Why has designer not used X2 capacitors in offline SMPS design?

[Flyback]

Hello,
We are getting a 100W, 90-265VAC SMPS designed for us by a contractor. It is a Boundary conduction mode boost PFC followed by a quasi resonant flyback (output is 40V, 2.5A).
The contractor has used a small common mode choke and a 10n X2 capacitor upstream of the mains diode bridge. That, as well as the NTC and fuse, is all he has fitted upstream of the mains diode bridge.
Then immediately downstream of the mains diode bridge he has a 47mH common mode choke, which has a 150n,400V metal poloypropylene capacitor either side of it. Here is the datasheet of these two 150n,400V polyprop caps…..
Datasheet of the 150n,400V metal polypropylene capacitors: (ECWFD series)
**broken link removed**
Why has he done it like this?
As you know, in most offline SMPS’s, the main common mode choke usually exists upstream of the diode bridge, and has X2 capacitors either side of it, not plain metal polypropylebe capacitors.

 

[Cicero]

I dont really know, I'm mainly commenting to get updates to this thread

I'll add my input anyway. I am also sceptical, because I believe where he's placed these caps still constitutes a primary circuit, and as such the caps need to be properly rated and approved, X2 caps.
It only becomes a secondary circuit after the isolation, anything prior needs proper safety ratings, thats my understanding.

 

[JimB]

Flyback asks:

Why has designer not used X2 capacitors

He then goes on to state:

We are getting a 100W, 90-265VAC SMPS designed for us by a contractor.

JimB comments:
As this is an item being designed for you, do you not take part in a design review process, however informal?
Just ask them... "Oi, whats all this about? Should you not be using an X2 capacitor here?"

Another whining post by Flyback wanting us to redesign his day job work for him.

JimB

 

[Cicero]

Flyback asks:

He then goes on to state:


JimB comments:
As this is an item being designed for you, do you not take part in a design review process, however informal?
Just ask them... "Oi, whats all this about? Should you not be using an X2 capacitor here?"

Another whining post by Flyback wanting us to redesign his day job work for him.

JimB

Ey, this seems offsides. This is a place for asking questions is it not, I dont see why this question is so ridiculous that it warrants this response?

 

[Tony Stewart]

A diode bridge does many things including;
- Multiply the fundamental frequency x2
- divide the Vpp ac in half
DC shunt caps have ripple charged by rectangular current pulses whenever bridge voltage exceed Vcap

But during the conduction transition, when diodes turn Off , there is an RF resonance with a Vpp amplitude equal to the diode drop, due to series line inductance, load capacitance and back EMF pulse from the switched current cutoff in the line inductance. Although this is a differential current loop, imbalance in the bridge load when one side is grounded, causes a large CM noise leakage which when the impedance is raised by a CM choke is then suppressed in both directions by the C-L-C or "Pi" filter ,which is symbolicly shaped like the Greek letter, ∏ or π.

Contractor's get paid to do work, not teach newbies.

 

[Cicero]

A diode bridge does many things including;
- Multiply the fundamental frequency x2
- divide the Vpp ac in half
DC shunt caps have ripple charged by rectangular current pulses whenever bridge voltage exceed Vcap

But during the conduction transition, when diodes turn Off , there is an RF resonance with a Vpp amplitude equal to the diode drop, due to series line inductance, load capacitance and back EMF pulse from the switched current cutoff in the line inductance. Although this is a differential current loop, imbalance in the bridge load when one side is grounded, causes a large CM noise leakage which when the impedance is raised by a CM choke is then suppressed in both directions by the C-L-C or "Pi" filter ,which is symbolicly shaped like the Greek letter, ∏ or π.

Contractor's get paid to do work, not teach newbies.

"Vinny: Well, thank you for that. That's a great weight off me mind."

Apart from stating the obvious in an attempt to look superior, you really just come off as patronising where it is not warranted.

Seriously Tony, what you have said has little to nothing to do with Flybacks question, which is all centred around safety approvals and the use of a non safety approved cap prior to isolation.

 

[Tony Stewart]

Your perception may be valid for you, Cicero, but if you don't understand the subtle byproducts from transformation from AC to DC as I described, then you won't understand why the current limiting of diode ESR will prevent a crowbar failure on the grid. Which is why they don't need X2 ratings, but have ESR much lower than the bridge diodes.

Whereas the grid may have much lower ESR than the caps unless a DM line filter is inserted, and thus SCR mode failure will have explosive effects from follow on currents unless internally fused self-healing X2 rated caps which are mandatory for primary side of bridge across the line.
External fuse offers extra safety.


I answered the question, why it was designed this way.

No ego here, and please no more drama.

 

[Tony Stewart]

Also the secondary return cannot be earth grounded otherwise it becomes a Y rated line to ground hazard for ground induced voltages in the presence of a groundfault.

Y1 caps are tiny 0.25mA line current snubbers, which may be added to earth ground to suppression CM noise between earth ground and bridge secondary DC, which will have each side 2 diodes drops from line peak voltage and Neutral, so it is preferred Y1 caps if used for EMI resonant issues in bridge , added on primary to bridge.

These capacitors are divided into 2 classifications, X and Y. X class capacitors are capacitors that are connected line to line and in the event of failure of the capacitor the potential for electrical shock is not present. X capacitors are further subdivided into three subcategories X1, X2 and X3. X1 capacitors are used where the peak voltage the capacitors will be greater than 2500 volts and less than 4000V. Class X2 capacitors are in applications where the peak voltage is equal to below 2500 volts. X3 capacitors are used where the peak voltage is less than or equal to 1200 volts. X2 capacitors are the most common.

Y capacitors on the other hand are connected from line to ground. They are typically a very low capacitance value. In the event a Y capacitor fails the potential for electrical shock is present. Y capacitors are also subdivided into four subcategories, Y1, Y2, Y3 and Y4. Y1 capacitors are used with voltages up 500Vac, Y2 are used with voltages up to 300Vac,Y3 are used with voltages up to 250Vac and Y4 capacitors are used up to 150Vac
. ref. Illinois Caps. Inc.

 

[Cicero]

Your perception may be valid for you, Cicero, but if you don't understand the transformation from AC to DC as I described, then you won't understand why the current limiting of diode ESR will prevent a crowbar failure on the grid. Which is why they don't need X2 ratings, but have ESR much lower than the bridge diodes.

No ego here, and please no drama.

Well why include "Contractor's get paid to do work, not teach newbies"? It sets the tone, and invites drama. I will drop it for now.

In all seriousness, a direct quote from UL 60950:

A capacitor connected between two line conductors in a PRIMARY CIRCUIT, or between one line
conductor and the neutral conductor or between the PRIMARY CIRCUIT and protective earth shall
comply with one of the subclasses of IEC 60384-14 and shall be used in accordance with its
rating. This requirement also applies to a capacitor bridging DOUBLE INSULATION or REINFORCED
INSULATION elsewhere in the equipment.

The caps referenced are all X and Y rated. And this is my concern...and I think Flybacks as well.

 

[Flyback]

It is a concern, though i am also wondering why the common mode choke has been moved to downstream of the mains rectifier. The two X caps which one normally sees either side of the common mode choke have been converted to simple metal polyprop caps and moved downstream of the mains rectifier with the common mode choke.

There is one x cap upstream of the mains rectifier, and that is a 10n, x2 cap. It has another common mode choke upstream of it.

I thought all of the the common mode chokes had to go upstream of the mains rectifier because the high frequency stuff could otherwise bypass it and get in through the input connector?

I also thought that all the x rated caps had to go upstream of the mains rectifier, because there they can quench transients before the transients get downstream of the mains rectifier. Also the mains rectifier itself needs protection from transients.

As you know, this is PFC'd supply, so there is very little capacitance downstream of the mains rectifier

 

[Tony Stewart]

Well why include "Contractor's get paid to do work, not teach newbies"? .

This explains perhaps why he didn't get answers from Contractor.
Did anyone not understand reasoning for post bridge filtering and lack of Y rating.

X,Y rated caps are also Polypropylene, Polyester or Some kind of plastic film. The difference is subtle in the conductor distributed fusing for microscopic defects that cause breakdown first.

 

[AnalogKid]

It is a concern, though i am also wondering why the common mode choke has been moved to downstream of the mains rectifier.

Are you sure that's a CM choke? In a boost converter PFC supply, the first thing after the bridge is the boost inductor. Also, any reactive components (caps, CM chokes. etc.) introduce a phase shift between input voltage and current that introduces an error into the PFC control chip circuits. Quieting a PFC supply is a game of tradeoffs.

ak

 

[Flyback]

It is a CMC., one of three in this PSU