Safely discharging those big caps in SMPSes

[Christopher]

Measure it with a meter, most discharge within 5 minutes of turning off with a bleed resistor. If the meter shows a voltage, screwdriver (or 10R 50w panel mount resistor) in parallel!

[brabus]

AND... always operate with one hand behind your back.

And maybe a friend nearby, who can see if you need help. 

Seriously, whenever I work with something "powerful", I am never scrupulous enough. I turn on the equipment with a remote switch, to avoid ANY consequence of anything gone horribly wrong.
I always keep in mind this phrase: a 100 uF cap can easily rip away an eye, be aware.

Obviously I never discharged a cap with a screwdriver, unless I had to throw it away: short circuit is not healthy for yourself and for the cap as well... and not also for the screwdriver! 

Maybe a complete video tutorial on safety can be helpful. 

[justanothercanuck]

When I was taught to repair CRT TV's, I was also taught the same-old screwdriver technique.  "Jam the screwdriver against a leg and keep touching it until it stops popping..."  Same goes for the tubes, except the screwdriver was also connected to the internal chassis ground.  It's probably not best way to do it, but I'm still alive. 

[grumpydoc]

an isolation transformer doesn't magically make everything safe;

No, it makes one specific action safe - i.e. connecting a grounded 'scope lead to an arbitrary point in the DUT without letting the magic smoke out of the DUT or your 'scope.

It's best to think of it as a device that protects test equipment (from the specific action above) - if the operator finds himself connected between two points in the DUT with enough potential difference between them life will become painful, short or both depending on the impedance of said PD.

Also, having provided a new ground reference for the DUT it becomes possible to complete a circuit by touching a single point in the DUT if you are also grounded. Again if there is enough voltage and low enough impedance life becomes,..... well you can figure that bit out yourself 

More embarrassingly the single point of interest might be the chassis.

Eg consider an audio power amplifier with ± 100v rails. OK, connect it up to an isolation transformer and ground the -100V rail. You now have the "0V" rail 100V above ground and it's likely the chassis is connected to 0V - enough for a nasty nip. The +100V rail is now 200V above ground, enough to kill considering this is DC and the PSU will probably be capable of several amps.

[Doug_in_Minnesota]

The idea mentioned of using a lightbulb is BRILLIANT !!!

I never thought of that. Gives you a visual indicator of the charge dissipating. I'm going to put together a rig like that, with different nominal voltage bulbs, and a probe. Lot's easier than what the one guy was describing of doing formula's to determine resistance, wattage, TC's, etc.

Thanks!

[David_AVD]

The idea mentioned of using a lightbulb is BRILLIANT !!!

Haha... pun intended?

But yeah, we use the light bulb trick many times a week when dealing with audio power amplifiers.