Safely discharging those big caps in SMPSes

[Giddie]

I really love the videos where you've repaired LCD monitors by replacing the dodgy SMPS caps; brilliantly practical and instantly useful information, even to those of us with limited experience.  I'm generally pretty safety-paranoid, so I've done plenty of reading up on the dangers of SMPS repairs since watching those videos, but I wonder if it would be worth making a video explaining how to ensure high-voltage caps are discharged and other safety advice for those that might be tempted to reach inside their dead monitor unaware...?

[jahonen]

When I did fix similar things in the past life, I used a large-body high-power resistor (something that is easy to hold and does not get too hot, maybe 4-10 watts) around 500-1k ohms. Just put this resistor in parallel with the cap to be discharged for some time.

Regards,
Janne

[c4757p]

I'm reckless. Screwdriver short, wait 60 seconds, screwdriver short again. Anything I suspect of high dielectric absorption or very high voltage (CRT, etc) gets a ground lead clamped on and left in place while I work. Anything I suspect of a big enough spark to make me wet myself may or may not get a nearby, handy bigass resistor instead.

[Giddie]

Thanks; as I mentioned I've done plenty of reading now, so I'm not really asking how to do this myself.  I'm suggesting that maybe Dave could do a video so that people watching cap-replacement videos can be warned about the dangers and how to avoid them.  In particular, important things such as to not power an exposed SMPS without an isolation transformer.

In fact an ATX power supply repair might make for an interesting video in itself, and a good platform for this kind of information?

For anyone interested, this is a great source of information: http://www.repairfaq.org/sam/smpsfaq.htm

[c4757p]

I don't really think cap safety is enough to warrant a video. "Hi! Um, capacitors can store high voltage. Don't lick them."

[Giddie]

Yeah, true.  Just that information on its own would be a bit lonely, but for instance it took me quite a while to be confident in my calculations of the value of resistance I need to discharge a given capacitance at a given voltage in a sensible length of time, plugging everything into formulas and figuring out time constants, how much heat will need to be dissipated (so that the resistor doesn't melt), etc...  Maybe I went waaay overboard, but it could at least be an interesting lesson?  Just a thought, anyway.

Since SMPSes are one of the most common forms of failure, I think a few repair videos would be really handy, too.

[tom66]

Even with an isolation transformer I would not recommend the average individual power up an SMPS with no casing. Only if you know how to be safe around one. Testing the low voltage side, for example, is absolutely fine if you don't touch any exposed part of the SMPS.

[c4757p]

Yeah, you went overboard. If you use a resistor, discharge for at least 5 or 10 RC, then come back in a minute and repeat. Resistor power rating? "Big" - a brief overpower impulse isn't going to wreck a wirewound. I'm personally a fan of the "short with screwdriver, piss pants if >50V" method. It's served me well for years, though my underwear aren't so grateful.

[c4757p]

Even with an isolation transformer I would not recommend the average individual power up an SMPS with no casing. Only if you know how to be safe around one.

This. Would you lick a Tesla coil behind an isolation transformer? Didn't think so. They aren't the instant safety some people seem to think.

[Giddie]

OK, maybe there's a little confusion here. My intention was not to indicate what I thought was safe and ask for critique, but to suggest a possible video topic. I certainly do understand that an isolation transformer doesn't magically make everything safe; I only mentioned it because it's one of the topics that could be discussed in safety advice. Ah, never mind.

[c4757p]

  Sorry, I'm not ranting at you specifically. No offense meant. I've just heard enough retarded safety tips lately to send anyone over the edge...

[Giddie]

You're forgiven :p  I guess this is my thinking: I'm a newcomer to electronics.  I was really excited especially by the tutorials (e.g. soldering) and the repair videos on the EEVblog, and watched them several times and ended thinking "huh, I reckon I could do this, but is it really as safe as he makes it look?"  I spent quite a lot of time trying to figure out the answer to that last question, and would have really valued a basic safety briefing tutorial; SMPS and capacitor safety seems particularly important, given that pretty much everyone has come across at least a failed computer power supply or a monitor with no power, and after viewing those videos will probably (like me) be inspired to open it up to check the caps.

[c4757p]

Safety videos are hard because everyone disagrees. You're being too careful, you're about to kill yourself, back and forth endlessly. You could change a cap in the full safety gear of an industrial electrician or you could short-discharge it, or anywhere in between, and there will always be a milllion people (like me ) bitching no matter what you do. I'd like to see one if it could be done right. Personally I think the best suggestion is to not attempt to repair something unless you have at least a passing understanding of how it works. Dangerous voltages can be found in the most "unlikely" (to a newbie) places.

Case in point: I saw a kid from my school the other day messing with something power supply-ish, not giving a shit about the caps but treating the (inactive) transformer like it carried the Plague. I'll admit hoping a cap would (harmlessly) nip him and teach him the leason better than I could... If he had a clue what the parts did he would have known better.

2nd case: the Jim Williams pulse generator popular here contains 90V, a complete surprise if you didn't know it used an avalanche transistor.

[peterthenovice]

dave should do a general safety video on the basics.

[David_AVD]

We usually use a 240V lamp with clip leads (or probes) or sometimes a 5W resistor in the tens of Ohms range.

[vk6zgo]

Yeah, you went overboard. If you use a resistor, discharge for at least 5 or 10 RC, then come back in a minute and repeat. Resistor power rating? "Big" - a brief overpower impulse isn't going to wreck a wirewound. I'm personally a fan of the "short with screwdriver, piss pants if >50V" method*. It's served me well for years, though my underwear aren't so grateful.

*Pretty common method among people who do this sort of thing for a living!

Always a few fusspots who want to use resistors,though!

[MacAttak]

I agree with the OP and think some safety content is a great idea. There is a LOT to be learned as a newcomer, and yes getting nipped is a good lesson but there is a fine line between being nipped harmlessly and lying on the floor unconscious.

There are some general rules of thumb like "if you see wide traces, thick wires, heatsinks, big caps, or large coils then be careful!", but it's very much not obvious which parts of an unknown circuit are likely to be energized after the power cord has been disconnected. At least not to the novice. Yes, the leads of those big caps is fairly obvious thing to not touch - but equally important is what other components and wires are likely to be on the same net (you can get zapped just as easily by touching a solder joint on a trace inches away from an energized cap as you can by touching the cap leads directly).

Sort of a "traps for young players" topic. I'd watch it. I've been bit in the ass by 120 volt mains (spent a few summers as an electrician's apprentice in high school) - it wasn't pleasant at all and I would like to avoid repeating that experience.

[jahonen]

Discharging those caps is important if the power supply does not start at all. Then if you start desoldering components with charged caps, there might be some inconvenient collateral damage (been there, done that  ). Yes, I know that all big boys like to do the discharge with screwdriver etc. but I like it nice and smooth

Regards,
Janne

[c4757p]

There are some general rules of thumb like "if you see wide traces, thick wires, heatsinks, big caps, or large coils then be careful!", but it's very much not obvious which parts of an unknown circuit are likely to be energized after the power cord has been disconnected.

As I said:

Personally I think the best suggestion is to not attempt to repair something unless you have at least a passing understanding of how it works. Dangerous voltages can be found in the most "unlikely" (to a newbie) places.

Any electrical circuit can potentially (heh...) be dangerous. There are more applications for voltages over 50 or so than people think. It really is very hard to give anybody a general understanding of what is and isn't safe, and the danger with safety lessons like that is always that if you forget to mention something, people will take it less seriously than if you hadn't said anything.

Just don't poke around in a circuit you don't understand. I know it's tempting to see a blown cap and "know right away" that you can just whack in another one, but you really should spend a little time figuring out how the circuit works before you dig into it. Then you'll know where the dangerous parts are.

[nitro2k01]

Any electrical circuit can potentially (heh...) be dangerous.

A much more interesting question is if it's currently dangerous. (Puns, haha!)

[Giddie]

I remain unconvinced that discharging a large cap with a screwdriver should be the first port of call.  I shorted a camera flash cap with a screwdriver in my teens and even though I was expecting a spark, it gave me enough of a fright that I'd rather not repeat that.  Any situation where you're not sure what to expect (how big of a bang), and you're kind of hoping for the best doesn't seem like good practice to me.

Besides the "being in control" factor that I feel is crucial to safety, I believe that the sudden discharge caused by short-circuiting caps can damage semiconductors, and maybe even the cap & traces.  (That's certainly the warning I found in "How to Diagnose and Fix Everything Electronic", which I've found to be a very good book for beginners, by the way.)

[psycho0815]

I shorted a camera flash cap with a screwdriver in my teens and even though I was expecting a spark, it gave me enough of a fright that I'd rather not repeat that.

I shorted one of those with my thumb when i was 12 or so. Not something i can recommend. It did however teach me a valuable lesson. Just because something runs off batteries, doesn't mean it can't hurt you.
So i'm with c4757p on this one. Unless you have at least some basic understanding of what's happening in a circuit, don't go poking around in it.
And also, treat everything that might be dangerous, like it is dangerous unless proven otherwise.

[vk6zgo]

I remain unconvinced that discharging a large cap with a screwdriver should be the first port of call.  I shorted a camera flash cap with a screwdriver in my teens and even though I was expecting a spark, it gave me enough of a fright that I'd rather not repeat that.  Any situation where you're not sure what to expect (how big of a bang), and you're kind of hoping for the best doesn't seem like good practice to me.

Besides the "being in control" factor that I feel is crucial to safety, I believe that the sudden discharge caused by short-circuiting caps can damage semiconductors, and maybe even the cap & traces.  (That's certainly the warning I found in "How to Diagnose and Fix Everything Electronic", which I've found to be a very good book for beginners, by the way.)

Between 1965 & now,I've never seen such damage occur.

[AlfBaz]

I use to carry out repair and maintenance on these large 600kW VVVF drives, each phase had a bank of caps in metal enclosures that you would rack out (after waiting for the discharge time). There wasn't much room in this control room so I would place them on the floor. I had been working on the drive for about 10 minutes placed a six inch shifter on one of the cap enclosures and nearly shit my pants. Damn bleed resistor had gone open circuit. Don't know how much voltage was in it when it shorted but the DC bus operated at 600VDC 

[AlfBaz]

Oh and speaking of stored charge, heres a story I've been told by a bloke who worked at a cable repair shop.
They use to repair High voltage trailing cables and test them with high kV insulation testers. Once tested you would discharge the cable with a proper hi voltage probe.

This friend of mine was working with this guy who decided to short the charge out with a large Stanley screw driver, he tensed up briefly and fell down dead!