power supplyrepairsafetytesting
I have a 750 watt ATX power supply that I will be testing with a multimeter to make sure the right voltages are present. How can I make sure that I do this safely? Is there anything that's threatening my safety that I should be aware of before I go prodding away at the leads?
Is the power that the PSU provides through its connectors enough to cause injury?
I don't believe USB power supplies require any negotiation to ask for current. When working with a PC, devices are supposed to draw no more than 100 mA and ask permission to draw more (and in practice they usually just take what they need up to 500 mA). I only mention that because you can't expect this same test to work with your computer.
The test you're considering should not damage the power supply. On a cautionary note, the 5W in a 10W-rated resistor will generate plenty of heat over a small area, more than enough to cause skin burns and start fires. Left alone for several minutes, even in free air, it will easily exceed 100ºC. There's a reason power resistors are usually made of ceramic materials. Take some precautions. Work on a metal surface, wear gloves, use a small fan or attach a heat sink, and you should be fine.
22 AWG can tolerate 8-13A depending on the insulation. Another StackExchange question indicates that breadboards have about a 1A current limit, so I refer you to that question for alternatives. (The voltage in breadboard does not matter as long as it's low.)
Bad idea.
The TNY268P along with the optocoupler is the heart of the regulation circuit. What makes the circuit 'green' is that minimizes the amount of power taken by constantly monitoring the load current. These supplies are about 78% efficient. Removing the heart is a sure way of killing the patient.
The left hand side of the circuit board (top picture) looks like a 240V mains AC input that is rectified by D1 to D4 then smoothed by the two big caps to produce a high voltage DC supply.
The TNY268P uses this high DC voltage to switch the primary of the transformer (GVG-TNY268M-A) on and off. Pin 5 - [DRAIN] to transformer and pins 2 & 3 [SOURCE] to ground (mains side of circuit)
The optocoupler monitors the voltage from the output side transformer and sends a control signal to pin 4 [EN/UV]
To the right of the transformer it looks like a bunch of schottky diodes, zeners, smoothing caps and inductors. They will determine the actual output voltages at the socket.
If there is something stopping the output my guess would be there's some form of link required in this socket (probably something that went off to the on/off switch).
More info:
I've found the whole circuit - just google tny268_122 and download the pdf
Best Answer
Many novices seem to want to measure the current output capability of power supplies by setting their meter to read current, then connecting it directly across the power supply terminals. YOU MUST NEVER CONNECT AN AMMETER DIRECTLY ACROSS A POWER SUPPLY Doing so will cause very large currents to flow, and may permanently damage the meter, power supply, or user!
If you keep the meter in voltage mode, and are careful to not short the test leads or power supply wiring, you should be able to measure the supply voltages without any hazard.
Many PC power supplies (and other switch-mode supplies) require a minimum load to operate correctly. Be sure you have a suitable load connected during your tests (or any time the supply is on.)