Correctly Testing Electronics

[Tardis]

It's been a long time since I dug into an extensive electronics project. I remember most of the basics of circuit and component testing, but there's a lot of stuff I have forgotten. I have a few questions...
Resistors:
Can a resistor be properly tested while in a circuit?
If it is tested in a circuit will it give you the correct reading?
Diodes:
Same question, can it be tested in a circuit?
What reading should you get from a good diode?
Capacitors:
What's the best way to discharge and test a capacitor?
Transistors:
Same question, can it be tested in a circuit?
What reading should you get from a good transistor?

After I get answers to these questions I'll have more....

[haaplaj]

Well, anything that is connected to circuit, cant be that
properly tested.

I think only diodes can be tested in the board,
diode should conduct only to one direction and
usually it's usage is that on most circuits.

Use analog ohm meter on diode with low ohm reading,
or digital meter, with the diode range.

Transistors can be tested as diodes, either PNP or NPN
is an two diode circuit. FETS and MOSFETS cant be tested that way.
However, bigger transistors can have inner bypass
diodes or resistors. You should check datasheets of those by model.

If you want to check an capacitor, in most cases you have to
take it off from board to get capacity readings.

You can test capacitor overall with analog multimeter
with low ohm ratio, by connecting leads to it and then
reversing the leads from capacitor connections.
The meter needle should jump, as it first
charged it and then re-charged it.
This usually can only reveal shortcircuited or dry
electrolyte capacitor.

Do not use this on any heavy loaded big capacitors
that may still have charge.

You can safely discharge a capacitor by
short circuiting it with resistor having low
ohm reading (About 22 - 100 ohm).

[nathanscribe]

If you want to check an capacitor, in most cases you have to
take it off from board to get capacity readings.

You can test capacitor overall with analog multimeter
with low ohm ratio, by connecting leads to it and then
reversing the leads from capacitor connections.
The meter needle should jump, as it first
charged it and then re-charged it.
This usually can only reveal shortcircuited or dry
electrolyte capacitor.

Ah, so that's how you do it... Would this work with a cheapo digital meter too, or is the response too slow?

[MrHope]

Be extremely careful with power supply capacitors. Those can melt the tip of a screwdriver in some cases and could hurt you. I don't know what the proper way is for discharging those, so ask a professional. But they can be dangerous.

[haaplaj]

Yes as I mentioned:

Do not use this on any heavy loaded big capacitors
that may still have charge.

You can safely discharge a capacitor by
short circuiting it with resistor having low
ohm reading (About 22 - 100 ohm).

I think, analog meter is essential.
You could use a light bulp (Used in cars)
with resistor in series to see is there charge.

[Tardis]

I think, analog meter is essential.
You could use a light bulp (Used in cars)
with resistor in series to see is there charge.

A headlight? What kind of bulb are you referring to?

[haaplaj]

Sorry, I ment the headlight.

You could use an flaslight bulp also with
the resistor having more ohms.

[Anaki_Muon]

In terms of the car light-bulb trick... the first project I made was a box that plugged in to a wall outlet, had a light-bulb fixture in it and had a plug on the other side. It works just like a breaker, but if I got a short in any project I plug in to the box, and the bulb is the same voltage as the project, the bulb will burn brightly, otherwise it will be dull (it's on the return circuit from the plug for the project to plug in to).

Wow. That's a garbled description. Should I rephrase that or does somebody know what I mean?

[Maschinengeist]

I'm tired and don't feel too much to write:

http://www.cirris.com/testing/guidelines/testcomps.htm

That should help you out a bit