Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Offline 40W Flyback SMPS PCB near to Earthed metal enclosure…good or bad?

Status
Not open for further replies.
T

treez

Guest
Hi,

1..........If there is an offline 40W flyback SMPS on a PCB which is inside a large rectangular earthed metal enclosure…then should the Flyback PCB (and specifically the flyback switching node), be as near as possible to the earthed metal enclosure walls, or as far as possible from same?

The earthed metal enclosure is not being used for heatsinking of the flyback SMPS components.
............................_________________.............................
Also,
2......If you have an offline 90W Laptop SMPS in the usual plastic enclosure, with aluminium foil “coated” completely round the inside walls of the plastic enclosure for the purpose of Radiated emissions shielding…..then should this metal foil shield be connected to either....

A…Primary DC Bus ground?
B….Earth?
 

1, far from the walls

2, If it extends from pri to sec side - then earthed.
 
  • Like
Reactions: treez

    T

    Points: 2
    Helpful Answer Positive Rating
Thankyou. Also, for the PCB in (1) above, where should the earth wire connect to the earthed metal enclosure?
Should the incoming earth wire only be connected to the earthed metal enclosure at the earth wire's entrance to the enclosure?
And regarding the PCB in the earthed metal enclosure, should its only connetion to earth be via the PCBs incoming earth wire?...(ie, preferably not having the pcb also earthed via several metal pillars from the earthed metal enclosure to the PCB)?
 

Hi,

To question 2):

I'm no HF expert.
You talk about radiated noise, HF noise. Probably coming from the switching node(s), with all it's overtones.
So I think a short and low impedance return path for this HF should be a good start.
Safety connection to the metal case is something different, but in best case you can do all in one.
Where the safety path needs to be wide enough (to carry a high fault current) and low ohmic (for low frequencies).
...the HF path may be purely capacitive.
Now it depends from which place of the PCB (or transformer outer winding?), primary side or secondary side, the noise comes from...
exactly to this place (PCB GND plane) I'd wire the capacitor.

But often there are several sources of noise, maybe one part is from the primary side and a part is from the secondary side...then you need to find a good compromise.

Maybe build a couple of devices with different capacitor connections and run EMI tests to find out which one is the best.

Klaus
 
  • Like
Reactions: treez

    T

    Points: 2
    Helpful Answer Positive Rating
a long earth wire - before connecting to the case - is problematic, the shorter the better ...
 
  • Like
Reactions: treez

    T

    Points: 2
    Helpful Answer Positive Rating
be as near as possible to the earthed metal enclosure walls, or as far as possible from same?

Static answer: it does not matter. The electric field inside a conducting metal enclosure is zero.

Now consider the electric circuit placed within the box (no electrical DC connection to the metal box). The metal enclosure can be charged to arbitrary potential without affecting the functioning of the circuit board inside.

What happens when the potential on the metal box changes rapidly? We get both electric field and magnetic field and the associated EM effects still do not enter the inside of the box. It is basically a farady cage.

Now consider the circuit inside is connected to the enclosure with a zero impedance connection. Inside is still field free but the potential of the enclosure will be seen by the circuit.

Basically grounding is a safety feature of the common man.
 
a long earth wire - before connecting to the case - is problematic, the shorter the better ...
Thanks, sometimes we will have a mains lead containing L,N&E coming several metres from the wall socket to the unit.
Or do you mean the length of the earth wire inside the metal enclosure?, ..ie its length inside this enclosure to the SMPS?

Also, is it a good idea to use say 10cm lengths of shielded cable at the input and output of the flyback SMPS?. That is, shield the 10cm of mains cable coming to the Flyback maisn input, and the 10cm of dcdc output cable going from the flyback output. I mean surely if you can keep the high frequency radiation off the cable near the SMPS PCB, then you keep it off the whole cable?
 
Last edited by a moderator:

a long earth wire - before connecting to the case - is problematic, the shorter the better ...




if the earth wire is inside the case and long before it connects to the case or metal spray ...
 
  • Like
Reactions: treez

    T

    Points: 2
    Helpful Answer Positive Rating
Thanks, sometimes we will have a mains lead containing L,N&E coming several metres from the wall socket to the unit. Or do you mean the length of the earth wire inside the metal enclosure?, ..ie its length inside this enclosure to the SMPS?

What matters is the effective impedance of the line. That is all that the circuit inside the box can see and the parameter that matters.

The ground symbol in the schematic is a device with characteristic properties. That lumped property is to be considered along with the wiring that finally goes deep inside mother earth.

Also, is it a good idea to use say 10cm lengths of shielded cable at the input and output of the flyback SMPS?. That is, shield the 10cm of mains cable coming to the Flyback maisn input, and the 10cm of dcdc output cable going from the flyback output. I mean surely if you can keep the high frequency radiation off the cable near the SMPS PCB, then you keep it off the whole cable?
The metal box itself is an excellent shield; it may not, however, be very effective to shield the noise from one part to another within the box and the energy transfer between the components as noise. But the metal box has some effect (actual calculation I do not have).

But certainly the SMPS module produces more noise than the amount that comes in from the power lines: but you have not explained your idea of the origin of the noise that you are trying to save the SMPS?
--- Updated ---

By the way, it is convenient to treat noise as temperature. And it works well most of the time.

Higher noise corresponds to higher temp. And heat flows from high temp to low temp. Noise current should flow in a direction so that higher noise corresponds to higher potential. The net result will be higher noise will be flowing out of the box. And unless it is an infinite source, it will keep on dissipating.

So to tame a source with high noise, couple it with a sink of low noise. And pray that something desirable happens. (it works mostly).

Many amplifier noises are reported in temp. And there is a method in the madness. Equivalent noise can be considered as a thermal source (under equilibrium).

Sometimes we can pump noise out but that needs lots of effort.
 
Last edited:
  • Like
Reactions: treez

    T

    Points: 2
    Helpful Answer Positive Rating
Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top