Legal requirement; stored energy in capacitor

[Dryjoint]

I recently encountered a manufacture and test situation where a small pcb requires a 230VAC test prior to shipping.
The circuit uses a series 0.33uf cap as the dropper for a number of leds. No bleed resistor is present on the board, so a random level of charge remains when the power is removed following the test. The packers regularly act as unwitting discharge loads(!) I've searched (as best I can) the BSI site for any reference to the limits of retained energy, but without success. Any words from the wise on the legislative side of this one?

Cheers.

 

[OperaHouse]

I won't comment on the design, but a resistor accross the switched line of the test fixture would discharge the cap to a safe level. Likely what you are doing is just sticking wires into a contact hole. I can't cite the section, but not having a discharge path wouldn't pass UL.

 

[Dryjoint]

OperaHouse, I'm with you on the test rig mounted resistor; and yes - right now, they just poke the wires on it as it lays on a bench! I've recommended a covered/interlocked rig; we'll see what happens.... BTW, the discovery was made whilst I was being given an out of hours tour of the works. The MD's first reaction to being informed by an employee of the likely breach of health & safety protocol was "Well, that's because you've come from a company that has to comply." When it was pointed out that EVERYBODY has to comply, the reaction was apparently: "****ing hell, That's why I don't let ****ing visitors poke around!" Perhaps he should test a few

 

[jsummerfield]

This may not help much, but OSHA has a standard on the control of hazardous energy titled 29 CFR 1910.147. You can find this topic at

http://www.osha.gov/pls/oshaweb/owadisp.show_document?p_table=STANDARDS&p_id=9804

John

 

[IRstuff]

How much voltage is on the capacitor? Most requirements for discharge circuits ignore anything below 30 Vdc

TTFN

 

[busbar]

 
Although it sounds like it may not be a US installation, typical [NEC 460] boilerplate for premises wiring is: [ol]Capacitors shall be provided with a means of discharging stored energy…

The residual voltage of a capacitor shall be reduced to 50 volts, nominal, or less, within 1 minute after the capacitor is disconnected from the source of supply…

The discharge circuit shall be either permanently connected to the terminals of the capacitor or capacitor bank, or provided with automatic means of connecting it to the terminals of the capacitor bank on removal of voltage from the line. Manual means of switching or connecting the discharge circuit shall not be used.
[/ol]

 

[OperaHouse]

We had to test devices like this and my solution was to build an oscilator that operated at about 1khz. With a capacitive voltage drop, an increase in frequency will allow you to effectivly do the same test with only about 30V from a 1-2W amplifier chip made to self oscilate. Of course that only works if you have not much more than LEDs in the circuit.

 

[sdmays]

...we use similarly sized caps for a similar reason, and UL makes us put 1M - 10M resistors across the caps...

 

[Warpspeed]

I believe it is always good sensible design practice to include a bleeder across any large capacitor operating at hazardous voltage levels. A 1 Meg, or 10 Meg resistor does not cost much, and will have negligible effect on most circuits. The service guys will love you too.

If it is REALLY dangerous, a neon indicator or LED with suitable dropping resistor conspicuously wired right across the capacitor terminals will act as a last resort warning if all else fails. I am talking here of very high powered equipment maybe operating at several Kv and definitely lethal.

 

[ruggedscot]

Id be wary of the resistor and neon accross the capacitor - a neon opens when the voltage drops below a level - de-strickes and thats you lost the discharge path. You need a decent wattage resistor of a mohm or so just to provide a discharge path, and some form of sign to state that persons should wait so long after disconnection of supply to allow residual charge to dissipate.

Power factor correction caps are my worst fears - an open cct discharge resisotor can leave some horrendous charges present on these beasts and they can bite with some umph. Now use the last ditch method a nice big screwdriver - accross terminations and then when you do find the one with the faulty resistor it still makes you jump !

Rugged

 

[ScottI2R]

Warpspeed, YOU ARE CORRECT! I am a service guy and I always appreciate well design discharge circuits! I work with laser and EDM systems and the lasers run into the several Kv range. On some of the older (80's) systems, we have "Chicken sticks" that we use. I ALWAYS check the resistor with my meter before use! After all, 3Kv can really put an end to a bad day at work!
Ruggedscot, Just make sure the owner of the company is not in the room when you screwdriver the cap! They dont understand the loud 'POP'! Oh, BTW, also be sure that you didnt borrow the screwdriver! The chunk you take out tends to tick off the owner.
Scott

In a hundred years, it isn't going to matter anyway.

 

[Warpspeed]

ScottI2R, hehehe, I know exactly what you mean about those chicken sticks.

After you have disconnected the power and removed all the covers and are staring at the nasty looking threaded studs and big brass nuts on the top of that evil looking capacitor, it takes a very brave man indeed to just short it out with a chicken stick or screwdriver.

In my misspent youth, I once had my eyebrows copper plated doing something like that. That was in the days of high power commercial HF radio transmitters and laser flash tube power supplies. Age certainly puts fear into a man.

 

[ScottI2R]

Warpspeed,
Isnt it strange how the concept of MORTALITY tends to creep in and replace the youthfull concept of IMMORTALITY as we start to age as well as wonder,"Why isn't my 401(k) doing as well as that guy told me it would???" And BTW, NO youth is misspent, just misplaced and sometimes misguided, but all the time, MISSED!
Scott

In a hundred years, it isn't going to matter anyway.

 

[Warpspeed]

I am just glad microprocessors and logic circuits run on low voltages. Back in the old days of vacuum tubes, getting zapped or burnt fingers seemed to be an almost daily occurrence.

Ah yes retirement is sweet indeed, at least for those fortunate enough to have planned far enough ahead for it.

 

[ruggedscot]

To ScottI2R

Hey I always borrow other peoples tools - like to keep mine in their wrappers, keeps em shiny and new - The capacitor thingy Id tend to use the test lamps first and then finally the screwdriver before manhandling the beasts. One bitten twice shy - The muscle contraction tends to lift you off of your feet.

Rugged

 

[Skogsgurra]

I can sympathise with all nostalgia here and I have also had a few shocks in my life so I am not going to mock the precautionary measures that you recommend. BUT, it is all about energy. And the truth is that a .33 microfarad capacitor charged to 230 peak voltage (325 V) carries a very low energy, less than 0.02 joules.

The human body can stand a 230 V fault voltage safely during the time it takes an RCD (GFI breaker) to open the circuit. The time is around 100 milliseconds.

The RCD operates at 30 mA which gives a rough idea about human resistance; 230V/30 mA is about eight kohms, let's say 10 kohms for simplicity.

To sum up: The RC of the .33 uF capacitor and a human body (10 kohms) is 3.3 milliseconds, which is a very safe time even if the voltage were on top (325 V) when the capacitor was disconnected.

So, there should be no need for an extra discharge path - at least not for physiologic or physical reasons. But then, there are the legislators. So it may very well be that local codes say that a discharge path is needed regardless of capacitor size.

Before you all run to the drawing board to change your designs: make sure that the discharge path requirement really applies to your sub microfarad capacitors.

For nitpicks: Yes I have simplified the math, but not to hide or falsify anything. If you do it more accurately you will have an even more reassuring result.

Gunnar Englund

http://www.gke.org

 

[Warpspeed]

Fair enough, but how about a solid state Laser power supply that has 150uF charged to 3.8Kv ? Something like 1,000 Joules stored in there. The pulse forming network may be built to continuously deliver 800+ Amp pulses into the flash tube load, so the source impedance is rather low too.

No such thing as a slight tingle off something like that. I would not be checking for discharge with a screwdriver either. I know it is an extreme case, but I was once responsible for a beast exactly like that many years ago. It sure deserved some serious respect.

Defibrillators and large HF radio transmitters can kick fairly hard as well, if given the chance.

 

[ruggedscot]

I dont think that we need reminding that the implications of any discharge through the body is severe, Was always told that no matter how small a current in the right cicumstances it could kill - all it takes is for the current to disrupt the heart in some manner, you may have a weak heart a genetic defect or latent unknown. So at the very end I like to be absolutely 100% sure that there is no stray current out there.

The old screwdriver gives that last 100% confirmation that there is no power there, when working on high energy systems I always and I mean always use the check test and check again with the testers just to be sure. Being witness to a bad electrical accident really does waken you up and make you stop and think, especially when you know that it could have been avoided.

Id like to know that Id taken every step the right way the safe way and ensured that everything was right. We crap and bitch about legistlation permits method statements and risk assesments but hey there for a reason.

Rugged

 

[Skogsgurra]

Don't get me wrong! I am all for safety - that's for sure. And I have had a cardiac operation so I know about how a weak heart reacts to shocks.

BUT, the OP used a 0.33 microfarad capacitor as an example. And such a capacitor will not be a problem for anyone. I have actually designed triac controlled bath-room fans where the power for the processor was derived directly from the 230 V mains using a 0.47 microfarad capacitor.

The TCF has been checked by the Competent Body and the design has been tested for EMI, safety and fire hasard. The only thing that we had to change was that we needed to add a resistor. Yes, a resistor. But not a discharge resistor, a series PTC resistor to limit current in case the capacitors shorted out.

So, if the Competent Body - SEMKO it was - wasn't completely wrong, there is no law saying that a 0.47 microfarad capacitor that works off a 230 V mains needs to have a discharge resistor. At least not in Europe.

I might be falsely educated here - then tell me. And show me the paragraphs.

Gunnar Englund

http://www.gke.org