Scope blow, why?

[Chriss]

What do you mean with this missing return gnd?

[Armadillo]

See attached.
The current must return back to the capacitor or the bridge.
The schematic was wrong or I don't know if that was the actual wirings.
Without the return connection, it could answer your original question regarding the negative probe clip.

Edit: The Germany already answered, less than 30mA, the RCD will not tripped [apprehended the missing return line].   

[Chriss]

Yea, that's correct.

I think I shorted out the BR through my GND of my scope probe through the PC USB port, where the GND is connected to earth.
The delayed big bang was just a coincidental situation because of a bad contact of the alligator clip.

[Armadillo]

I am glad that you found the answer to your question.   

[ArthurDent]

There seems to be confusion or misunderstanding about how A.C. works and is referenced to ‘N’ and what function the diode bridge performs.  First, the ‘L’ is 220VAC rms referenced to ‘N’ , earth, ground or whatever you want to call it. If you connect a scope probe ground clip to ‘N’ and the tip to ‘L’ you will see a sinewave that is symmetrical around the ‘N’ or zero reference so you will see the voltage swing smoothly from zero to about +311 peak and then drop through zero to -311 peak. The effective heating or rms value of this waveform is the rated 220VAC.

The diode bridge isn’t magical but can be thought of as a voltage controlled very fast double-pole double throw reversing switch. A diode will conduct in the forward-biased direction and not conduct when it is reverse-biased. If you draw a bridge rectifier as 4 separate diodes, one pair of opposite diodes will be ‘on’ during the positive half of the cycle and the other opposite pair will be on during the negative half of the cycle. What a full wave bridge does as opposed to a single diode is increase the efficiency of the rectifier by using the energy from both halves of the cycle and make filtering the pulsating D.C. easier because the ripple frequency is doubled.

Armadillo should look at the first schematic posted by Chriss where the ground connection is not missing and I'm assuming that is the way Chriss wired it. Ian M. is correct in that I wasn’t concise in my drawing as I only meant to show that the negative on the capacitor and the 555 were connected through the diodes in the bridge to the 220VAC line, not that there would be 220VAC at that point. It’s immaterial because there is a short of the L line through the 555 circuit and the scope with no fuse in the line for protection. The attached photo should make it clearer. 

[Ian.M]

Actually, the bottom right circuit with scope ground connected to neutral has a high risk of blowing your scope.  You'll probably get away with it in a single dwelling (with no wiring faults - are you feeling lucky?) in the USA, because for the usual US domestic and small business supply the NEC specifies that Neutral is to be bonded to Ground at the main panel, and the US split phase 115V system typically has relatively low neutral currents, but for countries that don't use split phase supplies, heavy loads can result in enough Ground-Neutral differential to burn tracks off boards etc.

[Armadillo]

Yeap, unfortunately there are misunderstandings.

Let's start from the beginning the original OP Questions;

why all the stuff not blow up when I connected the GND clip from my scope probe but when I connected the positive tip to the NE555.

Which implies he knows that connecting the Gnd Clip to the rectified DC ground will create a big bang!. He is not asking why it bangs but why it did not bang!.

Ok let me explain why I said your concept was wrong. Firstly, it was not the question asked. Secondly, the rectified DC voltage at the capacitor negative rail is at the lowest of the lowest voltage [don't agree?]. So the concept that another -220v will flow to the negative rail of the capacitor is a wrong concept leading to some people thinking AC voltage more than DC voltage, a confused concept.
However, it is a fact that the rectified "DC" voltage at the negative capacitor rail/circuit ground is lower than the Neutral line. So regarding whether it will bang when the neutral ground lead of the probe touches the rectified DC ground, I will refer you to the underlined and highlighted text above, because no one had denied it.

Cheers, let put the matter to rest. thanks

[Ian.M]

Without a photo of both sides of the fax board with connection points marked we can only speculate.  One possibility is that the O.P. was using a x1 probe, got the ground clip on a reasonably safe secondary side 0V point, then probed a point on the primary side, grossly exceeding the scope's input voltage rating, which blew right through its MCU and took out the host USB interface.

[Chriss]

I asked a friend of mine to send me a schematic about the section of the psu where the problem was.
He draw it by hand and I'm draw colored lines on it, which should represent my thinking where the problem was.

Check the picture.

I assume:
Scene1.:

The whole circuit is not galvanic separated from the main,so when I put my probe to measure the NE555 that was the same as I would measure a BR direct with my scope, putting the alligator clip to the DC gnd and the DC positive of the BR.
That would finish up in a big bang if the alligator clip of my scope is connected to earth.
The green line on my drawing is the path where the current was travel on every half cycle of the L line, also on the N line.
That means from my opinion does I shorted out the D1 and than the D2 or vise verse through the earth, and so I was connected directly to the main.
Al these happend even if I connected the scope to the GND of the NE555.

The pcb is from the upper side filled very hard with some white rubber think, so it is very hard to identify the component labeling etc.
The bottom side is clear.

Scene2.:
Maybe an over voltage was traveled through my scope tip.
But, I don't really think so.
The max voltage of the NE555 should be not more than 18V.
My tip should handel much more voltage around 600V DC.
So, that should not be the case, only if the NE555 was bad and on his output was more voltage then suspected.
But, in that case, I assume the NE555 would be burned, I mean exploded.

Conclusion:
I stand on the theory does I made a mistake and shorted out one of the diodes in the RB through earth connection of my PC.
The reason why the big bang didn't happened immediately  after I put the alligator clip to the minus connection of the capacitor
and powered on the psu, was because a bad connection on the alligator clip and the minus pole of the capacitor.

After moving the probe to connect the tip to the NE555 output i finaly made a nice connection on the alligator clip and the capacitor,
but at that point I was very close to the NE555 output pin, even maybe I touched or not, I don't know because the big bang begin. 



So, what you think about my theory?
From my side, it is a simple and common mistake when it's comes to measuring with scope on mains.

[capt bullshot]

So, what you think about my theory?
From my side, it is a simple and common mistake when it's comes to measuring with scope on mains.

Looks plausible to me.

[Ian.M]

So, what you think about my theory?
From my side, it is a simple and common mistake when it's comes to measuring with scope on mains.

.... and to me.   

However, because its so simple and common, you need to give some thought to methods of avoiding repeating it!   

If you are going to do further experiments with mains,  you need CAT II or better isolated differential probe, which takes most of the worry out of measurements in non-isolated or on the primary side of PSUs, as long as you do all connecting/disconnecting with the D.U.T powered down and stand clear before applying power.   They aren't cheap, but are far far cheaper than your funeral!

It is possible to get by with an isolating transformer, but that has several safety disadvantages that are too complex to go into here - there have been *many* threads on the subject and I doubt this forum needs another one! 

Whatever you do you need an effective and reliable method of deciding if its safe to connect your scope and non-isolated probe to a circuit.   The problem here is isolated (floating) mains powered circuits, where the permissible AC leakage current through the class Y HF return cap between primary and secondary sides of a typical SMPSU and simple inter-winding capacitance in linear PSUs, tends to leave the output floating at some fraction of the mains supply voltage when measured with respect to ground on a high impedance voltmeter like a modern DMM. 

The cure for that is to connect four 4.7K 3W wirewound resistors in series, and connect them across your DMM input.  They'll swamp the leakage current and drag the floating PSU down to ground, but if you inadvertently probe a high voltage node, they'll pass a safe 12mA or so if you are measuring mains, or typically 18mA if you are measuring the high voltage DC bus in a SMPSU.   They are also good for discharging SMPSU reservoir capacitors in a controlled fashion.    Its worth getting a small project box and butchering a set of economy DMM leads to add such a shunt resistance across the input.  You should check the resistance is close to 18.8K before and after use.

Test between the point you want to clip the scope ground to and a good ground point (not the scope itself if its a USB one without its own mains lead and ground wire).  If you see more than a few volts AC or DC, *STOP*, do *NOT* probe that circuit with the scope.  If you get a few volts AC or DC, connect a torch bulb of a similar voltage between the point you want to clip to, and ground - if it glows at all there is a real voltage differential with enough current available to be a problem, so again STOP.   Also, without the 4x 4.7K shunt, (switch back to ordinary leads), you should check that the point you want to probe is within the input voltage ratings of your scope and probe, testing between there and the point you want to clip ground to.

Hopefully you'll avoid future scope destruction and 'ElectroBoom' style excitement on the test bench

[Chriss]

Ian.M:
Can you give some drawing how to actually use that four 4.7k/3w resistors?
I'm not really understand your advice about.

What do you think if I would connect a bulb in series with the unit on the main side?
If I get a short in the circuit the bulb will go on and consume the current but nothing should be killed.
I do this when I repair a device and I'm not sure does there some other shorted parts too in the circuit.

You know what I mean... 

[Ian.M]

A bulb in series reduces the peak current if you short something, but as the cold resistance of a tungsten filament is typically an order of magnitude less than the hot resistance, a 40W 220V bulb can still let through a surge of nearly 2A, which can easily blow sensitive parts.   Higher wattage bulbs make the problem worse.  Also if the D.U.T. is starved for power, waveforms may be distorted making many scope measurements pointless.

I've attached the meter + resistor hookup.   Why so many big resistors? - well that will let them survive gross overloads.  Most wirewound resistors can handle a x5 overload for at least one second, so the resistor values and types chosen will survive brief connection to a 1000V source, and as there will only be 250V across each resistor they wont flashover and create a further hazard.

[Armadillo]

They bridge the Y capacitor there for EMC compliance and this leakage current is not fixed value, so I am wondering by altering the impedance of the meter will results in how many degrees of measurement accuracy of the meter right there left?
Any forms of documented papers there?

[Ian.M]

You aren't concerned about accuracy, you just want to be able to tell the difference between Y cap leakage and either faulty insulation, or one side of a direct from mains bridge rectifier.   e.g. for any double-insulated (Class II) D.U.T. the indicated voltage will be under 4.7V if its due to leakage, as the safety limit is 0.25mA.   The only thing that could be ambiguous is a direct short to Neutral, and the torch bulb test will indicate if there is currently an excessive Neutral-Ground differential.   N.B. Check the torch bulb before and after use!

There's a chance this testing method will trip an up-stream GFCI (RCD), but that also tells you the proposed ground clip connection point is unsafe.

[Armadillo]

With that un-toleranced resistors selection, I should figure.
At a max of 4.7volts and weighing between faults situations and this code compliance small leakage current.... ehm... oh yeah....some cannot differentiate and would definitely need it.
thanks for sharing.   

[Chriss]

If you see more than a few volts AC or DC, *STOP*, do *NOT* probe that circuit with the scope.

How many voltage is that few volts?
I assume 3-4V or maybe more voltage could be a trigger not to scope..

If I understand your idea of your drawing:
1. the resistors should be connected in series with the GND of the unit where I should connect the GND alligator clip from my scope.
2. the scope alligator clip should be connected then to the resistors in series too.
3. with my dmm I should check for AC voltage
4. if few voltage is present I should stop scoping.

Ok, but at the moment of the voltage checking with the dmm, the alligator clip is connected, and if few voltage is present something can goes wrong.

I feel I don't understand good enough your idea.
Sorry, the language difference make it for me a bit harder to understand the theory . 

[Ian.M]

The series string of resistors need to be in parallel with (across) the meter input jacks.

Take a small plastic project box and a pair of cheap multimeter leads.  Lay the leads across the box  and check the meter end is going to be long enough to reach the meter sockets without straining the wires.  Mark the leads in the center of where they cross the box.  Cut the leads there.  Drill the box half way up its sides, two holes in each long side, one at either end to feed the leads in and out the other side.  Rough up the inside of the box round the holes with sandpaper.  Thread the cut lead ends in through the holes and pull through enough length to work with.   Strip 1/2" of the end of each lead.  Solder the four 4K7 3W  wirewound resistors together in series, twisting each joint before soldering, then slip some glassfibre heat resistant sleeving over the whole lot.    For heat resistant supports for the resistor, take two pieces of thin ceramic tile, cut rectangles that just fit inside the box vertically and take up the full width, and drill one hole through the middle of each of them (use a small glass drill).   Thread them onto the resistor assembly end leads.  Twist and solder both red wire ends to the protruding resistor lead, then the same with the black wires, to the other resistor lead.   Glue the tiles in place with Epoxy or silicon caulk so the resistors cant touch the sides of the box.   Pull the wires tight, and fit a cable tie right up against the box inside wall at each hole so they cant be pulled out.  Secure each wire to the box wall at the cable tie with a blob of epoxy or silicon caulk.  Let all the glue set.  Screw the lid on the box and label it:

Low Impedance Voltage Test Lead for DMM
Max 500V for FIVE seconds
Check resistance before use: 18.8K (nom)

Usage is as described earlier.

The risk is minimal as long as you use a known good earth ground point for the tests.  If you aren't sure if a ground point is good don't use it.   Worst case: take a mains plug, and a length of hookup wire, connect the hookup wire to the ground pin, and screw the other two terminals down tight so the screws cant come loose, bulk up the hookup wire with rubber tape till the plug's cord grip can grip it securely, reassemble the plug, and use that, plugged into the same wall socket plate (or an extension off it) as your test bench equipment, and use the free end of that hookup wire as your test earth ground, AFTER checking the socket with a socket tester.

How much voltage is too much?  Well that depends.  Once you get below about 10V its not the voltage that's the problem, its how much current it can supply, which is why you then need to do the torch bulb test.  If the bulb lights, even dimly (or blows), *DONT* proceed with connecting the scope ground to that point.   Don't try to measure the leakage current with a multimeter - that's a good way to blow $expen$ive$  multimeter fuses if its more than expected.

If the above description of the apparatus required and methods to use it is still incomprehensible to you then *PLEASE* do not attempt to do any work on mains circuits or PSUs except under the direct supervision of a trained and qualified Electronic or Electrical appliance repair technician, with teaching experience.

[Chriss]

Thank's for this nice explanation, I read and translated several times your writings and now I'm on the road with it.
I'm familiar now with your thinking and suggestion.

I don't have many times to repair such thinks with mains power supply, but enough times to kill myself or my equipment.
So, It is better to ask and learn than to smell the flowers from below. 

My best regards.

[StillTrying]

Have you blown the DL+ up yet?

[Chriss]

If you mean my new scope?  No. I even don't get it for now. I have to wait 30-45 days for delivery.  :-)

Sent from my GT-I8260 using Tapatalk

[StillTrying]

I thought being near the center of The Universe Europe you'd get it in days, it's only ~£250 inc. in EU.