I come from a power supply background, where powering up an assembly for the first time could be a dicey affair. Not only is the unit at risk, but anything hooked to it is also fair game for any sort of error that causes lots of current to flow. After I saw a capacitor blow up (it was not in backwards, but another hook up error that got it), I started wearing safety goggles. If you are firing up an unknown power circuit, I suggest you wear some, and maybe a long sleeve shirt, too.
Perform a visual inspection: Use a quadrant method. Even if it has been inspected, traditional workmanship inspectors only inspect the quality of the solder joints. They miss things that are pretty obvious to an electrical engineer. Missing parts, parts across wrong pads, bad orientation - operations inspectors typically do not inspect for this. Extra parts too - Do Not Installs are a source of constant confusion to prototype assemblers. Optical inspection can pick these kinds of errors up in production, but on the first item, this is typically not set up, since there is no reference. You have to do this.
Check for improper chassis connections: If you have a heatsink or shield plane, this where a lot of workmanship errors appear. In most designs, the chassis is isolated from the electronics at the module level, and connected at the system level.
Attach one lead of a VOM to the chassis or heat sink, and the other to various points in the circuit. Check for dc continuity, should not be any. If you have EMI capacitors, you may have to wait for them to charge up. Even if you want to connect them together electrically at final, it is good to be able to isolate them for this test by designing the pwb to use a jumper or zero ohm resistor to connect them.
Sometimes, the board can be contaminated (dirty) and you will get high resistance continuity everywhere. It will drive you crazy if you are trying to characterize the performance, especially of RF circuits. When I was troubleshooting production of very low power boards, I fixed a lot of problems just having a badly functioning board cleaned.
By the way, we had some problems with cracking EMI feedthrough pins - and this method would find them. If you measure from the feedthrough pin to chassis, you should see the cap in the filter charging up. Unless it is cracked. This is common due to over torque of the fastening hardware. Cracked EMI caps are hard to find any other way. The unit will work properly in every regard, but may not a pass FCC screening test. Checked for cracked filter pin capacitors before you run final test.
Bringing the Power Up:
Estimate the expected current load on VCC.
Set VCC to the proper voltage while unloaded. Set the Current limit on the supply to minimum
Turn VCC off with power switch, connect assembly, and turn power switch on. VCC should drop due to the current limit function.
Bring the current limit up slowly until the load current matches the expected current load, or VCC comes within regulation.
If the current exceeds your estimate and the voltage does not come into regulation, you have an active circuit driving something it shouldn't. There is a problem. You can find this the way you find any short.
Could be a solder splash, or a defect in the solder mask. Could be two tracks are shorted together. An examination of the artwork will show the likely suspects, and a little trial and error gets you there. Usually a feed thru next to an underetched track or bad solder mask. "When in doubt, dyke it out" (cut the trace). You can put the right connection back in with a jumper and continue.
If the short is to a ground, voltage or shield plane, the situation is different. If VCC (typically a plane) is shorted to Gnd (also, typically a plane), it could be anywhere. Which part do you pull? Where do you drill?
Why not just chuck it an get another? Why do you want to find where it is? Because the fault is probably due to a subtle error in the way the PWB was made, or a problem with the solder mask, etc, and is likely to occur again. There is a process or drawing problem to get to the bottom of.
Sometimes shorts between ground planes be found with a very accurate volt-ohm meter. But there is a faster way. There is a tool you can buy or make that makes finding shorts between planes much easier.
You connect an AC current source (low voltage pulses, like from a signal generator, work fine) between the two points that are not supposed to be connected together. Using a very simple magnetic probe, you measure the intensity of the magnetic flux induced by the pulsing at different spots on the board. In this way you can trace current through wires, pwb traces, components, even ground planes and components. On ground planes, the current will spread across the plane, but the current density will become very high at the point of the short. The highest flux at one point is your short. Of course, you may have more than one short, if things are really sloppy.
Set the voltage and current levels so it will not hurt the components. You can measure the flux and follow the current path from one end to the other. You may have to reposition the current pulse stimulus probes several times, but it is pretty fast and painless, and will even find shorts under or within BGAs.
You can make one of these with a signal generator and a wire loop or an old cassette tape playback head attached to a scope, or you can buy a device specifically designed for this. The HP547/HP546 is what I used. Practice with a wire to calibrate your setup and test technique.
This technique is also useful in finding wiring errors in backplanes or trays.