Why was i electrocuted?

[treez]

Hello,
I went to disconnect the BNC of a coaxial probe from the scope, and received a snappy electric shock. Do you know why?

I have been disconnecting this same BNC connector on previous days, and wasnt electrocuted, even with the same overall setup.

I had to disconnect the BNC with the circuit "live", so i could not switch everything off to do the disconnection.

At first i couldnt touch the BNC because of the shocks.....then i kind of "pawed" at the BNC, touching it lightly then off again, then i was able to disconnect it without getting shocked.
The setup is as attached.

 

[KlausST]

Hi,

That's why you should use series resistors with ESD earthing.

In your case there may be a couple of possible effects causing this problem.
Maybe the ESD mat is more conductive than needed and feeding the voltage from the DUT to your arm.

Klaus

 

[treez]

Maybe the ESD mat is more conductive than needed and feeding the voltage from the DUT to your arm.

Thanks, it is a very conductive ESD mat...if you touch it with live and neutral 1 inch apart, the mat smokes.

But the circuit under test was on an earthed heatsink which sits on the esd mat.

I am wondering if the inside of the BNC has come apart a bit, though the supplying PSU is isolated, so i dont see why that would be a problem either.

 

[betwixt]

I suspect the problem is with the "isolated 230V AC". How exactly is it isolated?

You oscilloscope, mat and wrist band are probably at mains Earth potential but the 'isolated 230V" probably isn't. Isolating it means there is no direct current path between the input and output sides but it doesn't mean the output is at ground potential. Even if it was Earthed (at least on one of the output lines), if your LED driver has a bridge at it's input, the sense resistor is probably at elevated voltage.

Four things come to mind:
1. When working on high voltages, a static dissipative mat should be used, not a conductive one, especially one with such low resistance.
2. Your ESD strap should have a resistor in line with the wire to limit the current that can flow under worst case conditions to a safe level.
3. You should use a second leakage path (like a second ESD strap wire) to the 'ground' of the LED driver so it can't float if the probe ground is removed.
4. Use a good probe instead of your 'stripped coax' or you will get unwanted capacitive effects on your circuit, ideally use a x10 probe for minimum loading.

On point 4, bear in mind that cables like RG58 and RG59 typically have a capacitance of between 75pF and 100pF per metre and it will appear in parallel with the scope input capacitance. They are also normally rated to only 300V so be careful when using them at mains voltages. If you are using audio grade coax expect a much higher capacitance and lower voltage rating.

Brian.

 

[treez]

I suspect the problem is with the "isolated 230V AC". How exactly is it isolated?

Thanks, the manual doesn’t say and doesn’t give isolation voltage etc.
We know its isolated because I connected earth to live_out via a 1 amp fuse and turned the output up to 230vac and the fuse didn’t blow..therefore we know its isolated.
The sense resistor is in the negative_DC_bus, just downstream of the mains rectifier.
As you know, the BNC of the coaxial probe is connected to the scope, so the sense resistor “should” have been at earth potential.

When working on high voltages, a static dissipative mat should be used, not a conductive one, especially one with such low resistance.

Thanks, I assume this is for safety reasons?

You should use a second leakage path (like a second ESD strap wire) to the 'ground' of the LED driver so it can't float if the probe ground is removed.

Thanks, that sounds a good idea though it wouldn’t be possible with this measurment as we have the grounded sheath of the coaxial probe soldered to the “non-ground” side of the sense resistor….(so that it gives a positive voltage on the scope because having negative going voltages makes things too confusing for our poor, low_IQ heads)
…..but I think you are on to something there…..maybe when I went to unscrew the BNC, it still appeared to be on the scope at first but just maybe the actual connection from BNC ground to scope ground was not made and so the BNC floated up and then I got zapped?

Use a good probe instead of your 'stripped coax' or you will get unwanted capacitive effects on your circuit, ideally use a x10 probe for minimum loading.

…thanks, I’d like to do just that…but you cant find x10 probes that are coaxial…..well, you can get scope probes as you know, which are indeed coaxial, but they have the dangling ground clip which picks up noise……the coax x1 probe is good because you can strip the end and solder it right on to the sense resistor……yes the capacitance is not wanted but its not so bad in our particular case setup at the moment.
Ive tried stripping a x10 scope probe so I could use it like I’m using the x1 coax probe , but the scope leads have a terribly thin and delicate ground sheath, which can’t be soldered to the sense resistor with any convenience.

They are also normally rated to only 300V so be careful when using them at mains voltages

Thanks, the voltage across the sense resistor is less than one volt so I presumed we’d be ok using the x1 coaxial probe?

 

[FvM]

Thanks, the manual doesn’t say and doesn’t give isolation voltage etc.
We know its isolated because I connected earth to live_out via a 1 amp fuse and turned the output up to 230vac and the fuse didn’t blow..therefore we know its isolated.

I won't worry about isolation voltage, it's surely adequate for 230 VAC. But what's the leakage current respectively insulation resistance specification?

You need a leakage current below 1 mA to safely touch the floating circuit.

 

[treez]

I won't worry about isolation voltage, it's surely adequate for 230 VAC. But what's the leakage current respectively insulation resistance specification?

Thanks, the manual of the AC PSU didnt say.
In fact, i asked that very question about the PSU on edaboard in about April or May this year...but i cannot seem to find it now.
I am not at work so cannot read the part number but its in my post from April or May on this forum, i will continue seeking this post.

 

[FvM]

It's easy to measure insulation resistance and leakage current for live and neutral output.

 

[SLK001]

Since it was isolated, the power supply ground connected to the DUT was at a potential higher than your earth ground - which you were connected to. As you "pawed" the BNC cable, you slowly brought the two grounds to the same potential - through your body. Once fairly equalized, you could touch the BNC without being shocked.

 

[treez]

Since it was isolated, the power supply ground connected to the DUT was at a potential higher than your earth ground

Thanks, but i must add that the coaxial probe, which was connected to the scope via its BNC connector, was , at its other end, soldered to the PCB at the sense resistor, so therefore the PCB of the LED driver should have been at earth potential...at least at the point where the earthed sheath of the coax probe was soldered to it.

- - - Updated - - -

It appears, on the face of it, that i was electocuted by touching Earth potential.

So i suspect that the answer above by Betwixt seems correct...that in fact, the outer bit of the BNC connector which i touched, must have not been at earth potential....maybe somehow the "thing" inside the BNC connector that should recult in it getting connected to the scope BNC connection was not actually connected. or it came away as i begun to unscrew the BNC .

But then again, i know what SLK001 means, because i did seem to stop myself getting electrocuted by dabbing and pawing the BNC shell with my finger....till the point where i could grab it without getting electrocuted.

- - - Updated - - -

The reason i had to do this live is because the led driver , for some reason, malfunctions when it is started up with the coaxial probe in place, and the only way you can get the led driver working and use the coaxial probe , is by disconnecting and then reconnecting the coaxial probe once the led driver has gotten started up.

- - - Updated - - -

I suppose i could start it up without the coaxial probe in and then connect it when the led driver has started up, , but presumably that woudl have still shocked me.

 

[betwixt]

The sense resistor is in the negative_DC_bus, just downstream of the mains rectifier.
As you know, the BNC of the coaxial probe is connected to the scope, so the sense resistor “should” have been at earth potential.

Double check this - based on your earlier posts you make a conflicting statement.
You state that one side of the isolated AC is grounded (that's fine) then say you connect scope ground to the sense resistor downstream of the mains rectifier. If, as I suspect, you use a bridge rectifier, the negative side sense resistor is NOT at ground potential. At best you are shorting out one diode in the bridge and worst case you are floating all the other nodes at an unpredictable voltage. When using a bridge rectifier, never assume the DC output can be grounded unless both AC inputs are floating.

Brian.

 

[ads-ee]

When working on high voltages, a static dissipative mat should be used, not a conductive one, especially one with such low resistance.

Thanks, I assume this is for safety reasons?

You should use a static disipative mat as a conductive one will still transfer charge quickly, which is exactly what you don't want for any kind of ESD protection. The 1 Mohm series resistor in a wrist strap is the safety feature, to avoid a low resistive path somewhere else through you.

 

[treez]

You state that one side of the isolated AC is grounded (that's fine)

Thanks but sorry i do not understand what is meant here. To clarify, i do not know if neutral_out of the isolated AC PSU is connected to earth potential. I will check this. I do know that live_out of the isolated AC PSU is not connected to earth....oh sorry,... i actually did a test whereby i connected the live_out to earth (via a fuse) and switched the isolated output on at 230VAC......so this confirms that both live_out and neutral_out of the isolated AC PSU are isolated from earth ground. As such, i cannot be shorting out any of the diodes in the mains rectifier.

 

[betwixt]

Sounds like I raised a false alert then - for which I apologize.

The concern I had was because if you are using an isolating transformer at the mains input and ground one side of the secondary to stop it floating, then connect a second ground connection via the scope cable after the bridge rectifier, it implies half the bridge is shorted out. On the other hand, if you don't ground one side of the isolation transformer secondary, but do ground the scope wire, you are making the current sensing resistor your zero reference which could have other implications. Without a schematic and details of the isolating transformer it is difficult to be sure but I wonder if this has anything to do with your reports of products failing when the scope probe is connected.

Brian.

 

[treez]

Thanks, though please dont apologize, your comments are of great use and really appreciated. I am going to look into this. Especially with regard to our products that malfunction when scope probe attached.

 

[treez]

You should use a static disipative mat as a conductive one will still transfer charge quickly, which is exactly what you don't want for any kind of ESD protection.

Thanks, though i wonder why the conductive ones we have are for sale?

After shocking myself again with the setup of the top post, i am beginning to wonder if the scope socket is actually at earth porential after all....maybe the earth wiring in our building does not actually go to "planet earth" (eg via an earthing rod), and maybe the "earth" connection is floating at who knows what voltage?

 

[FvM]

After shocking myself again with the setup of the top post, i am beginning to wonder if the scope socket is actually at earth porential after all....maybe the earth wiring in our building does not actually go to "planet earth" (eg via an earthing rod), and maybe the "earth" connection is floating at who knows what voltage?

I bet that the protective earth in your building is O.K. but it can be checked quite easily.

Are progressively pushing back from your mind that the expectable level of leakage current makes any circuit node of your "isolated" test setup a hazardous contact voltage?

 

[c_mitra]

....maybe the earth wiring in our building does not actually go to "planet earth" (eg via an earthing rod), and maybe the "earth" connection is floating at who knows what voltage?

Quite possible but unlikely. Please test with a "fuse" or a "electrician's lamp" as well as with a neon lamp tester. If the fuse blows or the lamp lights, you have a short. If the neon lamp show voltage, you have a leakage.

The body of the scope is connected to the earth ground at the three pin power point. You are not expected to float the scope (or connect with a two pin plug).

Protective mats with low resistance are useless because (i) if you are charged, you will get a shock; (ii) if you are not charged but touch a charged body then you will also get a shock.

ALSO

low resistance mats allow high currents and that produces large dV/dt and that can cause damage to equipments (semiconductors). - High voltage but const will not cause any damage (it will be floating)

That you are getting shock only for the first time suggests that some leakage some where is the culprit.

 

[Relayer]

Hello treez,
Does your scope and AC PSU have a 2 or 3 pin mains plug on them?
Regards,
Relayer

 

[treez]

Does your scope and AC PSU have a 2 or 3 pin mains plug on them?

thanks, it has a 3 pin plug

I won't worry about isolation voltage, it's surely adequate for 230 VAC. But what's the leakage current respectively insulation resistance specification?

Thanks, its a Lisun Group AC PSU, but they dont sell it any more, and there are no details on it..its not one of the following, and must have gone obsolete
https://www.lisungroup.com/prolist-id-169.html

You need a leakage current below 1 mA to safely touch the floating circuit.

..Thanks i appreciate that if its floating, but in the case of the top post, it wasnt exactly floating any more because i earthed it at a point with the scope due to the coaxial scope probe connection.

Surely at the end of the day, no matter how much leakage current is flowing, i touched a BNC plug which was connected into the scopes earthed BNC socket...and these should have been at earth potential, and yet i was shocked.