Before carrying out detailed damage tracking of a digital circuit, there are several things that must be believed first, namely:
- A new service manual is available complete with circuit diagrams, layout diagrams and specifications.
- Availability of necessary tools and test instruments and spare parts.
- Be careful with the type of logic IC used in the circuit. In particular, it is necessary to know the expected logic levels and the power supply voltage specifications.
- Avoid using large test probes to prevent short circuits during measurements.
- Do not remove or insert an IC while the power supply is active/on.
- Do not apply test signals while the power supply is turned off.
- Check the power supply voltage on the actual IC pins, not on the PCB tracks.
There are several interesting ways that can be used to help find a damaged IC, namely:
1. See and touch (with senses)
Use the eyes, nose, and hands (figure 5.32).
Figure 5.32: Look and Touch
Sometimes component damage causes discoloration or the appearance of bubbles or black spots. Also, burned components cause a distinctive odor. For example, the smell of a broken electrolytic capacitor. And short-circuited chips will feel hot or even crack on the top or side. With your fingers you can feel the hot area on the board.
2. Heat and Cool
Testing in this way is very fast and effective by heating and cooling an IC so that the cause of the damage to the circuit is immediately known. Often old components become hot after being used for a long time. Its performance decreases and finally begins to stutter and break down. If a certain area where the chip is suspected of being damaged is heated (with a hair dryer) so that the damage is really visible, and then each component is cooled with a cooling spray, then the damaged chip can be seen to function again. By alternating heating and cooling, it can be known which part is damaged quickly.
Be careful in using this technique, because heat treatment of the chip can cause stress and shorten the life of components that are still good. You only need to spray the coolant for 1-2 seconds so that the hot components can function again, and try not to spray the electrolyte capacitor because the oil liquid inside can harden so that it can change the characteristics of the capacitor.
3. Chip / IC Stacking
The characteristics of an IC that is damaged due to a broken connection (cable) in the container is that it can still operate when cold. To check this, you can do this by stacking similar ICs on the circuit, as shown in Figure 5.33 below.
Figure 5.33: IC stacking
Place a similar chip that is still good on top of the suspected bad chip. Remember, first turn off the power supply, then after the chip is properly installed, turn on the power supply. You must press the chip on top so that its pins make good contact with the pins of the chip below it.
If the damage is caused by an open connection, the chip above will react to the input data and produce the output it should.
4. Approach with Similar Chips
Very often we can localize the damage to some chips, but we have to determine again, which one is actually the culprit. If time is not pressing, replace the chip with a similar chip that is still good, then test whether the replaced chip is the cause of the damage. If it turns out not to be the chip, replace another chip. If time is pressing and some of these chips are available in your spare components and the price is not too expensive, then replace the chips so that the circuit will definitely work. If there is an opportunity, we can test the used chips from the circuit using an IC tester, to find out which ones are damaged and which ones are still good to be used again at another time.
5. Cable Measurements Up to Microvolts
If you have a meter with microvolt sensitivity and have isolated a "stuck low" problem on both chips, you can try the technique shown in Figure 5.34.
Figure 5.34: Microvolt meter to determine which circuit is short circuited to ground
Measure the voltage drop between the input of gate B pin 1 and the output of gate A pin 3. This means measuring opposite ends of the same track or piece of wire: you are interested in determining which end of the track is more negative. The end closest to a bad chip will be more negative, because a bad chip will short the track voltage to ground causing this point to be more negative than pin 3.
Some important things that cause a digital circuit to experience damage are as follows:
- Power supply overvoltage.
- Excess temperature.
- Excessive input voltage.
- Excessive voltage on the data bus.
- Overvoltage clock pulse.
The actual process of diagnosing a digital circuit error is by operating the IC gates sequentially, to compare the output results with the actual ones.
There are two ways to check:
- Dynamically: by applying test signals and checking the results using an oscilloscope with a wide bandwidth (BW). The lowest CRO bandwidth is 10 MHz, and the triggering must be good. Otherwise, some pulse information will miss its target. Testing in this way will narrow the scope of finding a fault in the system as a whole.
- Static: a gate or IC function at a time. This may turn off or slow down the clock generator system. At this stage, test equipment such as those described above can be used, namely IC test clips, logic probes and "logic pulsers".
And most importantly, if a measurement is made on the TTL IC using a multimeter, then for logic 0 it should be below 0.8 Volts and logic 1 should be above 2 Volts. So if there is a TTL IC output voltage between 0.8 Volts to 2 Volts, it means that the IC has a problem.
The fault conditions for a single gate are illustrated in Figure 5.35 (a) and (b). In (a) the output is "stuck" at 0 and the output should be logic 1. Possible faults. The transistor inside is shorted, or the +5V power line is open either inside or outside.
Figure 5.35: Possible Fault Conditions in a Single Gate
In (b) The output is "stuck" at 1, with logic 1 at the inputs, the output should be less than 0.8 V. Possible Errors: Transistor is open circuit or 0 V power path is open circuit either in or out.
In a system whose inputs are supplied by the outputs of other gates and whose outputs can drive several inputs of the control gates, as in Figure 5.36 where gate A has its output permanently "stuck" at 0. Checking that the correct inputs do not force a change of state, that is, take an input down to 0, we can assume that the fault is at gate A. But this may not be true, because a short circuit to 0V from the inputs of gates B, C, or D also leaves the output of A at 0 V.
Figure 5.36: Outputs Supplying Multiple Inputs