After successfully completing my first quite crude short circuit tracer I decided that I could make a better one: This one will memorise the best result so far and flash a LED and sound a buzzer if you better it.
It works by forcing a current through 2 Kelvin clips attached to the jack sockets and sensing the voltage. Kelvin clips are special crocodile clips with each jaw electrically seperate. The forcing current is connected to one jaw on each clip and the sensing is done on the other. The force going through the shields of two coax cables and the sense coming back along the cores. The force current is controlled by R1 to about 10mA. You can change this value: The lower the resistance the more sensitive the tester and the shorter the battery life. The sense voltage is amplified by about a thousand in IC1A and presented to the input of IC2B. The voltage on C1 is initially at V+ set by the reset switch. When the voltage on the non-inverting input of IC2B falls below the voltage on C1 the output discharges C1 through D2 to the same voltage as the non-inverting input and flashes the LED. If the probes are not detecting a short circuit closer than that stored in C1 the charge will remain in it. There is a hidden oscillator in here which keeps the LED flickering and the sounder warbling whilst on the best result so far. If you want to puzzle it out don’t read the next bit yet. The oscillator is dependent on a characteristic of a capacitor called dialectric absorption or DA. When the IC2B charges the capacitor and has reached the desired voltage it backs off releasing the force on C1. The DA in C1 forces the voltage back up and IC2B detects this giving it another jolt and so on.
IC2A is driving a guard ring, this is indicated by the dotted line, it is a circuit around the bit of the parts which are connected to C1. This is held at the same voltage as C1. The guard ring reduces the charge leakage from C1 across the printed circuit as it is at the same voltage as that on C1. If you are not using a printed circuit then the guard ring might be difficult to implement. In this case it may be omitted, IC2A is redundant and IC2B can become IC1B. The charge may not last as long if you do not have a guard ring though. All the components within the ring are very low leakage; about 5 pico amps each. With the ring in place the charge on C1 lasts a very long time making the readings last a long time and you can go away for lunch and carry on from the last reading.
D2 is a diode which has a very low leakage current, some 10,000 times less than a normal signal diode. C1 needs to be a polystyrene (PS) capacitor due to its low dielectric absorption and leakage, though Teflon and polypropylene are also pretty good they cannot get as good as PS can. D1 is a Schottky diode and is used to reduce the voltage across the probes to about 500mV when they are not connected to a short circuit. This reduces this risk of damage to the circuit under test; it can be omitted to reduce the drain on the battery if you are not going to be using any sensitive parts.
For the Kelvin clips I am using a pair purchased for about £3 each off the Internet. I am connecting them using miniture coax and jack connectors. The cable shield is connected to one jaw of each clip and the core of the cable is connected to the other. The length of the cables is not critical but may be limited by microphony.
The meters are optional and are uncalibrated; they only give an indication of your progress. That on the output of IC1A will give you a reading of the current point. That on the output of IC2A gives you a reading of the current shortest point. You could put a meter between both outputs which would give a reading of the difference.
Parts list:
IC1,2 LM6482, Chosen for its high input resistance.
Q1 any p-channel MOSFET that fits.
C1 33nF polystyrene for its low dielectric absorption and leakage.
D1 BAT85 for its low forward voltage.
D2 BAS45 has a very low reverse current.
R1 330ohms gives the force current.
R2, R3 4k7 and R4 4.7 Meg set the gain of IC1A to 1,000.
LED1 indicates a short. R5 limits the current in it.
SP1 is a cheap electronic buzzer.
S1 resets the voltage on C1 ready for the next short.
R6 increases the charge time for C1 so that the LED lights for longer.