Adding diodes to reduce the current loop area is a waste of time?

[treez]

Hello,
We have an Offline LED driver which also has a 24V, 3W output for connection of a radio receiver module. Also, the 24V gets converted (externally) to a 16V DALI bus, which then gets fed back onto the LED driver. The customer may accidentally feed the 16V DALI bus in the wrong way round. So therefore there is a Bridge rectifier inside the LED driver.
Unfortunately, you can see that the prescence of the bridge rectifier means that the ground path is broken by the diode of the rectifier bridge….therefore, the return path for the ground from the 16V external DALI module is actually the ground terminal of the 0-24V power supply output….this unfortunately means that there will be a wide area current loop. This can be seen in “DALI 1 schematic” as attached.

As you know, this is not a good thing to have at the output of a noisy offline Switch mode LED driver. (a wide area current loop)

Therefore, we have decided to add diodes to the 24V output, in order that the ground return of the 16V DALI bus follows the 16V DALI bus. (thus giving a narrower area current loop). This can be seen in “DALI 21 schematic” as attached

Do you think that this is a good idea? (adding the two diodes)...or are we just being overly pedantic?

 

[dick_freebird]

If this is all about EMI, might a good quality capacitor be
a better shunt? Strapping diodes will let the loop swing
stop to stop before conducting. Why would there be that
kind of voltage "noise" at baseband (other than shoddy
design and construction)?

Distributing capacitors could turn your big loop into N
smaller "HF AC shorted" ones.

 

[treez]

Thanks, we dont have access to any of the "external" stuff...that is what the customer would put in, and we wouldnt be able to add capacitors there.
As you can see, when we put the bridge in to handle reverse polarity, we then find that the "return" of the 16V line, wont follow the 16V line any more......due to the ground connection inside the unit as shown....this internal ground connection , as shown, unfortunately shorts out the bridge rectifier diode.

- - - Updated - - -

As you can see, there is a connection back to GND inside the unit, (via the return of the 0-24V connection), which unfortunately shorts out the lower rectifier diode of the bridge.
As such, i put in the lower diode in schem 2 to mitigate this...putting it in the 0V return of the 0-24V connection.

 

[CataM]

unfortunately shorts out the bridge rectifier diode.

Why unfortunately ?

Isn't that how it is supposed to work ?

The 2 diodes one will behave the same way as the first scheme (you have 2 Net-Ties instead of 1).