Discharging a capacitor

[Greg Crawford]

I have the plug-in flourescent light fixtures, and it seems almost a sure thing that whenever one is unplugged, I touch the plug and get hit by the capacitor. Is making a dead short from one side of the plug to a ground an acceptable way to discharge one of those? It really gets irritating.

Thanks,

Greg

[Ken Higginbotham]

I'd get the wife to plug 'em in :D

[Greg Crawford]

Maybe that's why she left.

[Dewey Torres]

Greg,
That’s the way we do it in the Navy only we use a shorting probe that has a long non conductive handle. Once you short the capacitors collector to ground it will discharge and you will be safe. I have never seen this done on lights though.

[Paul Atkins]

Don't touch the plug. Take a voltage meter and see if that is where the juice is coming from. I haven't seen one with a capacitor - are you sure it's not static?

[Dan Friedrichs]

What?! :eek:

You should NOT be able to feel anything by touching an unplugged light fixture of any sort. While I've seen plenty of stupid designs, anyone smart enough to design a flourescent lamp balast would not do something like place a capacitor directly across the line. Was this fixture made in China, per chance? :rolleyes:

I think there is something wrong with it, and you should replace it.

That being said, a screwdriver works. If there's enough stored energy, you might start to melt chunks out of your screwdriver and/or plug, though.

[Ken Fitzgerald]

I work with high voltage daily. We have a special tool used for discharging. It has a couple of resistors in parallel that are in series with a cable with an alligator clip on it and a "metal wand". By having the resistors there, it limits the current thus suppressing instant unlimited discharge currents. You attach the alligator clip to ground and then holding the insulated handle, you touch the "metal wand" to the device being discharged.

That being said....you should not being seeing the effect that you are. Something's not right.

Do all of the lights do this or just one. Somethings not right.

[Chris Padilla]

Well, capacitance is just two conductive materials separated by a space. The space could be air or some dielectric material. This "device" can store energy and hence can also discharge the energy when the right situation strikes.

I think it could be static electricity you are experiencing but maybe not....

[Greg Peterson]

I work with high voltage daily. We have a special tool used for discharging. It has a couple of resistors in parallel that are in series with a cable with an alligator clip on it and a "metal wand". By having the resistors there, it limits the current thus suppressing instant unlimited discharge currents. You attach the alligator clip to ground and then holding the insulated handle, you touch the "metal wand" to the device being discharged.

That being said....you should not being seeing the effect that you are. Something's not right.

Do all of the lights do this or just one. Somethings not right.

+1.

When I am discharging caps on my amplifiers I use a conductor with high value resistor in series to ground. May not be as exciting as shorting out with a screw driver, but I don't really care all that much for sparks.

If you're getting a jolt off the fixture, something is definitely not kosher and discharging any surface charge that may exist is the least of your problems.

[Dan Friedrichs]

Well, capacitance is just two conductive materials separated by a space. The space could be air or some dielectric material. This "device" can store energy and hence can also discharge the energy when the right situation strikes.

I think it could be static electricity you are experiencing but maybe not....

Very true, Chris. The parallel-conductor cable running to the lamp is a capacitor. However, if you've ever taken a "can" type capacitor apart, you'll find that they manage to stuff a LOT of foil (conductive surface area) in there. Also, the insulation separating the foils is VERY thin. C=e*area/insulation_thickness, so both of those factors together is what gets you a usable amount of capacitance. Even then, many capacitors use a wet electrolyte to further increase capacity by chemical means. Thus, if the OP can touch a screwdriver across the plug terminals and see a spark, there is a sizable capacitor inside that has somehow become connected in parallel with the cord.

However, I agree that static electricity may be the culprit.

[Larry Edgerton]

I remember learning about discharging a capasitor........

I still have a twitch.......

[Gary Click]

Always liked to charge an ignition capacitor off a spark plug and leave it laying around on the bench with the wire folded back along the side. Some people never seemed to catch on or at least remember for more than a week or so.

gary

[Ian Gillis]

+1.

When I am discharging caps on my amplifiers I use a conductor with high value resistor in series to ground. May not be as exciting as shorting out with a screw driver, but I don't really care all that much for sparks.



How many ohms/kilohms are you talking? I have an old start capacitor I'd like to dispose of safely. I have the right screwdriver to do the job, but I'm not as fond of light shows as I used to be. And where do you go to buy stuff like that since Radio Shack isn't around anymore? Cheers, Ian

[Rob Young]

If you know how much voltage is across the cap, you can apply Ohms law to get the instantaneous DC current.

V = I * R

You want the current to be pretty low, 10 micro amps is a nice number. Milliamps, properly applied is enough to stop your heart. :)

Worked example:
assume cap is at 100Vdc
100V = 10uA * R
R = 100V / 100uA
R = 1Meg Ohm

The larger the resistor, the longer it will take to bleed down the cap (drops by about 68% every 2*pi*R*C seconds, all things being equal).

If you want to do more math, the equation for current flowing from a capacitor is I = C * dV/dt. That is to say, the current is proportional to the change in voltage over time. Mathematically, the current is infinite if the voltage is instantaneously brought to zero. In the real world, this isn't quite possible but shorting with a screwdriver does produce a sufficiently fast discharge so you get nice, high currents. Enough to ionize the air at the gap and viola', a spark.

Oh, and if you want to size the resistor for power dissipation, the DC power is P = I * V or P = I^2 * R. Since you are dissipating and not conducting the current continuously, this isn't quite as critical, you can figure the peak power and just use that value.

A good source for small quantities of components would be www.digikey.com (http://www.digikey.com) (minimum order $25), www.mouser.com (http://www.mouser.com), www.newark.com (http://www.newark.com) and there are lots of others.

By the way, there is an effect in capacitors where they will revert to a voltage after the short has been removed, sometimes called "bounce back". This isn't terribly critical unless you are working around ultra high voltage, oil filled jobs. It can be easily observed at the bench in smaller electrolytic and tantilum electrolyte caps but don't expect anything like 90% bounce back, think 1% or less. Still, when working with 10,000V, that leaves you with sufficient charge in a capacitor to push enough current to do damage.

As to the spark you are experiencing, if it isn't due to static electricity, there could actually be capactors across the AC line. If the balast is not using a transformer but instead was designed to run directly from the AC line, they sometimes include capacitors across the line in an attempt at power factor correction. If not done correctly, I suppose this could give you a pop each time. They should have included a tranzorb, SCR or some kind of back-to-back zener diode to snub out the caps when the power is removed but to save a penny, maybe not.

[Ian Gillis]

If you know how much voltage is across the cap, you can apply Ohms law to get the instantaneous DC current.

V = I * R

You want the current to be pretty low, 10 micro amps is a nice number. Milliamps, properly applied is enough to stop your heart. :)

Worked example:
assume cap is at 100Vdc
100V = 10uA * R
R = 100V / 100uA
R = 1Meg Ohm

The larger the resistor, the longer it will take to bleed down the cap (drops by about 68% every 2*pi*R*C seconds, all things being equal).

If you want to do more math [SNIP]

Man, that was an answer and a half!! And, personally, I couldn't possibly do any more math than that :O
I'll have to look in the the surplus shop, or risk ionizing some air. $25 minimum plus S&H is enough to make me appreciate making sparks. Never thought I'd be nostalgic over Rad Shack and Heathkit.
Thanks for the info!

[Greg Crawford]

Wow, I never thought this question would generate answers like this.

First, I have to say I assumed it to be a capacitor. It doesn't feel like any static shock I've ever gotten (I've gotten some real doozies off sandblasters), rather just like touching a hot 110 volt wire, complete with the tingling in the muscles. There's also no spark or "click" like static generates.

Does every light do the same thing? I can't say all of them for sure, but it is more than just one. I don't feel like being a human volt meter, so I'm not going to go grab all the plugs to find out.

Are they made in China? Of course, isn't everything? I have about half a dozen of these, all the cheapest you can get at the BORGs. I've tried the higher-dollar lights, and one of them gave off such a fireworks show I didn't think I'd get it unplugged before it started a fire! The other one worked for about a year and now it just goes off after about 15 minutes. It has to be unplugged for 10 minutes or so to re-light, then just does the same thing. The cheap ones at least work. I've also tried the hard-wired type, but replacing ballasts is more maintainence than replacing these, plus more expensive.

So far though, it sounds like I'm the only one experiencing this. I'll try to tag any that do give me a jolt and see if there's some common thread. If so, I'll try to determine which kind they are and avoid them in the future.

One other question, if they are holding voltage, wouldn't that light an incandescent bulb on the same outlet for a short time after the switch was off? For that matter, would it make a difference if I simply unplug the light verses turning off the circuit before unplugging it? I'll have to try to keep that in mind.

Thanks for all the brain power. Some of you really, REALLY know your stuff about electrical matters.

Greg

[Eric Gustafson]

First, I have to say I assumed it to be a capacitor.


Right! Excepting the electronic ballasts, these fixtures use an inductive ballast to generate the higher voltage needed to light up. I have never seen the circuits, but I bet there should be a discharge resistor across the inductor to bleed off voltage when you turn off the light. Perhaps the ones you have are either missing the bleed resistor or are designed improperly. I wonder if the lights you have are UL listed?

[Chris Padilla]

If you know how much voltage is across the cap, you can apply Ohms law to get the instantaneous DC current.

V = I * R

Ah, good old Circuit Theory 101. :D

[Chip Lindley]

Umm, why are we touching the plug ends anyway??? DUHH! If getting *bit* is abnormal, take the fixture back. If its broke, DO Fix it!!

[Paul Atkins]

-one of them gave off such a fireworks show I didn't think I'd get it unplugged before it started a fire
What actually are you touching when you get this jolt? Is this conduit wiring? Sounds like maybe a ground/neutral problem with your wiring. Are they plugged into 220? Use a meter and check the grounds and hot leads. Or let Mikey do it. We don't want to se you get fried.

[Rob Young]

Ah, good old Circuit Theory 101. :D

Sometimes known as Circuss Theory when things aren't working right.

Likewise, Electro-Magnetic becomes Electro-Magic... :)

[Chris Padilla]

Sometimes known as Circuss Theory when things aren't working right.

Likewise, Electro-Magnetic becomes Electro-Magic... :)

After finishing grad school in EE, my job was as an EMC (electromagnetic compliance) engineer and everyone called it Black Magic. Now I'm on the PCB getting chips talking to each other at multi-gigabit speeds (SI). :D

[Ken Garlock]

If you want some fun, take a piece of steel wool, and while holding it with insulated pliers, jam the steel wool across the offending two terminals. :cool:

In a college physics class I took a 5 mfd 5,000 volt oil filled capacitor to class. I had charged it to about 500 volts in the morning before leaving for school. The physics class was the last class of the day, when I placed the steel wool across the terminals of the cap, it sounded like a firecracker going off. Of course, some of the steel wool went up in smoke.

Ah, the good old days before boring TV lectures. :(

[R Mousel]

I seem to remember a recall on some 4' lights for this very reason. Don't remember where I saw it though. I think they were from HD but can't be sure.