Type of switch for short circuit test?

[treez]

Hello,

I wish to do a sudden short circuit output test on a LED driver which has Vout = 40V, and Iout = 125mA. (the output cap is 15uF ceramic)

There will obviously be a sudden surge from the cap when the shorting switch is suddenly switched on.........this could arc over the switch contacts and damage the switch...................what type of switch will i need to use....eg current rating and material?

Will it need to be gold plated contacts, or just tin?

After the surge, the switch will be carrying 3A rms due to the staircasing of the sepic converter inductor current into the shorted output.

 

[mtwieg]

Well a good low Rds mosfet would probably work fine... but if you want a mechanical switch then for those ratings gold plating will be useless. Are you looking for an electrically controlled switch (relay) or a manual switch (toggle, momentary)?

 

[treez]

seeking a manual switch...not momentary

 

[dick_freebird]

You will do contact damage on opening, not closing, the
switch. A snubber network may limit the voltage rise on
opening, to below arc voltage and minimize that.

Personally I'd go with the power MOSFET because you
may later take an interest in the dynamics of short
circuit protection (esp. looking at what the output
response to a momentary short is - overshoot being
more widely destructive than the short-event itself,
perhaps, and the loop being possibly wound up to the
point that normal stability analysis does not predict
the behavior).

 

[treez]

the thing is it bounces when it closes, so it will be opening and closing there-such

 

[emontllo]

How many short-circuit cycles are you planning to do? 100 or 100000?
A properly oversized switch may do the job perfectly with no damage for say... <1000 cycles, 40 V 3 A is not too much.
Just as examples that will work: Arcolectric C3950BBAAA, Carling DK284-73.
If this is a serial test system (i.e. lots of shortcircuit cycles) I would suggest using a MOSFET or SSR.

 

[emontllo]

Just a few calculations to understand better what will happen during the shortcircuit:
The power dissipated during the initial capacitor discharge will be: (1/2)*C*V^2 = 12 mJ. Considering that Rswitch >> (Rwires + ESR of ceramic capacitor), the energy dissipated in the switch will be approx. 12 mJ... not too much. After that, if short-circuit current is 3 A which should be lower than the switch rated current.