Diff-Complete Shield

If there is an EMI problem, one of the methods to correct the problem is to shield with ground patterns. However, in most complicated designs, a perfect shielding is not easy to accomplish. Therefore, designers should carefully design grounds to achieve the best shielding effect not only for a single net but also for a differential pair net. Differential pair nets are employed to resolve SI noise, return path, PI, EMI issues that appear often in high speed signal nets. While differential pair nets take care of return path and impedance issues by themselves by being side-by-side with uniform distance apart, noise from other nets may disturb the differential pair nets, if they are not properly shielded, then prevent the paired nets do the intended roles. Therefore, it is important to shield the differential pair nets as much as possible. In order to shield signal nets or traces on the same layer, shielding grounds cover around the signal traces with a certain uniform clearance. However, for differential pair nets, shielding grounds should not present between the two pair nets. As a rule of thumb, it is recommended to shield at least 80 percent of signal net length. This item checks whether a design meets a specified shielding ratio.
  • Item: Input item name.
  • Check Type: Define the shield checking method of net.
    • Separate Net: PollEx DFE detects existence of shielding net(s) within specific distance per every segment. The DFE regards only the portion of shielding net which is parallel with target differential net.
      Figure 1.

    • Merged Net: The DFE extends the searching area with the same amount of specific distance, then PollEx DFE detects existence of shielding net(s) within that region. This method is more accurate than the Separate Net method, but it takes more time.


      Figure 2.
  • Net: specify a differential net group.
  • Filter: Enter a filter to choose differential pair net from selected net group.
  • Shield Net: High speed signal will be shielded with nets in this Shield net group.
  • Component Keep IN/OUT: Define test region. For example, you can test inside or outside the breakout region of CPU with different test values.
    • Component Group: Select the required component group which is used for defining the test region.
    • Region: Define the required test region.
      • IN: Test inside of the breakout region of target component.
      • OUT: Test outside of the breakout region of target component.
  • Range (COC+distance): Enter the distance value to define breakout range. The COC (Component Overlap Check) plus this distance value is considered the test range.
  • H-Shield Check: Option for horizontal shield check.

Horizontal Shield Ratio Check

  • Shield Distance: Specify distance from differential pair net to ground.
    • Use Width Value (W): Upon selecting this option, the value in Maximum Shield Distance is considered as multiple of Pattern Width.
    • Max: Specify maximum distance from differential pair net to ground. The DFE searches shielding net within this range.
    • Min: Specify minimum clearance from net to ground. If shield net exists closer than this clearance, it reports as a fail.
  • Min Horizontal Shield Ratio(%): Minimum shielding net length ratio against total net. If the shielding ratio is smaller than given value, it is reported as failed.

    Ratio = The total length of shielded portion of net/Total length of net

  • Shield Ratio Calculation Type: Determine shield ratio calculation method.
    • All Layer: Calculate shield ratio for total net.
    • Per Layer: Calculate shield ratio per each layer.
  • Exclude/Include Option:
    • Pin Escape: Enter a radius of circular region around pins to be excluded for the rule check.
    • Via Escape: Enter a radius of circular region around vias to be excluded for the rule check.
    • Exclude Short Pattern: Exclude the short length trace shorter than this value.
    • Include trace corner: The ground shield length at which signal bend is always longer than signal trace length. This leads to results that are not accurate. In order to eliminate this problem, the DFE ignores the bending position. The DFE includes the bending position when calculating shield ratio.
    • Include Inner VIA Shielding: In the case of Inner Layer VIA where the net is not connected, check the shielding.

Guard Via Distance Check

The DFE check maximum via distance of shielding net.
  • Guard Via Range: Define the maximum distance from trace edge within which guard via should be exist. The DFE considers only the vias within this range as a guard via.
  • Max Via Distance: If the via distance is greater than this value, it is reported as failed.
  • Except PAD: Exclude pin pad for testing.
  • Merge Result Via Count: PollEx DFE shows the merged result of all errors on the same display.

V-Shield Check

Option for vertical shield check.
  • Vertical Shield Ratio Check:
  • Shield Distance (Extended V-Shield Width): Specify minimum required GND width of upper/below layer.
  • Min Vertical Shield Ratio(%):Minimum ground shielding ratio. If the shielding ratio is smaller than the given value, it is reported as failed.

    Ratio = Shield Area/Net Area

  • Shield Ratio Calculation Type: Determine test layer.
    • All Layer: Calculate shield ratio for total net.
    • Per Layer: Calculate shield ratio per each layer.
  • Exclude/Include Option
    • Check Layer:
    • L1 Layer: check one layer above and below.
    • L2 Layer: check two layers above and below.
    • L2-Hop Layer: check two layers one layer apart above and below.
    • Pin Escape: Enter a radius of circular region around pins to be excluded for the rule check.
    • Via Escape: Enter a radius of circular region around vias to be excluded for the rule check.
    • Exclude Short Pattern: Exclude the short length trace shorter than this value.
    • Consider both upper and lower direction: Upon selecting, above and below layer will be tested if not wider layer side will be checked (above or below)
The flux radiated from each of differential signals are canceled each other because of their polarity is reverse. But it still needs ground shielding for better EMI and signal return.


Figure 3.

Although crosstalk noise on a differential pair has a significant common-mode component, it also has a differential component because the distance between an aggressor and each side of the pairs differs. Consequently, each leg will experience slightly different crosstalk, which will not be rejected by the differential amplifier. The signal swing of differential signal is so small this crosstalk can cause failure. Ground shielding is needed for differential lines.

High speed signal’s 100 percent ground shielding is perfect, but it is impossible. Over 80 percent and shielding for all routed layers are recommended.


Figure 4.