0
\$\begingroup\$

Below in the photo both SMPS power supplies powers some DAQ boards. I named them as 12V PSU1 and 5V PSU2. As you see the PSU1 and PSU2 are daisy chained at their mains entry(line, neutral, earth):

enter image description here

And here is the basic diagram:

enter image description here

Above both power supplies powers the different parts of the remote device Device 1. In real there are six of them but or simplicity I draw one device. The PSU1 directly powers the device, the PSU2 powers the 5V part of the device through a USB Hub.

I named the nodes as A, B, C, D and E.

Knowing that in SMPS A E and C D nodes are capacitively coupled, is there a loop for high frequency across A B C D and E?

If there is can I eliminate that by cutting the daisy chain earth between D and E?

Edit:

Block diagrams of the power supplies:

enter image description here

enter image description here

And here is the power line filter.

\$\endgroup\$
  • \$\begingroup\$ is it malfunctioning? \$\endgroup\$ – Jasen Dec 10 '18 at 20:12
-1
\$\begingroup\$

It seems to me that there is at least one ground loop between the line filter housing - the yellow cable - the mounting plate with housing - back to the line filter.

I suggest you disconnect the yellow cable from the line filter.

| improve this answer | |
\$\endgroup\$
  • \$\begingroup\$ Thanks for the input I didnt think about it. Here is the exact entry module line filter eu.mouser.com/datasheet/2/358/typ_5707-1275588.pdf Are their housing coonnected to their erath terminal in general? See also my edit I also provided the block diagram of the two SMPS. \$\endgroup\$ – user1245 Dec 8 '18 at 21:48
  • \$\begingroup\$ disconnecting the saftey ground is very rarely the answer. and certainly not the answer to an immaginary problem. \$\endgroup\$ – Jasen Dec 10 '18 at 20:07
-1
\$\begingroup\$

Yeah there's loops but they don't matter. USB signaling is primarily differential: so long as the ground is good it should all work fine.

if you suspect problems with the ground add more connections to ground.

Cutting ground connections so that fault currents would flow through USB cables has the potential to end very badly.

| improve this answer | |
\$\endgroup\$
  • \$\begingroup\$ " add more connections to ground" what does that mean? \$\endgroup\$ – user1245 Dec 9 '18 at 0:57
  • \$\begingroup\$ if something is misbehaving, ground it better. \$\endgroup\$ – Jasen Dec 9 '18 at 1:15
  • \$\begingroup\$ @Jasen That is a terrible advice. Ground loops have to be avoided in first place, which is why more ground is not always better. Additionally, when there is a problem related to galvanic coupling, one should remove the common ground path of two interfering devices. \$\endgroup\$ – Stefan Wyss Dec 10 '18 at 5:14
  • \$\begingroup\$ Oooh scare ground loops! if that weere the case gound planes would never be used - they provide for an infinite number of ground loops. \$\endgroup\$ – Jasen Dec 10 '18 at 20:09
  • \$\begingroup\$ @Jasen A plane is not the same as a GND loop. There‘s no induction/EMF happening within a plane. E.g. I‘ve never seen an inductive coil of an RFID system being replaced by a plane. \$\endgroup\$ – Stefan Wyss Dec 10 '18 at 20:56

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.