Floating copper heatsink inside an offline PSU?

[cupoftea]

Hi,
We purchased an offtheshelf 750W Offline isolated PSU. The mains input rectifier is the 4-pin, plastic packaged GS1B 2560. Its metal cased and fan cooled.
The mains bridge is mounted flat, actually suspended above the mains input EMC filter. It is "sandwiched" between two 1.5mm thick copper plates which form its heatsink. one of these copper plates completely covers over the top part of the mains input EMC filter section. One copper plate is approx 4cm by 8cm and the other some 4cm by 4cm. These copper plates are both floating...not connected to any node of the circuit.
Is this OK from an EMC perspective? (ie, floating metal above a mains EMC filter).
The actual metal case of the PSU is connected to earth. The PSU is obviously open-grilled at each end to allow for air flow from outside.

 

[dick_freebird]

I'd think that a floating plate intimately connected to
switching devices (mechanically, if not electrically)
would be mysterious at best, capacitively coupled
to busy bits and hence an EM radiator most likely.

What about putting high voltage, low ESL/ESR caps
from these plates to the case ground, to at least
shunt-divide any such HF that may be riding on the
plates?

Have you put a "sniffer" to the plates to look for
EMI magnitude? That would tend to answer, or at
least bound the problem.

 

[Easy peasy]

the metal case of the BR is isolated - so the floating "heatsinks" are not an issue here - the capacitance to the mains diodes is quite low - and, you will find the thermal resistance to the diodes is quite high in this particular construction - as they rely on the resin - and it ain't that good ...

If the heatsinks were earthed there would be a slight extra capacitive path from the mains diodes to earth which may or may not increase the EMC signature ... as seen by a LISN ...