1
\$\begingroup\$

I have the following problem: I am currently debugging an electronic sensor board which is connected to the USB port of a laptop. The USB port is used both for communication via UART and to supply power to the sensor. The sensor shows some erratic behaviour and I want to check whether that is caused by an insufficient power supply of the USB port.

Therefore I want to connect the sensor board to an external lab supply like this:

schematic

simulate this circuit – Schematic created using CircuitLab

Since the laptop is not gonnected to mains ground (MGND) I would be connecting the floating laptop ground (LGND) directly to the mains ground (MGND). My question is whether that is okay or do I need to ground the laptop to be sure?

\$\endgroup\$
1
\$\begingroup\$

It is typically safe to connect the laptop chassis to mains earth. If it isn't already in the power brick.
Any peripheral might also ground chassis to mains earth, or create leakage current.

Just be aware of ground loops.


To prevent all of these power and noise problems I have a collection of USB isolators, and an isolated USB HUB*

*this one is mains earth referenced though (bb-elec UHR204)

\$\endgroup\$
  • \$\begingroup\$ Thank you for response, I´ll keep the points you mentioned in mind. \$\endgroup\$ – Mantabit Oct 24 '18 at 15:32
0
\$\begingroup\$

Laptop SMPS May be earth grounded on primary side to reduce egress, but they cause massive ingress with 100pF Torroid coupling to CM noise and worse with another primary ground source. Normally for a laptop with it’s ground plane , no problem but then cables become antenna. Ethernet has a DM and CM choke, DC and external mic’s may have a ferrite sleeve, (CM choke) but is not enough.

Due to leakage safe concerns of AC line with lightning this is how it’s done to limit Y cap leakage to 125 to 250uA.

Yet anytime you connected VGA monitor which has an earth grounded supply to signal gnd, when you connect to a laptop and it also becomes earth grounded.

So to satisfy both the SMPS noise + line + lightning noise to sensors, I highly recommend 1 to 10 nF to earth ground to 0Vdc or nearest Agnd. This because you cannot define to us how ground current nor CM transient current from impulse will flow.

The longish cables ought to be STP and the capacitance ratio of what you add 1nF ~ 10nF to ferrite coupling capacitance xx pF determines the best case amount of ingress + egress attenuation.

I have an excellent paper on What is CM and how answer every question on this subject ,but cannot post it.

Fully Understanding CMRR in DAs, IAs, and OAs Pete Semig Analog Applications Engineer-Precision Linear. ti.com

\$\endgroup\$

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.