Hello Group, We are having a debate concerning the best practice for grounding of a printed circuit board containing digital logic. These boards are multi-layer with a ground plane and a power plane.
One school of thought is to tie the ground plane to chassis ground in many locations, thus reducing the impedance. Another school of thought says to control the point(s) that is (are) tied to ground or risk upsetting of sensitive circuits with an ESD or other immunity event. The concept is that an ESD event may be decoupled to chassis at the I/O ground plane with the use of appropriate circuit elements to control impedances. Now consider the chassis to be steel, and the digital ground plane to be copper. If the digital ground plane is stitched to chassis in several locations, it appears that a lower impedance path (copper vs steel) will encourage the ESD to travel across the ground plane. If the ESD travels across the digital ground plane, there appears to be a good chance of upsetting sensitive circuits. So the thought might be to tie only one point of digital ground to chassis ground, thereby not providing a path for any immunity event to flow across this ground plane. The rest of the above concept is to use moats to segregate key circuits -- digital, I/O, analog, switch-mode power supplies. Again, some say to keep the ground plane in tact to provide the lowest impedance reference possible, so isolation is provided by carving up the power plane. The alternate approach is to "carve all the way through", i.e., if you have a moat around a particular circuit, if you are going to isolate, do it for all planes (stack, do not overlap). This latter approach, however, carves up the ground plane which would appear to increase the impedance of the overall ground reference. The argument is that carving up the ground plane is justified by eliminating the coupling of "dirty ground" to other circuits in an overlap situation. I would like to hear what you do for pcb grounding and why you do it. Don Umbdenstock Sensormatic --------- This message is coming from the emc-pstc discussion list. To cancel your subscription, send mail to majord...@ieee.org with the single line: "unsubscribe emc-pstc" (without the quotes). For help, send mail to ed.pr...@cubic.com, jim_bac...@monarch.com, ri...@sdd.hp.com, or roger.volgst...@compaq.com (the list administrators).
