Y Capacitor, current leakage & Electromagnetic Interference (EMI) problem.

[davidwyl]

over the years i purchased many phone charger from china. Most of the charger will give me shock "tingly"  sensation when touching the phone while charging.definately there is some current leakage and can give me unpleasant feeling.
Since im not using all these charger, so i decided to investigate and tear it apart. in order to reduce to "tingly"  feeling when touching i need to use a really low value for y capacitor or remove it completely. After i removed the capacitor, the "tingly" sensation is gone. From my knowledge remove y capacitor will increase EMI interference.
my question is:
Is it safe to use the charger without y capacitor??
how bad is EMI Interference to my phone/raspberry pi/arduino??
why manufacture put that capacitor in the first place?? to reduce the EMI but give people a shock every time touching the phone ?

[Simon]

I think the problem is the chargers are not earthed but they have connected the connector body up as though it were earth. these things should be outlawed, we have an earth for a reason but we don't do quality control anymore and safety laws are not a concern for the chinese.

[orolo]

I saw something related just yesterday:



In this case the fault was in the transformer, apparently. Be careful with those cheap wonders.

[Zero999]

The Y capacitor provides a return path for the capacitively coupled energy from the primary to the secondary transformer inside the switched mode power supply. Without it, the secondary will float at an AC voltage equal to the SMPS's switching frequency, with respect to the mains voltage or earth.

Yes, the chargers aren't earthed. One problem is some countries have two pin sockets for many devices, which aren't earthed and the power supply circuit is designed so that it can be used the world over. The tiny shock is not harmful and it's perfectly safe, so long as the Y capacitor meets all the relevant safety standards.

The are other potential problems with Y capacitors though. One is the tiny current can damaged sensitive devices which don't have ESD protection, although this can be solved by proper design. Another problem is, if many unearthed appliances with Y capacitors are connected together, the leakage currents will add up and can then exceed the safe levels.

Removing the Y capacitor is generally a bad idea. One alternative is to connect the Y capacitor to earth or, if the mains connector is polarised, neutral. That way all the EMI will be diverted to earth and there will be no leakage current.

Another possibility is adding a screen between the primary and secondary of the transformer and connecting that to the mains rectifier, via a Y capacitor, but you'd need to rewind the transformer to do that. It's something which could be done by the manufacture but it would cost more.

[davidwyl]

I believe these power supply is not damage . For example this Orico brand, I bought at least 6chanrger from them all give me the shock feeling . Initially I thought just device faulty but recently I bought another 2 from them, sadly it giving me same shock feel.
I believe this is something to do with the manufacturer try to control the emi under the requirements and totally forget about the current leakage.
This Orico sold thousands of charger under their company name I'm wondering why I'm the only one feel the shock and no one else. I didn't see any complaint whatsoever.
Is this something to do with the country I'm in is humid? Main socket voltage? Earth? I dunno, maybe if I take these charger and use in China the electric shock will gone. I seriously dunno.

[davidwyl]

Since removing the y capacitor is a bad idea , Is there a safe way to modify these charger to eliminate the shock feeling?

[Simon]

yea, give it an earth.

[Zero999]

Since removing the y capacitor is a bad idea , Is there a safe way to modify these charger to eliminate the shock feeling?

Removing the Y capacitor is generally a bad idea. One alternative is to connect the Y capacitor to earth or, if the mains connector is polarised, neutral. That way all the EMI will be diverted to earth and there will be no leakage current.

[davidwyl]

Thanks for the information...
All  These charger i owned are a 2pin non-polarised type so my option is to connect the y capacitor to earth? 
How to Connect it?
Primary (AC side) of the transformer ---> Ycapacitor--->earth??
 So I disconnect the Ycapacitor on the secondary side and connect it to earth pin?

[Gyro]

Unfortunately Y capacitors are there to eliminate (decouple) noise at RF frequencies. This means that they need to be very close to the noise source (the switcher). You can't just wire them externally to earth on lengths of wire, that would most likely create a radiating loop antenna, making the radiated interference worse.

As long as the Y caps are genuine, then they can't pass enough current to be anywhere near dangerous. Y caps are highly specified parts with very high breakdown voltage. The far higher risk from no-name adaptors is the quality of the insulation in the transformer, safety clearances on the PCB, inadequate overcurrent protection (squiggly line thin PCB traces) on the primary side.

[Zero999]

Unfortunately Y capacitors are there to eliminate (decouple) noise at RF frequencies. This means that they need to be very close to the noise source (the switcher). You can't just wire them externally to earth on lengths of wire, that would most likely create a radiating loop antenna, making the radiated interference worse.

It's not ideal but I still think it's better than removing the Y capacitor altogether. Don't use a long piece of separate cable. Use three core mains cable. The close proximity of the earth conductor to the live and neutral should minimise the interference.

Ideally there should be an X capacitor and another Y capacitor from the primary side to earth.

As long as the Y caps are genuine, then they can't pass enough current to be anywhere near dangerous. Y caps are highly specified parts with very high breakdown voltage. The far higher risk from no-name adaptors is the quality of the insulation in the transformer, safety clearances on the PCB, inadequate overcurrent protection (squiggly line thin PCB traces) on the primary side.

True but that doesn't stop it from being a nuisance or a problem if the 0V of many devices is connected together, as in the case of audio and visual equipment.

[davidwyl]

Thanks hero999 for the diagram.
What value of x y capacitor should I use? 
I have bunch of  2nf  capacitor around, will that ok to use in this?

[davidwyl]

As long as the Y caps are genuine, then they can't pass enough current to be anywhere near dangerous. Y caps are highly specified parts with very high breakdown voltage. The far higher risk from no-name adaptors is the quality of the insulation in the transformer, safety clearances on the PCB, inadequate overcurrent protection (squiggly line thin PCB traces) on the primary side.

I'm fully aware the power supply is not dangerous but it always give me unpleasant feeling. If I sit on my bed it is fine, once I step on the floor I will get very unpleasant tickling feel. Worst is when I'm using metal earphones, I can feel the tickle feeling on my ear too.

[T3sl4co1l]

Add a ground pin and connect the negative side of the power supply to GND.

This safely shorts the ground leakage current, and provides a return path for the RF noise besides.  (Some more Y caps from AC line/neutral to GND may also be necessary for reduced radio interference.)

Tim

[davidwyl]

If I just connect the Negative on the DC side to the Earth pin, is this safe to connect this way?
 

Sent from my ASUS_Z00AD using Tapatalk

[T3sl4co1l]

Even better, find where the USB shields go.  Should be the same -- check it with a meter.

Tim

[davidwyl]

Even better, find where the USB shields go.  Should be the same -- check it with a meter.

Tim

That is connected to the usb shell. Tested the connectivity.

Sent from my ASUS_Z00AD using Tapatalk

[Zero999]

Add a ground pin and connect the negative side of the power supply to GND.

This safely shorts the ground leakage current, and provides a return path for the RF noise besides.  (Some more Y caps from AC line/neutral to GND may also be necessary for reduced radio interference.)

Tim

That will work too. The only thing you now need to be wary of is the negative is now referenced to earth, which won't be a problem in this case.

If I just connect the Negative on the DC side to the Earth pin, is this safe to connect this way?

The connections are fine but please use proper three core mains cable, with an outer sheath, rather than individual conductors like that for the real thing.

[davidwyl]

That will work too. The only thing you now need to be wary of is the negative is now referenced to earth, which won't be a problem in this case.

Is there any situation where DC negative  referenced to Earth might be dangerous / causing any issue?


Sent from my ASUS_Z00AD using Tapatalk

[Zero999]

That will work too. The only thing you now need to be wary of is the negative is now referenced to earth, which won't be a problem in this case.

Is there any situation where DC negative  referenced to Earth might be dangerous / causing any issue?

It could be a problem if the circuit you're powering has a rail splitter and is connected to another device who's negative line is connected to earth.

Look at the attachment. Device 1 has a virtual earth circuit which splits the 12V from power pack 1 to + & - 6V. The  0V rail is connected to another device running from a power pack 2 which has its negative rail connected it earth. The circuit will work fine, until power pack 1's negative rail is connected to 0V.

This won't be an issue for your circuit, as the 0V goes to some USB connectors which are quite often earthed anyway.

[davidwyl]

if the mains connector is polarised, neutral. That way all the EMI will be diverted to earth and there will be no leakage current.

If I make the socket polarised (live always connect to pin1 , neutral always connect to pin2 & not reverseable),  then I can connect the DC negative straight to the Neutral?  Is what what u mean?
Simply diagram will be much appreciated 

[Zero999]

if the mains connector is polarised, neutral. That way all the EMI will be diverted to earth and there will be no leakage current.

If I make the socket polarised (live always connect to pin1 , neutral always connect to pin2 & not reverseable),  then I can connect the DC negative straight to the Neutral?  Is what what u mean?
Simply diagram will be much appreciated 

Oh no, never, connect the 0V directly to neutral.

I meant, change the Y cap so it goes to neutral, rather than the rectifier.

[davidwyl]

if the mains connector is polarised, neutral. That way all the EMI will be diverted to earth and there will be no leakage current.

If I make the socket polarised (live always connect to pin1 , neutral always connect to pin2 & not reverseable),  then I can connect the DC negative straight to the Neutral?  Is what what u mean?
Simply diagram will be much appreciated 

Oh no, never, connect the 0V directly to neutral.

I meant, change the Y cap so it goes to neutral, rather than the rectifier.

Is the x capacitor necessary? What range of value normally? 1nf - 2nf good enough?

I have bunch of 1nf capacitor can I use that for x and y?


Sent from my ASUS_Z00AD using Tapatalk

[Zero999]

if the mains connector is polarised, neutral. That way all the EMI will be diverted to earth and there will be no leakage current.

If I make the socket polarised (live always connect to pin1 , neutral always connect to pin2 & not reverseable),  then I can connect the DC negative straight to the Neutral?  Is what what u mean?
Simply diagram will be much appreciated 

Oh no, never, connect the 0V directly to neutral.

I meant, change the Y cap so it goes to neutral, rather than the rectifier.

Is the x capacitor necessary? What range of value normally? 1nf - 2nf good enough?

I have bunch of 1nf capacitor can I use that for x and y?


Sent from my ASUS_Z00AD using Tapatalk

The X capacitor should already be built-in to the power supply.

[Giaime]

That will work too. The only thing you now need to be wary of is the negative is now referenced to earth, which won't be a problem in this case.

Is there any situation where DC negative  referenced to Earth might be dangerous / causing any issue?

It could be a problem if the circuit you're powering has a rail splitter and is connected to another device who's negative line is connected to earth.

Look at the attachment. Device 1 has a virtual earth circuit which splits the 12V from power pack 1 to + & - 6V. The  0V rail is connected to another device running from a power pack 2 which has its negative rail connected it earth. The circuit will work fine, until power pack 1's negative rail is connected to 0V.

This won't be an issue for your circuit, as the 0V goes to some USB connectors which are quite often earthed anyway.

Direct connection between ground and earth (PE - protective earth) constitutes a PELV-type system: usually it's the most common practice for class I equipment (that is, with protective earth), for example IT stuff, audio/video equipment, etc. Just to give you an example, in lighting appliances (at least in Europe) it's not allowed, and only SELV systems are allowed (>1MOhm insulation resistance @ 500V between ground and PE).

One of the potential dangers of PELV systems is that if you have an earth fault, with huge currents flowing into the PE building conductors, it is allowed for the PE terminals to "shift" from the "real" earth up to 50V. This additional potential can "shift up" all potentials attached to it: if you have a 24V PELV system for example, it is safe to touch since it's less than 60V. But if you have an earth fault, some points in the 24V PELV system can reach to 74V which by some safety standards is not safe anymore to touch!