Best way to discharge caps?

i was recently working on a guitar amp (not plugged in, or hooked up to
any power source) and got a nasty shock. I always heard putting a metal
screw driver with a non conductive handle between the two prongs on
the power chord will discharge whatever voltage that is left in the
caps. I found that this was not correct, and got a 100-200v shock from
a totallly dissconnected circuit board (ouchie!) that was off for a few
hours.. I do not plan on repeating this experience (im lucky i only got
a minor shock) a 2nd time..

Whats the best way to discharge these caps? i measured hours later and
there was still about +180v DCV being stored.. Is there a easy to do
method?

thanks

richie


--

+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Richie086
http://www.richie086.com

"The only thing better than sitting outside and
playing a banjo is sitting outside playing a banjo
made of the skulls of people that made fun of you in
elementry school."

 

That was bad advice! You need a 1000 to 10,000 ohm resistor and a set
of insulated alligator clips. Ground one clip, then connect the other
clip to the positive terminal on the capacitor. I would use a 2 watt (or
higher) resistor in case you run into a higher voltage. Leave it
connected for a minute, then check the voltage. If it is gone, or down
to a couple volts, it is safe to unhook the resistor.

--
Merry Christmas!

Take care, and God bless.
Michael A. Terrell
Central Florida

 

While the screwdriver-across-the-prongs method can be used, there are
two considerations: first, precisely which prongs? Not the power cord
prongs, but the actual "guilty party", the capacitors. The power cord
may drain SOME capacitors when shorted like that, but I wouldn't want to
be making any bets on it getting them all (or even the most dangerous
ones) Second, I feel it probably does bad things to the long-term life
of the capacitor to "spark it off" like that. I could be wrong, but...
<shrug> that's the way I feel, so that's the way I do it, and it gets
the job done just as well.

Find the biggest caps on the board. Use whatever suits you to grab onto
a large-ish wattage resistor rated around 10K (not critical - just want
to have a decent bit of resistance in the circuit, and hopefully of a
power-handling capability that won't burn up or explode in the process.
If you don't have a 10K, go with whatever you've got that's kinda
close.) and use the leads of the resistor to jump the capacitor
terminals. Hold it there for a second or two. Now do it again, just to
be sure. Be aware that depending on the size of the cap and ratings of
the resistor, the resistor may get rather toasty. Or it may stay cool.
Which is why I suggest you use something to grab the resistor, rather
than just grabbing onto it bare handed.

Repeat on any capacitors you don't trust for whatever reason, then play
on the board to your heart's content.

 

If the cap is 10,000 uF (and many are bigger) and you use a 10K
resistor, the discharge time constant will be 100 seconds, so there
will be almost no voltage reduction in a second or two. Use a 50 or
100 ohm power resistor and hold it for 10 seconds.

John

 

The OP mentioned 180 volts, so I doubt there is a 10,000 µF
capacitor.
--
Merry Christmas!

Take care, and God bless.
Michael A. Terrell
Central Florida

 

I like this method myself. I usually use a 15 ohm or maybe 100 ohm 25W
wirewound power resistor. With such low values a couple of seconds is
adequate to do the job without doing any calculations first.

But the real advantage to using such a low resistance is it will usually
make a small spark when you make initial contact. When you see the spark
you can be pretty confident you successfully discharged the capacitor,
whereas without a spark it could be there is a small insulating film or
oxide layer on your terminals and you don't make any contact at all. As
such the capacitors remain charged, and even if you try to measure the
voltage afterwards it is also still possible the meter electrodes will not
make contact due to the film/oxide layer and you may mistakenly believe they
have been drained.

Of course, the amp should have some bleed down resistors on it. If it
doesn't have any then it would probably be prudent to install some.

 

Bleeders can be nasty. With lots of stored energy, as in a big power
amp, a bleeder that has a decent time constant (say, not too many
minutes) can wind up dissipating an impractical amount of power. On
our bigger NMR gradient amps (running at up to +-180 volts, tons of
caps) we have a DISCHARGE pushbutton or rocker switch and LEDs to show
voltage on the caps. We fuse the discharge resistors in case somebody
decides to press the button while power's on!

In a lot of cases, you can use a single discharge/bleeeder resistor
from V+ to V-.

John

 

Wow i really like how i got so many responces over a 24 hr period and
not one of them was flaming me saying i was dumb for even shocking
myself in the first place! thats what i was expecting, but instead i
got nothing but helpful responces! you guys rock!

I'll try using a big wirewound resistor like mentioned.. i have a few
huge ones lying around that i pulled out of a old tv set..

and someone mentioned a spark when putting the leads onto the negitive
and positve pins. Thats exactly what happened when i first went to
meausure the 180v with my DMM.. scared the shit outta me!!!!

And for this is for all the people who replied with helpful ideas, the
cap in question is only 22uF but rated at 450v.. So i guess im lucky
that i didn't get a way bigger shock, or mabye i took 200v or so off
the caps when i accidently shocked myself



--

+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Richie086
http://www.richie086.com

"The only thing better than sitting outside and
playing a banjo is sitting outside playing a banjo
made of the skulls of people that made fun of you in
elementry school."

 

Snipped good response.

The best way is to use a resistor to ground. The industry calls
them "bleeders" or "bleed resistors" in this application.

The best resistor is calculated for a given discharge time, which
depends on charged voltage and capacitance available, and how hot the
resistor will be "allowed" to get for a given environment (discharge
stick, or part of circuit).

Our HV products bleed off inside 5 seconds IIRC. It is easy to
calculate the resistance required to bleed off a given capacitance at
a given voltage in a given time frame.

The problem is that if the capacitance is too high, the resistor
will heat a lot more than our HV modules with only nF capacitances in
their output stage.

In such cases, longer bleed times should be used, and or higher
wattage rated resistors, Which can also be figured out.

It all depends on the conditions that the bleed resistor will be
facing.

In HV circuits, bleeders are almost required by law. At least UL.

In LV circuits, they can rob power, cause circuit heating, or both,
so calculating the correct value is always good, particularly if you
are installing it to be a added circuit element, instead of merely
discharging your banks at the bench during test and run procedures.

 

You surely did reduce voltage if your shock had any duration to
speak of.

You should measure it just after a power shutoff, or even while
running.

That voltage, and the known capacitance should allow you to
determine if your resistors are going to see a lot of power during
their capacitor discharge events.

Since your capacitance is fairly low, I'd say that you won't heat
'em up to much, if they re not to low in wattage and value.