AC Leakage Detected at AC/DC Adptor DC output negative

Thread Starter

Jonathan Foong

Joined Mar 13, 2021
50
This issue had been many times happened to me whenever I bought AC/DC Adaptors. I am not sure what is going wrong with these units. Hope there is an answer from this forum.

I tested my power source is perfect.
Please refer to my attached pictures for my AC Leakage issue from Adaptors.

Thank you.
 

Attachments

DickCappels

Joined Aug 21, 2008
10,153
The neon light only tells you whether there is a path to line voltage, as opposed to a path to neutral or earth ground.

To know whether the leakage current is dangerous you need to know how much current is available through a "human body model".

Basically, you put an impedance between the point of suspected leakage and earth ground, measure the voltage across the impedance then calculate the leakage current I = E/Z.

The body model (IEC60601-1), the component values, and acceptable maximum leakage currents for a variety of cases is explained here:

(Go to the “Measurement of leakage current” section near the bottom of the page.)
https://www.chromausa.com/applications/safety-compliance-test/leakage-current-test/
 

Thread Starter

Jonathan Foong

Joined Mar 13, 2021
50
The neon light only tells you whether there is a path to line voltage, as opposed to a path to neutral or earth ground.

To know whether the leakage current is dangerous you need to know how much current is available through a "human body model".

Basically, you put an impedance between the point of suspected leakage and earth ground, measure the voltage across the impedance then calculate the leakage current I = E/Z.

The body model (IEC60601-1), the component values, and acceptable maximum leakage currents for a variety of cases is explained here:

(Go to the “Measurement of leakage current” section near the bottom of the page.)
https://www.chromausa.com/applications/safety-compliance-test/leakage-current-test/
Hi, It is my pleasure to receive your explanation and information for further measurement of the leakage current amount.

I learned a lot about this issue from your reply.

Thank you very much.
 

Ian0

Joined Aug 7, 2020
9,668
That’s normal. There is a capacitor between the rectified AC negative and the output for EMI suppression. It is allowed to leak 700uA to the output. Connect 22k resistor from output to earth, and make sure that the voltage across it is <15V.
But don’t go connecting a load of those supplies in parallel!
 

Thread Starter

Jonathan Foong

Joined Mar 13, 2021
50
That’s normal. There is a capacitor between the rectified AC negative and the output for EMI suppression. It is allowed to leak 700uA to the output. Connect 22k resistor from output to earth, and make sure that the voltage across it is <15V.
But don’t go connecting a load of those supplies in parallel!
Hi, thank you very much for your response, and advice solution to rectify this issue. I am most interested to resolve this issue, because I had bought more than 10 AD/DC adaptors in the past having this similar issue, therefore it will be good if you could provide me a typical sample circuit together with your advice to demonstrate for this solution.

Thanks.
 
Last edited:

Ian0

Joined Aug 7, 2020
9,668
No need to do anything unless they fail the >15V across the 22k resistor test.
If they pass, they are OK to use.
If they fail, return them to the supplier.
 

Dodgydave

Joined Jun 22, 2012
11,284
Every switchmode psu which is in a plastic case has a BLUE EMI suppression capacitor across the negative output and negative mains side , (as you can see on the picture), the only way to stop this mains leaking is to remove it, Not Recommended.USB-Charging.jpg
 
Last edited:

Ian0

Joined Aug 7, 2020
9,668
Every switchmode psu which is in a plastic case has a BLUE EMI suppression capacitor across the negative output and negative mains side , (as you can see on the picture), the only way to stop this mains leaking is to remove it.View attachment 273984
But don't remove it unless you want it to radiate interference everywhere. That class-Y capacitor makes the biggest difference of any component as to whether it passes EMC regulations or it doesn't.
 

DickCappels

Joined Aug 21, 2008
10,153
But don't do that...the capacitors are in there for a reason - to prevent conducted and radiated signals that might interfere with wireless and in some cases wired communications. The value of the capacitors in question were chosen so that the ground leakage current limit would not be exceeded. If the leakage current is excessive, something is wrong. I it might be the capacitors are of a higher value than intended (this is probably extremely rare) or some other defect in the circuit. If you detect excess leakage the wise choice is to just don't trust it any more. You can toss it into the trash, return the unit to whomever you sold it to, or leave it in the junk box for spare parts.

You might even try to fix it.

You will never see zero leakage current power supplies, unless your instrumentation is defective.
 

Thread Starter

Jonathan Foong

Joined Mar 13, 2021
50
But don't do that...the capacitors are in there for a reason - to prevent conducted and radiated signals that might interfere with wireless and in some cases wired communications. The value of the capacitors in question were chosen so that the ground leakage current limit would not be exceeded. If the leakage current is excessive, something is wrong. I it might be the capacitors are of a higher value than intended (this is probably extremely rare) or some other defect in the circuit. If you detect excess leakage the wise choice is to just don't trust it any more. You can toss it into the trash, return the unit to whomever you sold it to, or leave it in the junk box for spare parts.

You might even try to fix it.

You will never see zero leakage current power supplies, unless your instrumentation is defective.
Hi, it is my pleasure to receive your further explanation in detail and remedy actions.

Thank you again.
 

Thread Starter

Jonathan Foong

Joined Mar 13, 2021
50
That’s normal. There is a capacitor between the rectified AC negative and the output for EMI suppression. It is allowed to leak 700uA to the output. Connect 22k resistor from output to earth, and make sure that the voltage across it is <15V.
But don’t go connecting a load of those supplies in parallel!

Hi, I refer to your advice on the connections and test measurement method, I am most interested to sort out the failed PSU.

Can you please confirm what I understand per my attached illustration, please correct me where I am doubt or wrong.

1. 22K resistor connections
2. Voltmeter connection and AC or DC type
3. >15V Fail
PSU AC Leakage Test.jpg

Look forward to your value confirmation.

Thank you very much.
 
Last edited:

Ian0

Joined Aug 7, 2020
9,668
measure AC voltage.
You have drawn your PSU with an earth connection. Generally what you describe will not be a problem with PSUs with earth connections. It's only the double-insulated ones that cause this problem.
 

Thread Starter

Jonathan Foong

Joined Mar 13, 2021
50
measure AC voltage.
You have drawn your PSU with an earth connection. Generally what you describe will not be a problem with PSUs with earth connections. It's only the double-insulated ones that cause this problem.
Hi, thanks for your prompt reply, the type of Adaptor with AC leakage problem per the attached picture, is not sure if it is a double insulated one.
 

Attachments

Ian0

Joined Aug 7, 2020
9,668
Hi, thanks for your prompt reply, the type of Adaptor with AC leakage problem per the attached picture, is not sure if it is a double insulated one.
Maybe - maybe not - the bit I need to see is further down the label out of focus. If it is double insulated then there should be a symbol of two concentric squares.
I also would not expect a 3-pin connector if it wasn't earthed.
Is the negative of the output connected to earth?
Even if it isn't I would not have expected much leakage current on the output, a lot less than a double insulated product.
 

Thread Starter

Jonathan Foong

Joined Mar 13, 2021
50
Maybe - maybe not - the bit I need to see is further down the label out of focus. If it is double insulated then there should be a symbol of two concentric squares.
I also would not expect a 3-pin connector if it wasn't earthed.
Is the negative of the output connected to earth?
Even if it isn't I would not have expected much leakage current on the output, a lot less than a double insulated product.
Hi, I am trying to attach pictures for you to evaluate further. Hope can be assisted me to understand further about this issue.

I attached 2 pictures :
one is BSY-1203 which is an AC leakage problem and the other one is A140 which is perfect without AC leakage determined by a test Pen with no neon light visual.

Thanks.
 

Attachments

Last edited:

Ian0

Joined Aug 7, 2020
9,668
The exact measurement procedure is part of a British Standard, but I can‘t remember which one. If I find it I’ll post a scan of the relevant paragraph.
 
Top