PAGE 9

LOOKING
AT FAULTS
Part 1

INTRODUCTION
Looking at the signal through a coupling capacitor
Biasing a transistor

Maybe you have seen a circuit diagram showing the signal at various locations.
If you are are fault-finding the circuit and connect a CRO to different points, quite often you find the waveform is completely different to that on the circuit diagram.
What do you do?
Is the fault in the circuit?
Is the fault in the CRO?
Or is the picture incorrect?

This is a major dilemma so we have decided to produce this chapter: "Looking At Faults."
It will guide you through some of the simplest circuits and show you how components have an effect on a waveform.
The signals flowing through a circuit are very complex and it is very difficult to show them on a circuit diagram, due to many factors.
The amplitude will depend on the supply-voltage and this can decrease as much as 50%, especially in battery-operated equipment.
The other problem is the shape of the waveform. This will depend on the "quality" of the CRO (the maximum bandwidth) and the setting of the probe (1x or 10x). Your probe may put a load on the circuit and distort the waveform. Or your CRO may not be fast enough to reproduce the waveform.
If the observer does not see the exact-same waveshape as shown on the circuit, he starts to worry.
I know. I have many overseas circuits showing waveforms and seldom do they correspond to those on the CRO.
This causes enormous frustration.
Is the fault in the section under observation?
Is the fault being introduced by another section?
Or is the fault due to the signal being processed?
We cannot definitively answer any of these and that's why you have to use a CRO with a lot of understanding.

Because of this, we will limit our discussion to general situations of what to expect when you see a waveform and how to interpret it.
Some things, such as the waveform around a faulty diode in a power supply bridge, will appear almost perfect but all you have to do is put your finger on each diode and the cold diode is faulty!!!

The same with the reservoir electrolytic in a power supply. You can spend an hour looking at waveforms. But all you have to do is turn the supply off, connect another electrolytic across the original and switch on. Hey Presto. It's fixed!

And the regulator in a power supply. It may be a transistor or 3-terminal regulator.
Simply replace it. These items fail as soon as current starts to flow and cannot be tested out-of-circuit. They will seem perfect when no current is flowing.

Don't try to use a CRO to solve a problem until you have done at lease some of the following:  (not necessarily in this order)
1. Lightly tap the PC board.
2. Lightly twist the board.
3. Use coolant (freeze) to cool the components.
4. Replace all the items in a particular section.
5. Replace all electrolytics.
6. Feel the temperature of components with your finger.
7. Re-solder joints and connections.
8. Measure voltages (and current) with a multimeter.
9. Replace IC's - especially if they are in sockets.
10. Try all sorts of other things - like substitution, comparison with another board, re-building a section or removing various sections. 

If all this fails, you can use a CRO.

Let's start at the beginning.
There are two main waveforms. All others are made up of variations of these.
They are:
1. A gradually rising and falling waveform called a SINEWAVE.
2. A rapidly rising and falling waveform called a SQUARE WAVE.


Fig 1: A Sinewave and Squarewave

It does not matter if the waveform is above or below the X-axis. This can be moved by adjusting the left/right and up/down controls on on the CRO.
We are only concerned with the shape of the wave.


Fig 2: A wave can be anywhere on the axis. It's only
the shape of the wave, we are discussing.
 

When any waveform enters a resistor or capacitor, it emerges UNCHANGED!
This amazing fact has never been emphasized before. A single resistor or capacitor has no effect on a waveform.


Fig 3: A waveform entering and emerging from a resistor
or capacitor is unaltered!

To change any of the characteristics of a waveform, the resistor or capacitor must be connected to OTHER COMPONENTS.

If two resistors are placed in series, the arrangement is called a Voltage Divider, Buffer, or Attenuator.
The signal appearing at the join of these two components is equal to the ratio of the resistances.
The two resistor symbols in Fig 4 have been shown as large and small to make the description easy to follow.
The signal at the 0v rail will be zero (this is obvious) and as the "tap-off" point is raised towards the input, the amplitude of the signal will increase, until it is 100% at the top of the two resistors. The frequency is not affected - only the amplitude changes.
Both the input and output signals will be IN PHASE. In other words they will both rise and fall at the same time.
This type of circuit is used to reduce the amplitude of a signal.  This is shown in Fig 4. 


Fig 4: The waveform from a "voltage divider."

A square wave will also be attenuated with this circuit:


Fig 5: The waveform is attenuated.

Two capacitors can also be placed in series to produce an attenuator. This is not done very often but it has the advantage of providing isolation, such as in this TOUCH SWITCH circuit.
Note the position of the small and large capacitor to create signal attenuation. If you compare this circuit with Fig 5 above, you will notice the large and small values are in different positions. This is important to remember.


Fig 6: Waveform attenuation using capacitors

Now we come to a complex part of the discussion. By placing a resistor and capacitor in series, different output waveforms will result. The two components can be placed in two ways:


Fig 7: Two ways to connect R and C

If we apply a square-wave to each circuit in Fig 7, the following results are obtained. (The value of R and C must be selected to achieve these results - but this will not concern us at the moment.)


Fig 8: These circuit will be named in a moment.

The capacitor in Circuit A is pulled up very quickly by the action of the leading edge of the digital pulse and the value of the capacitor is selected so that it charges fully during the high-time of the pulse.
The capacitor is actually CHARGING during the FALLING part of the graph. And it is fully charged when the output of the circuit is zero.
This is not obvious from looking at the graph - if you don't know what to look for -  and that's why you have to be able to read a graph AND look at the the circuit at the same time.
This is the most important part of this discussion.
Waveforms are not easy to read and not easy to interpret.
This is just the beginning of UNDERSTANDING A WAVEFORM.
The input waveform now falls to zero and the voltage across the capacitor creates a negative pulse across the resistor. To see how this works, we will replace the capacitor with a small battery. It is identical to a small battery with the positive terminal being pushed down to the 0v rail.
The negative terminal will go below the 0v rail as shown in the animation below.
This is how the output gets a negative pulse.


Fig 9: The differentiator in operation

This is the amazing part of this circuit. It is creating a NEGATIVE PULSE!

Circuit B creates an entirely different output. The capacitor is designed to charge fully during the high-time of the pulse and it will then discharge when the pulse returns to zero.
The current though the resistor is charging the capacitor and when the pulse returns to zero, the resistor discharges the capacitor.

These two circuits have technical names, they are DIFFERENTIATOR and INTEGRATOR.


Fig 10: A differentiator and Integrator

We are not going into the operation of a differentiator or integrator. They have been included to show a negative voltage can be produced by a differentiator circuit.
This is important as many circuits are not designed as differentiators but create a negative voltage due to the arrangement of the capacitor and resistor.
This type of circuit is very important as it used in many circuits to connect one stage to the next.
The following diagram shows a capacitor with an input and output signal. This type of arrangement is often used to couple two stages and is called "inter-stage coupling."


Fig 11: The output follows the input

Inter-stage coupling is also know as "AC coupling" or "Capacitor Coupling."
The following circuit shows capacitor coupling:


Fig 12: Each stage is capacitor coupled

Capacitor coupling allows each stage to be biased so the DC voltages on one stage do not affect the next stage. Note: the voltage on the leads of each coupling capacitor is different. 
In theory, the waveform on one side of a capacitor will be the same as the other when a sinewave is applied. In practice, the shape of the output will depend on the impedance of the circuit it is connected to.
Let's take the case of a sinewave generator and look at the waveform on both side of a "coupling capacitor:"


Fig 13: The waveform at point "X"

The 6v p-p waveform generated by the sinewave generator passes thorough the capacitor and appears at point "X." The DC potential at point "X" is zero (it is connected to the 0v rail) and thus the waveform appears equally above and below this level.
If the 10k resistor in our circuit under investigation is connected to the 9v rail, the waveform peaks at 12v and dips to 6v.


Fig 14: The waveform at point "Y"

You can see the 6v p-p waveform is super-imposed on the DC level (9v) of the circuit contained within the red dotted rectangle. This is shown in Fig 14.

THE COUPLING CAPACITOR
The purpose of a coupling capacitor is to separate DC voltages between two stages and transfer the signal from one stage to the next.
The waveform on the output of the capacitor in Figs 13 and 14 has been delivered to a high impedance load.
In other words it is not delivering any energy to the load and the waveform will not attenuated.
But in a normal circuit the capacitor will be delivering energy (commonly called "a signal") to the stage that follows and this will reduce the amplitude of the signal.  


Fig 15: Typical capacitor-coupled stages

If we take the case of a typical capacitor-coupled amplifier shown in Fig 15, we see the first transistor is "self-biased" and transistors 2 and 3 are "heavily turned-on."
Now we will see what to expect from this type of biasing:

 BIASING A TRANSISTOR
Basically there are three ways to bias a transistor:
1. TOO MUCH
2. TOO LITTLE
3. JUST RIGHT

There are many reasons why a (transistor) stage is not operating correctly. It may be incorrect biasing, transistor damage, high or low ambient temperature affecting the bias-point, incorrect supply voltage or any of a number of other factors.
If a transistor is biased just right  it will process a signal correctly.
If a transistor in an arrangement shown in Fig 16 is not biased correctly, the collector voltage will be too high or too low and the positive or negative parts of a signal will not be amplified correctly.
Biasing a transistor has been covered in our Basic Electronics Course and you can read more about this in the pay section of Talking Electronics website.
In general, a transistor must be "turned-on" so that it can amplify both the positive and negative parts of a signal.
For a simple "self-biased" stage, the collector voltage should be approximately half-rail voltage, as shown in Fig 16(b).
If the voltage is below half-rail, as shown in Fig 16(b), the transistor is turned on too much.
If the voltage is above half-rail, as shown in Fig 16(c), the transistor is not turned enough.
The following diagram shows three biasing arrangements:


Fig 16: Biasing a transistor

A stage can have faulty biasing due to the gain of the transistor. If a self-bias arrangement is designed for a transistor with medium gain, and a low or high gain device is fitted, the bias-point will change. This is shown in Fig 17, below:


Fig 17: Bias-point depends on the gain of a transistor

The correct biasing of a stage is very important. You must know if a transistor is biased correctly BEFORE taking CRO readings. If the waveform is distorted you need to know if the cause is a faulty component or a biasing problem.


Fig 18: Transistor "saturated" - turned-on too much.

For the transistor in Fig 18 above, the 22k base resistor is a low value and will turn the transistor ON fully.
The positive excursion of the waveform from the capacitor will not be able to turn the transistor ON any more, and thus these portions of the signal will not be processed. The output will be distorted, as shown in the waveform above.


Fig 19: Transistor biased correctly.

The transistor in Fig 19 above is "self biased" via the 1M base resistor and the collector voltage will be approximately mid-rail. This will allow the transistor to process both the positive and negative excursions of the waveform on the base.


Fig 20: Transistor only turned ON when base above 0.7v

The transistor in Fig 20 above is not biased via a base resistor. 
The positive excursion of the waveform will not be processed until they are above 0.7v and the negative portions will not be processed AT ALL.
The output will be distorted, as shown in the waveform.
Since 0.7v represents only a very small portion of the amplitude of some waveforms, it may not be considered as creating distortion.
The arrangement is Fig 20 is a commonly-used configuration called "Class C" in which the transistor is not biased at all but is turned on via the input signal.
This arrangement has the advantage of drawing zero current when no signal is being processed and will deliver a large output power.

NEXT
When viewing the waveform on the output of a transistor stage, you need to find out where the fault is coming from. This will be covered in the next section.

 

Page 10

10-03-04