"Noisy" ground and the theory behind it.

Thread Starter

odm4286

Joined Sep 20, 2009
241
Hello everyone, a thought popped into my head today and I was hoping someone could tell me if I'm on the right track.

Can the sinking of a large amount of current generate a noisy ground? Especially at certain frequencies? My intuition tells me this causes ground to "bounce" as the current is sunk.

Also, I know that isolated dc-dc converters help rectify this by supplying an isolated ground. How exactly is an isolated ground physically different from the "normal" ground?

Hopefully my questions are clear enough. Thanks everyone.
 

danadak

Joined Mar 10, 2018
3,627
There is a lot written and dealt with concerning ground systems in the
same environment as sensitive analog circuitry and EMI/EMC consid-
erations.

Large currents, for example H Bridge Motor Drivers, Inverters for welding
and Plasma cutting, to name a few. Generally the approach is to split the
grounds and only join them at the source of power. To eliminate ground
drop and inductance.

Isolated ground -

https://www.samlexamerica.com/documents/application_notes/Input_output_grounding_of_isolated_DC_to_DC_Converters.pdf

http://www.ti.com/lit/an/slyt298/slyt298.pdf

http://www.manson.com.hk/getfile/index/action/product/name/5166517149dbb.pdf

Layout techniques -

http://cache.freescale.com/files/analog/doc/app_note/AN3962.pdf

http://www.analog.com/en/technical-articles/a-short-course-in-pcb-layout-for-high-speed-adcs.html

https://www.e2v.com/content/uploads/2014/09/Board-Layout.pdf

http://www.cypress.com/file/136286/download


Regards, Dana.
 

MrChips

Joined Oct 2, 2009
19,421
A perfect connection to ground would have zero resistance and zero reactance.
The world is not perfect and hence we have to live with something less than perfect.

If you have 10A flowing in 0.1Ω that will result in a voltage drop of 1V. This can be detrimental depending on specific circumstances.

You have to distinguish between a signal or supply return path versus a common connection to a zero volt reference point. Every power supply must have a current path from the supply and a return path back to the supply. Your circuit, whether it be analog or digital might require a zero volt reference point. It is important that the supply current does not interfere with the zero reference point.

In high current power supply design, it is common practice to monitor the supply voltage using a pair of sense wires attached directly at the load. In this way, the voltage drop appearing across the supply wires is taken into account in order to establish the correct voltage at the load and not at the supply.

Supply noise can be mitigated by paying particular attention to how you run your supply lines versus signal lines, particularly when laying out a printed circuit board. You have to make sure that the signal grounds are not part of the supply loop. This is one example of the infamous "ground loop" problem. You should never "daisy chain" signal and supply grounds. This would be best configured using a "star" connection.

In mixed signal applications, where there is a mix of analog and digital circuitry, one tends to use separate supply and ground planes in order to mitigate digital noise affecting sensitive analog circuitry such as high resolution analog-to-digital converters.

Ground loops can also be created when a signal cable is grounded at both ends. It is known practice to connect audio signals with a shielded cable with only one end of the shield connected to signal COMMON.
 

danadak

Joined Mar 10, 2018
3,627
One other consideration, capacitor technologies for noise suppression, and their
ESR performance. Not all caps of same capacity are "equal".



Regards, Dana.
 
Top