Very odd offline SMPS?

[cupoftea]

Hi,
A competitor has an isolated Flyback SMPS as follows....

100-265VAC
132kHz
Vout = 24V.
9W

The attached is roughly what it is...(LTspice and jpeg)

As you can see, there is no common mode choke and no Y capacitors. However, the Neutral is poured throughout the PCB on an inner layer, and goes under most of the primary and secondary sides. This neutral copper pour , with the PCB substrate, acts as a decent Y capacitor to the rest of the circuit.

The two 1mH inductors could have been put in Line and Neutral so as to give some common mode inductance, but this wasn’t done, as you can see.

The isolation from primary to secondary is 1500VAC , and the creepage/clearance between pri and sec copper nets is down to 2.5mm , since the particular application, only needs this.(?)

Also, the SMPS is on a PCB some 3cm wide by 13cm long…..the AC input is on the left….then the X2 capacitor…..then next in the layout is the secondary side output caps….then the flyback transformer…then the RCD clamp and the primary DC Bus caps are on the “secondary” side….also the primary side filter caps are on the “secondary” side……(please don’t ask me why its higgledy piggledy like this…I don’t know)

Do you think we are being “stuffy” by refusing to accept this design from the place that did it for our competitor?

 

[FvM]

Most likely you failed to reverse engineer the circuit correctly. As shown, the input rectifier is shorted.

 

[D.A.(Tony)Stewart]

very odd schematic. poof. Neutral cannot be DC ground , L1 must be on Cathode side of bridge.

It can't be isolated as shown.

But this is my result. Not very safe for line transients. But it runs.

 

[cupoftea]

Most likely you failed to reverse engineer the circuit correctly. As shown, the input rectifier is shorted.

..Thanks, sorry, my bad..i drew it wrong...there is no connection of pri and sec grounds.....transformer is isolated, and neutral is connected to secondary.
There is a triac switching the neutral......so the triac is switched from the sec side, hence the sec side is referenced to neutral.
Here it is drawn properly...my apologies.

 

[Calm down]

This is a deformation of a non isolated power supply, and BUCK is prone to problems when producing output voltages above 15V (I am not sure if it may be a driver level issue), so chip manufacturers usually recommend this approach. Alternatively, a single winding inductor can be used to output at both ends of the primary winding, and the high-voltage DC positive pole can also function normally as the output ground wire.

 

[cupoftea]

very odd schematic. poof. Neutral cannot be DC ground

Thanks, Neutral must be gnd in our application.

L1 must be on Cathode side of bridge.

Thanks, as you know, if L1 is in the opposite conductor to the L2, then you have a more balanced filter effect, which is better for CM emissions. (ie L1 in line and L2 in neutral)

Also, overall, there is less ripple current in the two electro caps if L1 is moved upstream of diode bridge.