Unexpected AC Voltage on SMPS Output - Why?

[thannara123]

Forum Question Title:

Unexpected AC Voltage on SMPS Output - Why?

Question Body:

Hi everyone,

I'm working with a 230V AC input SMPS that provides a 12V DC output. The SMPS is functioning correctly, delivering the expected 12V DC. However, when I measure the output voltage with respect to ground (both on the positive and negative output pins), I'm also getting a 20 Volt AC component.

I am using the same Ground connection in some of electronics circuit , with the smps dc output , it kills my circuits components .
How to avoid it ?

 

[barry]

First you say it's functioning correctly, then you say it's got 20 VAC on the.
output. That sure doesn't sound like it's "functioning correctly" to me.

Your second paragraph makes no sense, could you rephrase it?

 

[thannara123]

I am driving an Hbridge with the above said SMPS. That smps gives 12 volt DC and required current .

I am powering controller (for pwm) from usb of PCB ,the PC has connected the ground ,these ground with respect to the smps out put pin shows 20 volt ac .
So the controller board killed by this supply .or heating up controller board components.
So I tried a power bank as a controller power source ,so there is no problem .working well

 

[KlausST]

Hi,

As so often: A sketch would be so helpful.

***
I can only guess that you are talking about some pri to sec stray current.
If so, it makes a huge difference what kind os power supply you are talking about.
* Is it in a metal case with three inputs (L, N, PE)
* or is it a plasic boxed with only two inputs (L, N)
* or is it some open frame - resued one
* or anything else
Best if you can give a link to the datasheet, atleast give manufacturer name and exact type.

Then We need to knpw how you measured the voltage, which voltmeter did you use?

In any case "heating up controller board and killing it" makes no sense.
If you need help in this you need to provide more informations what this "controller board" is and you need to tell about all it´s connections, wiring, WHAT exactly heated up .. and so on.

Klaus

 

[BradtheRad]

I guess the setup works well in some conditions but not in others?

Does anything change when you reverse the wall plug? Perhaps your house wiring consists of one live and one neutral?
This reminds me of years ago when we used 2-pronged plugs (in the U.S.). Some appliances carried a slightly detectable amount of high voltage on their chassis. I say slightly because it was just a constant tingle, not a bad jolt. Grounding yourself through a live chassis was annoying since the tingle could approach house voltage yet the current was so weak it did no harm.

Audio equipment usually produced a soft hum making it necessary for us to test each wall plug's orientation until we found the correct arrangement that was quiet (especially in headphones). I owned tube-type guitar amplifiers that had live chassis (or felt that way) when the plug was in the wrong position.

 

[FvM]

PE connection wasn't mentioned but measured voltage suggests there is none. Without PE-connection, a small common mode current is injected through Y-capacitor between primary and secondary. A high impedance meter shows a respective common mode voltage.

 

[thannara123]

Sorry for the Late
Here is the block where I am using the circuit connection as above said .
In that circuit Point Anad B or Point C and D am getting a 20 volt ac , which some times kills the st link programmer or the controller board

 

[BradtheRad]

I remember wondering whether I should expect to read 120VAC on my tube amplifier chassis and why I saw much less... I decided measuring with my VOM (10,000Ω/V or 1.5MΩ) loaded the high voltage, causing a reduced reading because the rogue AC came through several MΩ. And so many commercial schematics show a high ohm resistor (or tiny value capacitor) going from circuitry to chassis (for one reason or another).

I think you're seeing house voltage dropped drastically because it comes through several MΩ (somewhere). So your meter reading is 20VAC. I'm pretty sure you can install back-to-back zener diodes rated for a voltage at a safe level for your equipment.

 

[danadakk]

Isolation problem you are seeing with PC supply ?

Page 21.....

 

[D.A.(Tony)Stewart]

Two unknown SMPS converters with a USB data link in between is likely to experience stray coupling unless both ends of DC 0V are also PE grounded.

Gnd is 0V from DC to any frequency of importance ( by definition until we inspect real ground noise to some reference with proper measurement methods) .

If you have a common mode current converted to a differential voltage where the impedance is unbalanced, we need a real picture of what is used and how it is connected.

You'd be better off using the PC supply from an HDD spare Molex connector.

Or if using a laptop disconnect the charger or PE ground the case.

Medical SMPSs with good isolation use a couple of different CM chokes to span the BW after an isolated design to balance the lines without a ground and reduce CM noise. You don't have this.

Unknown supplies must be tested for noise and fixed before using relative to the ground at the other end of your link.