Quantcast
Channel: Cadence Custom IC Design Forum
Viewing all articles
Browse latest Browse all 4890

mtline 2D solver ignoring dlosstype/dloss

$
0
0

I have a mtline instance in a spectre netlist that looks like this.  Note dlosstype=tangent and dloss=[0.1] (so really lossy)

ITL (net1 net2 net3 net4 0 0) mtline len=100m m=1 fmax=10G linetype=stripline \
    modeltype=wideband numlayer=1 numgnd=1 er=[4.0] \
    layerthickness=[500u] linewidth=[200u] lineheight=[250u] \
    linethickness=[25u] linespace=[200u] linesigma=9e7 \
    dlosstype=tangent dloss=[0.1] gndthickness=[25u] gndsigma=9e7

In the RLCG file that spectre creates I see the following before the actual numbers.  Note that the dielectric loss settings seem to have made it but there is no "G" generated, just "RLC".

; The RLCG matrices produced by mtline 2-D field solver
; The Inputs are:

; lmgLineType = Stripline
; lmgModelType = LossyWide
; lmgNumLines = 2
; lmgDielectricPermittivity = 4
; lmgDielectricThickness = 0.0005
; lmgNumLayers = 1
; lmgDielectricLossType = tangent
; lmgDielectricLoss = 0.1
; lmgNumGndPlanes = 1
; lmgGndThickness = 2.5e-05
; lmgGndSigma = 9e+07
; lmgConductorWidth = 0.0002
; lmgConductorGaps = 0.0002
; lmgConductorThickness = 2.5e-05
; lmgConductorHeight = 0.00025
; lmgLineSigma = 9e+07
; lmgMaxFreq = 1e+10

FORMAT Freq: L1:1 L2:1 L2:2
R1:1 R2:1 R2:2
C1:1 C2:1 C2:2

am I missing something obvious?  Some other flag needed?

Thanks

-Dan


Viewing all articles
Browse latest Browse all 4890

Trending Articles