part but didn't know it because it was so far out of adjustment. So let me reiterate, don't turn any adjustments. They don't turn themselves and they are not the reason the instrument stopped working.

Circuits which have never worked are a slightly different story. An example of such a circuit could be an R and D (research and development) prototype. In such a case there may or may not be a defective component in the circuit. It may not be working because someone installed a resistor of the wrong value or a diode or transistor backward. In a case such as this you are justified in modifying the circuit.

If the transformer is good no matter if the device uses tubes or transistors, check for shorted diodes and filter capacitors. If the diodes and capacitors appear to be good but fuse blowing persists try the dim bulb test. Connect a low wattage, say 25 watt, tungsten light bulb in series with the power line. The bulb may light up at full intensity. You now have current flowing in the transformer, rectifier, and filter, components. Find out what is going on in the power supply before you switch to a higher wattage bulb. Check for heating of diodes and filter capacitors. Measure voltages with an oscilloscope if possible. If you find it necessary to go to a higher wattage bulb don't go above the rated current of the device. For example if the device specifications say normal line current is 2 amperes don't go above a 250 watt bulb.

This test may reveal that a diode is going into reverse breakdown at a lower voltage than required for the supply. Contrary to popular belief a diode is not completely binary in its failure. Overheating may reduce its reverse breakdown voltage. The cause of that overheating may be an electrolytic that passed an ohmmeter test but is conducting a direct current at a much lower voltage than its normal operating voltage. Such a symptom is quite common in vintage equipment which has not been turned on for several decades. You may observe that the DC voltage across the filter capacitor is slowly increasing. If so let things run while occasionally checking the temperature of the capacitor. If the capacitor begins to heat change to a lower wattage bulb. What is happening is that the capacitor has become unformed from disuse and is being reformed. Reforming the capacitor may restore the unit to operation. If the capacitor refuses to reform to the necessary voltage it must be replaced. Even if it does reform after a few hours of dim bulb operation replacement is still recommended if at all possible.

In this procedure you check for the presence or absence of signal at a point halfway between input and output. If the signal is present, you know the trouble is in the second half. If the signal is absent, you know the trouble is in the first half. Then you split the defective half in half and check for signal. In a large system this procedure can save a lot of time over moving down the line checking each block or stage as you go.

Think of it as being similar to the old "guess the number" game. "I am thinking of a number between 1 and 1000. You make a guess and I will tell you if you are too high or too low." Your first logical guess would be 500. If I say "You're too low," your next guess would be 750, and so on. If your first guess was 1 and your next guess was 2, it would take a very long time to find the number unless I happened to be thinking of 3. If I am thinking of 998 it would take a very long time.

An example might be an AM radio. Although it may not be exactly half way, the volume control is a good place to start half-splitting. It is easy to find and test equipment is easy to connect. You are likely using a signal tracer or oscilloscope. Presence of audio means that the RF and IF sections are OK and the trouble is in the audio amplifier and speaker section. At this point the circuits are so simple it is impossible to distinguish between half-splitting and down the line testing. All but the most elaborate console radios have only two stages in the audio section. So check between them, at the plate or collector of the power output and at the speaker.

If there is no audio at the volume control check to see if the local oscillator is running. Do not connect your scope probe directly to the oscillator grid in a tube radio. Connect to a low impedance point such as the cathode in a Hartley oscillator circuit. Transistor circuits operate at a low enough impedance that connecting to the base of the oscillator will not stop the circuit from oscillating.

There may be situations where half-splitting cannot be applied. One is an audio amplifier with negative feedback. Any trouble anywhere inside the loop will propagate around the loop and the signal will be out of whack everywhere. If the feedback is AC coupled only the loop can be opened up by unsoldering a resistor and maybe a capacitor. Then normal half splitting can be applied. In large and complex systems half-splitting usually can and should be employed.

When AM radios were about the only thing to be repaired, a device known as a signal tracer was on every service bench. The device was capable of responding to the signals in the RF (radio frequency), IF (intermediate frequency) or AF (audio frequency) stages of the radio. The "readout" of the signal tracer was a loudspeaker. Signal tracers are seldom seen these days and I know of only one company which is still manufacturing them. No such device is available for television signals.

A signal tracer for radios can easily be constructed. Any simple two tube single ended audio amplifier can be a signal tracer. Even one that already exists. Just attach a shielded probe to it such as a times 1 scope probe. If you use an existing amplifier it probably doesn't have a DC blocking capacitor in its input. One must be added before the amplifier is used as a signal tracer. Don't just say "I won't touch any high voltage. Someday you will make a mistake and touch a 300 volt point and "zap", there goes your amplifier especially if it is transistorized.

For checking IF and Rf circuitry all you need to do is add a simple AM detector as shown below.

This probe can be used to test for an AM signal on the plates or collectors of IF amplifiers, mixers, and RF amplifiers. When connected to an unmodulated source of RF the detector will deliver a DC voltage approximately equal to the peak to peak voltage of the source. The probe can be used in conjunction with a high impedance DC meter to tell if the local oscillator in a radio is operating.

In the age of CD players and computerized industrial control systems, signal tracing is usually accomplished using an oscilloscope. Of course the troubleshooter must know what the signal is supposed to look like in order to know if anything is wrong. Service manuals for most modern equipment contain actual photographs of scope traces to show what the signals are supposed to look like at various points in the system.

Signal Substitution.

Signal substitution should not be confused with signal injection. In signal substitution the signal generator is connected to the input of a piece of equipment instead of the device or equipment which normally goes there. The signal generator is left connected at the input throughout the testing procedure.

For example suppose you are troubleshooting the vertical amplifier of an oscilloscope. You could use a function generator to inject signals at various points in the vertical amplifier and watch for the signal to appear on the scope's screen.

***  WARNING  ***

It is necessary that the injected signal be compatible with the stage into which it is being injected. It will be of no use to inject the output from an RF generator into an audio stage or the output from an audio generator into an RF or IF stage.

Returning to our example of an AM radio you could start by injecting a signal at the volume control. If you don't hear the signal or it is very weak even with the generator's output turned up to full the trouble is in the audio section. Once again the circuit is so simple that there is little use in half-splitting beyond this point. If the audio signal generator you have does not have a DC blocking capacitor in its output you must use one. The simplest way is to hold one lead of a capacitor in the clip lead from the generator and use the other lead as a probe. It will usually help to shorten the leads of the capacitor. Start with the generator output set low and touch the probe to the plate or collector of the output device. Turn up the generator until you hear something. In a tube radio the sound will not be very loud even with the generator turned up to full. Turn the output down a bit and touch the grid or base of the output device. The sound should be loud with a low setting of the output. Move back to the grid or base of the first audio stage. The sound should be even louder. These statements would be true if there was no trouble in the audio section. Obviously there is so the sound will not be louder in one of the cases where it is supposed to be.

If the trouble is not in the audio section you need to work with the IF and RF sections. Set your RF signal generator to the frequency of the IF which in most cases will be 455 kHz but in car radios and some older radios may be 262.5 kHz. Remember to use a DC blocking capacitor. Touch the probe to the plate or collector of the last IF amplifier. You should hear a good loud signal with the generator set to about 0.1 volt. Move the generator to the base or grid of the same stage. Touch up the frequency of the generator to be sure it is tuned to the center of the IF band. The signal should be much louder. If the AGC in the radio is very good you might not perceive much of an increase. Don't be fooled by this. Turn down the output of the generator until the signal begins to get weak. Move the probe back to the plate or collector to confirm that the stage has a lot of gain.

The next logical thing to check is to see if the local oscillator is running. One of the least equipment intensive methods is to listen to the oscillator on another AM radio. Tune the other radio to a weak station in the top half of the AM band above 1000 kHz. Tune the radio you are troubleshooting about 455 kHz below the other one and if the oscillator is operating you will hear a beat, descending tone followed by an ascending tone, as you tune across the frequency. If you are working on a pocket transistor radio the two radios must be sitting one on top of the other for the oscillator signal from one to be picked up by the other.

Another way is to look with an oscilloscope. This is the only sure test for a transistor radio. When looking at the grid in a tube radio be sure to use a times 10 probe. Even so the frequency of the oscillator will be thrown off by several kHz.

Yet another way to confirm oscillation in a tube radio is to measure the DC voltage at the oscillator grid. If you have a genuine VTVM it will have a 1 meg ohm resistor in the DC probe. You should measure approximately -20 volts at the grid. This figure is very approximate and your mileage may vary. If you measure only a fraction of a volt negative the oscillator is not running. If you only have a DMM you must use a 1 meg ohm resistor at the probe tip as described for a capacitor above. Touch the resistor to the grid. The reading will be approximately 10% low but you only want to determine if there is negative DC present at the grid. Its absolute value is not important.

If the oscillator is running but the radio is still not working you may be encountering a puzzling situation. You may be hearing stations when you touch your test probes to certain points in the radio but they go away when the probe is removed. This is most likely in a radio that has an RF amplifier stage. The trouble is there. If the radio does not have an RF stage this situation is still possible although the symptom was more likely to be that the radio would pick up very strong local stations but they sounded very weak. This would indicate trouble in the built in antenna or if the radio is very old and requires an external wire antenna the antenna transformer may be defective. You did remember to connect an external wire antenna to that old radio, didn't you?

The usual method of disturbance testing is to touch or scratch various points in the circuit with a screwdriver. It is sometimes helpful to touch a finger to the metal part of the screwdriver, but only in low voltage circuits.

If you are concerned about the hazard of electric shock (as you should always be), you can touch or scratch with one lead of a resistor while holding the other lead. The value of the resistor should be no less than 120 k ohms.

Disturbance testing should always be conducted with care because it is possible to destroy some MOSFETs (metal oxide semiconductor field effect transistors) by touching the gate terminal.

Back to the example of our radio. The impedance levels in a transistor radio are too low to effectively employ disturbance testing. You can test the audio section in a tube radio by touching the center terminal of the volume control. CAUTION! In most AC operated radios the on/off switch is shared with the volume control. The AC terminals for the on/off switch are on the back of the control. The audio terminals for the volume control are on the side. It is likely to be unhealthy to get them confused. Touch the center terminal on the side of the control and turn the volume up and down. You should hear the volume of the hum change. Note: If you aren't using a knob on the shaft but turning the bare metal shaft you will find that the hum will be considerably reduced when you touch the metal shaft. That's because you are grounding your body to the chassis of the radio and the signal that your body picks up as it acts as an antenna is not nearly as strong as when it is not grounded.

Disturbance testing on the IF amplifier is done by scratching on the grid terminal of the IF tube with the blade of a screwdriver. If this produces static in the speaker it is likely the IF amplifier is alright. And that takes us back to that pesky local oscillator again.

Static testing is further subdivided into two types, testing with power on and testing with power off.

Testing with power on usually means making DC voltage measurements. Voltage measurements are only meaningful if you know what the voltages are supposed to be. That implies that you need a voltage chart for the particular piece of equipment. If you are without such a chart, do not despair; later chapters in this book will cover how to reason from the circuit diagram what the voltages should be.

In any radio, tube or transistor, if it stops working chances are that some vital DC supply voltage is missing. While you may not be able to say if the voltage is high or low, if it is zero the tube isn't going to do much amplifying. In a tube radio check for plate and screen grid voltages. Odds are good this will lead you to the problem. In a transistor radio check to see that base emitter junctions are forward biased and that the collector is higher, in the absolute sense, than the base. Remember that most older transistor radios used PNP transistors which means that all voltages are reversed from the way we have become accustomed to thinking of them in NPN circuits.

Static testing with power off means making resistance measurements with an ohmmeter. Because the amplifying devices used today conduct current (from the ohmmeter battery) even when no power is applied (unlike vacuum tubes), resistance measurements are seldom used.

This philosophy does not work quite as well in the transistor age as it did in the vacuum tube age. One reason why this is true is because transistors have such a low input impedance. The trouble can be one stage later than you think it is. Another reason shotgunning doesn't work as well as it used to is because of the increased use of feedback in transistor circuits. This makes trouble propagate around the loop, making it very difficult to pinpoint the defective stage. Negative feedback can also defeat signal tracing and half-splitting.