Diagnostics in electronics

[Anonymous No.1632409]

Can someone explain a bit about the process of finding a problem in a circuit. Tell about how you do it, what method you use (probing, search visual signs, etc).
>Pic: I have a old casio cmd 40 that is completly functional, except for the IR transmissor (although the LED light up when voltage was applied). **Tips about where to start looking** I have fixed similar eletronics but with more obvious problems, I also have never dealt with something as delicate as a clock before so I want to make sure I dont fuck up.

[Anonymous No.1632410]

Other side just in case it helps with something

[Anonymous No.1632415]

>>1632409
if its just the led part that doesn't work then start there. follow the traces from each leg of the led and draw a circuit diagram. seems from the front one side has a resistor and on the back a transistor, so likely the transistor is driven by the cpu in some way. see if you can get a scope on the line from the cpu to transistor gate and see if that is doing anything. then work your way along, check for voltage out of the transistor, if you say the led works the only other real thing it can be is that other resistor, check with a multimeter.

[Anonymous No.1632502]

>>1632409
I help out some people who throw street parties with a big old boombox collection. 90% of the time they're broken because someone dropped it and cracked the PCB. Sometimes it's obvious and other times I need to check every pair of points with a multimeter until I find a broken trace. You can also do this to identify components that are no longer properly soldered. The other common problem is electrolytic capacitors and the larger diodes and transistors used in power supplies failing.

[Anonymous No.1632513]

>>1632502
is this what hell looks like

[Anonymous No.1633167]

>>1632409
how i was taught in school to reverse engineer/troubleshoot is ask yourself the following

1. what is it and what is it supposed to do
2. what is it not doing that it is supposed to do
3. does it have power
4 and beyond depend on what it is.

if you can find a schematic that has wave forms at test points, you can use an oscilloscope to to narrow down to where the problem is and then look up components for equations and IC's for pin outs and do a shit load of algebra.

if you don't have a schematic your going to have to make your own and do a shitload of math. rarely is it something visual, and even if its visual like a broken copper traces, you'll need a magnifying glass..

my advice is find a schematic, or at least that IC's pin out and start taking voltage measurements from the IC's pin to the LED.

i don't know your skill level, but a word of warning, you can fry components by applying voltage to random components to see if they work. also avoid continuity checks, the meter applies electricity. that's low voltage and you can probably handle the thing when powered without getting bit, voltage measurements are fine. besides you have to break traces/unsolder components to measure current.