Hi everyone,
I was just thinking about how dangerous caps may potentially be!
I've been stung my a mains cap of 0.33uf/240v a couple of times (my cd player...) and its quite a nasty shock, I remember it making my arm feel funny for a few minutes!
What is a potenially lethal capacitance/voltage combination? Would this have to be a huge capacitance, or a seemingly small one? Or is it not so risky, being a quick discharge, rather than a steady input of current??
Stupid question maybe, but I'd like to know how careful I have to be, with say a 10,000uf/~50vdc capacitor.
I am aware of bleed resistors, I will add one for my cd player at some point, my DIY mains cleaner has one, that has a 10uf cap inside, and I wouldnt like to find out what that feels like going through my hand!!
I think it would be an interesting practical joke to charge up a small value mains cap and put it in someones pocket, but I like most of my friends too much to do that!
I was just thinking about how dangerous caps may potentially be!
I've been stung my a mains cap of 0.33uf/240v a couple of times (my cd player...) and its quite a nasty shock, I remember it making my arm feel funny for a few minutes!
What is a potenially lethal capacitance/voltage combination? Would this have to be a huge capacitance, or a seemingly small one? Or is it not so risky, being a quick discharge, rather than a steady input of current??
Stupid question maybe, but I'd like to know how careful I have to be, with say a 10,000uf/~50vdc capacitor.
I am aware of bleed resistors, I will add one for my cd player at some point, my DIY mains cleaner has one, that has a 10uf cap inside, and I wouldnt like to find out what that feels like going through my hand!!
I think it would be an interesting practical joke to charge up a small value mains cap and put it in someones pocket, but I like most of my friends too much to do that!
Safety should alway be a concern when working with electricity. Under the right conditions, 70V or less can be lethal. The power supply caps that we deal with in our hobby can hold enomous amounts of energy. If that energy finds a path through you to a lower potiential, it could very well be lights out for you even if the supply is not plugged in to the AC.Sch3mat1c said:
Be careful! Safety First!
Always use bleeder resistors!!! I say this from personal bad experience. The low impedance/high energy supplies so beloved in the DIY community are particularly dangerous. Alawys check voltages on caps (including coupling caps, especially in tube circuits) with a meter before poking around.It's the milliamps that'll get you, not necessarily the volts.
Given the right circumstances, 12V DC can kill you no trouble. Safety is a must and prevention is always the best course of action. As always, if you dont know what you are doing or safe practice for what you are doing, dont do it until you find out how to minimize the risks to you and those around you.
Hi Mark,Variac said:
Not being a circuit designer, I can’t give you all the details on why one value would be better than another, but being an EE, I can give you the basics.
I believe you would select a resistance value first. Not too low or it will continuously draw too much current from the PSU, and not too high or it will take too long to bleed down. I’m not sure what effects this resistance will have on the amp circuit. Someone else will have answer that one.
I would think that something between 1kohm and 5kohm would work, depending on the rail voltage. If you have a rail voltage of 25V, the 1kohm will draw a continuous 25/1000 = 25mA with the power on. If you have a 5kohm, it will draw a continuous 25/5000 = 5mA, again with the power on. Once the power is turned off the voltage will start to drop as the bleeder drains the energy from the caps. The current will be as calculated above only when the power is first turned off. After that, the current value will decrease as the voltage drops as more and more energy is pulled from the caps.
The amount of time it will take to completely discharge will depend on the amount of capacitance and the bleeder resistance value. Lets assume the resistances and voltage from above and a total (per rail) capacitance of 100kuf = 0.1F. From the capacitance and the resistance we can calculate the RC time constant by t = RC. This is the time it takes to discharge to 37% the original voltage, in this case 25V. So t = RC = 1000 X 0.1 = 100sec for the 1kohm resistor and 500sec for the 5kohm resistor.
It takes 5 time constants to fully discharge (<1% the original voltage) the capacitor, but only 3 time constant to discharge the cap to 5% the original voltage. So we will use 3 time constants. In the case of the 1kohm resistor, it will take 300sec or 5 minutes to discharge the cap and the 5kohm will take 1500sec or 25 minutes.
The only thing left is to verify the wattage of the resistor used. The 1kohm resistor draws 25mA at 25V so it will be dissipating a continuous 25 X 0.025 = 0.625W and the 5kohm would dissipate 1/5th or 0.125W. Pick your poison.
Because I am a Moderator, I can go in and edit my post after the fact. So I added the missing "k" to "100uf".
tiroth said:
Here is another point that i guess have to be emphasised.
It have been common knowledge between any sort of craftsmen, and in particular those working with electricity, that any sort of jewellery is to be taken off before starting to work.
This goes as well for when you get the powerdrill/- beltsander out!!!
Magura
It has to do with your own constitution as well (how wet are your fingers, how good is contact with earth, what kind of shoes you're wearing etc.. ) That explains why one doesn't get killed everytime one touches phase accidently. I was bussy once in my bathroom checking the phase with phase finder (screwdriver with resistor and neon lamp) I was wearing leather boot's at the time and got schock of my life when I touched phase.. I guess I didn't get killed because of resistance (few megaohms) and because I'm right handed.. (left handed people have more chance to get leathal injury.) So from than on, I always use special phase finder when checking anything.
And yes.. it's miliamps in combination with frequency that will kill you ( even your heart beat frequency has to do with it..). That's why when messing around with tube amps, keep your other hand in your pocket, and take care that capacitors are empty before you go further with modifications/repairs.
And yes.. it's miliamps in combination with frequency that will kill you ( even your heart beat frequency has to do with it..). That's why when messing around with tube amps, keep your other hand in your pocket, and take care that capacitors are empty before you go further with modifications/repairs.
Thanks for all the safety input! I think I'll be even more careful than ever now, when messing with stuff!!
Thanks for the calculations roddyama, but when you say "100uf = 0.1F" you mean 100,000uf=0.1F right? I assume your calculations are for 100,000uf, as when I made a PSU I put in 4.7k for each set of 2 caps (2,000uf) and they discharged in not many seconds. I now realise a 10watt resistor was slightly OTT though!!!
We can't wait around for long on a mains filter, what if someone were to touch the plug pins when unplugging it? I've no idea what I used in my mains filter, but I've not been stung yet. Presumably a 5k resistor would drain the charge away in next to no time, right? (10uf/240v)
Funny how people always talk about how its the current that kills you/causes fibrilation(?) to occur, I remember seeing a chart that showed even a tiny current will if the voltage is higher, so you need current *and* voltage to be lethal, sort of. Anyway, with an earthed chassis, and bleed resistors, even my crap should be safe!
I remember the lid of my VCR once gave me a little nip, any logical explanation for that happening???
Cheers,
Thanks for the calculations roddyama, but when you say "100uf = 0.1F" you mean 100,000uf=0.1F right? I assume your calculations are for 100,000uf, as when I made a PSU I put in 4.7k for each set of 2 caps (2,000uf) and they discharged in not many seconds. I now realise a 10watt resistor was slightly OTT though!!!
We can't wait around for long on a mains filter, what if someone were to touch the plug pins when unplugging it? I've no idea what I used in my mains filter, but I've not been stung yet. Presumably a 5k resistor would drain the charge away in next to no time, right? (10uf/240v)
Funny how people always talk about how its the current that kills you/causes fibrilation(?) to occur, I remember seeing a chart that showed even a tiny current will if the voltage is higher, so you need current *and* voltage to be lethal, sort of. Anyway, with an earthed chassis, and bleed resistors, even my crap should be safe!
I remember the lid of my VCR once gave me a little nip, any logical explanation for that happening???
Cheers,
SimontY said:
I remember the lid of my VCR once gave me a little nip, any logical explanation for that happening???
Cheers,
You can get that off the rca leads of just about any component that isn't earthed, if you are touching something that is. I thought my TV was faulty the first time it happened to me. I was holding the amp heat sink in one hand and picked up the rca lead from the TV in the other and noticed it was giving me a very slight tingle. I put the multimeter across the amps heat sink and the rca lead (either ground or signal) and I read about 115V AC (the local voltage is 240V). measured the current and it was less 0.1 mA. when I connected one RCA lead to the amp the other would read 0V on both signal and ground WRT the amps earth.
I think that this is some very small leakage current from the unearthed components. I checked all of mine, and every one of them ranged between about 80V and 120V on the rca leads when not plugged into the amp, but 0V when one lead was properly earthed through the rca socket on the amp.
Oh and just my 2c worth on caps, 16,000 uF at 63V is enough to blow a chunk out of the end of a screwdriver if you are foolish enough to touch a non discharged one! I did it years ago, when young and naive, learned very quickly that it wasn't a good way to discharge caps!!!
Tony.
Ah, I see! It probably was the RCA plug to my amp, you're right, I think my DVD player attacked me in the same way once too, now I now what to avoid....!wintermute said:
My 0.33uf cap makes a real snap when you short it with a screwdriver, I can only imagine what 16,000uf would be like!!!
edit: is it safe for the caps to short them?
And even from some that ARE earthed/grounded. All grounds, unfortunately, are not equal. It's a nice demonstration of the pernicious nature of ground currents and a strong argument in favor of galvanic isolation.
Now, back to bleeders. The calculation of bleeder resistor values is very straightforward. First, you consider the RC time constant of the discharge. Understand that the time constant is NOT the time it takes to discharge the cap, it's the time it takes to discharge roughly 2/3 of the voltage from the cap. So, for safety's sake, you can't consider the cap safely discharged until several time constants have elapsed.
So, we pick a value that will discharge the cap pretty well in the time it will take us to unplug the amp, unscrew the cover, and start poking around (we, of course, VERIFY with a meter that the bleeding has actually taken place!). So for an example, let's assume we want a 15 second time constant. For our victim/amp, let's assume a 50 volt rail and a 10,000 uF storage cap. Time constant for discharge is t = RC, where R is ohms and C is in farads (or R in megohms and C in microfarads). In this case, we calculate the R as about 1.5K.
Now, here's the tradeoff- the shorter the time constant, the safer the supply, but the more power that gets wasted and dissipated. For this example, using Ohm's Law, P = V^2/R, which works out to about 1.7 watts. A 5 watt resistor would be safe to use, and for dual rails, we'll burn up about 3.5 watts. Make that 7 watts if you've got separate supplies for each channel. That's not too bad, we can make up for that by increasing the spec for the power transformer by 10volt-amps or so.
You can do a similar calculation for a tube supply. As an exercise, assume a 450V rail, 300 uF of supply capacitance, and a ten second time constant. Calculate the resistor value and dissipation.
It's dealer's choice. Just know that if you choose to bleed each cap individually, that you must do the math for each cap. If you do one to the rail, you must consider the total supply and any added resistance from the L or the R in CLC and CRC type supplies. That still doesn't answer the design question of what effect this bleeder will have on the amp circuit, if any.Variac said:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.