Capacitor Discharge Unit

[pokus]

Hey there,
I got a short, not really pedal related question.
I've seen what I believe was a built in capacitor discharge unit right next to the wall socket. It was just a light bulb and a switch.
Not for pedals but for tube amps or any other high voltage devices. Guess it's not hard to say how it's done. But it isn't as simple as just switch the two bulb connections instead of neutral and phase, is it?

[Ben N]

A light bulb current limiter, perhaps?

Edit: Read Phoenix's post below before following this advice.
As for capacitor discharge, you don't need anything that fancy. Stick a 1k 1W or more resistor in line between a pair leads ending in alligator clips, using shrink tubing as necessary to make sure you have no exposed lead. With the amp plugged in but off, clip one clip to the amp chassis and the other to the (+) end of each filter cap in turn, leaving it at each node for 20 seconds or so. Check the caps with a multimeter to make sure they are drained. Done.

[antonis]

More fancy should be a rotary switch with various resistors for various RC time constants..

[stallik]

I used to use a 10w 1k wirewound resistor, insulated as described to discharge banks of caps in studio flash gear. 10w made for a bigger handle as it was used as a hand tool. Never thought much about the process till a workmate mislaid his and used a screwdriver shaft for the same job. Molten metal flew everywhere

I doubled  up the insulation and used a rubber glove after that. The workmate received an award certificate -P45

[bluebunny]

I'm impressed that Antonis knows what a P45 is.   

[antonis]

I'm impressed that Antonis knows what a P45 is.   

You're serious salary tax spammer, Marc..

[pokus]

Thanks guys.
Yeah I guess it was more likely a current limiter to show if there are some shorts.
As I've seen capacitors discharge through a light bulb, I was just curious about how this can be done with a more or less simple switch for all of them at once. But the solutions you described are probably more useful.

[Phoenix]

A light bulb current limiter, perhaps?

As for capacitor discharge, you don't need anything that fancy. Stick a 1k 1W resistor in line between a pair leads ending in alligator clips, using shrink tubing as necessary to make sure you have no exposed lead. With the amp plugged in but off, clip one clip to the amp chassis and the other to the (+) end of each filter cap in turn, leaving it at each nde for 20 seconds or so. Check the caps with a multimeter to make sure they are drained. Done.

Sorry Ben, but that's dangerous advice. The amp does not need to be plugged in, that only risks shock or electrocution if one accidentally knocks any exposed primary circuitry. The capacitors and resistor form their own circuit, the amp does not need to be plugged in to complete it.

Also, tube amps commonly have negative bias supplies, connecting a discharge resistor between the bias filter cap positive terminal and the chassis will not discharge it, you must use the negative terminal to chassis to discharge these negative supply cap(s).
While these supplies are lower voltage than B+, and use lower value capacitors as they're lower current, they can still give you a jolt, and it is much more common for people to be injured or die after tripping over and knocking their head or similar rather than the shock itself, the actual magnitude of the shock is not a huge determinant of injury.

Checking with a multimeter is good advice though, especially for older amps with older caps which are more likely to exhibit dielectric absorption.
If you happen to have a multimeter with a LoZ volts range (as is common on electricians meters for avoiding ghost voltages), this can be done in one step, using the meters impedance (usually ~10k for LoZ) to discharge the caps, monitoring the voltage at the same time.
If your meter doesn't have this feature, you can achieve the same by putting your discharge resistor in a plastic enclosure with two pairs of banana sockets, so you can attach the leads for discharging to one pair, and you multimeter to the other.

[Ben N]

Thanks for the correction, Phoenix  . I was always under the impression that the chassis needs to be grounded for this to work properly; if that's not the case, all the better. And you are right that I neglected to mention the negative bias cap. In truth, I have generally worked on silverface Fenders, where the either the focus is on the big "doghouse" caps, or there is no negative bias supply (Champs). But you're right, the better practice would be to drain that one, too, and that is obviously from the negative lead of the cap.