# asymmetric stripline

Discussion in 'Electronic Design' started by John Larkin, Jan 29, 2008.

1. ### John LarkinGuest

Given a pc board like...

======================================= ground
^
12 mils
v
======= trace ----------

^
24 mils
v

======================================= ground

and assuming FR-4 and 1 oz copper, does anybody know what trace width
gives 50 ohms? My rough guess is 12 mils.

googling online calculators yields an impressive collection of
defunct/missing/silly web pages. I suppose I could fire up ATLC, but
that's a heap of work.

Help appreciated.

John

2. ### Robert LacosteGuest

12mils gives 52ohm. 13mils gives a better 50.158ohm, see here under
simulation done with the (very good) GPL 2D impedance calculator software
from Kirby Goulet (cf http://sourceforge.net/projects/mdtlc/) :

Simulation pix map 400 pixels high by 1600 pixels wide.
1921024 bytes allocated for bmp.

INPUT PARAMETERS:

Layer Thick Er Specifications
Copper Plane Top 1.30 3.20 Open w=0.0 off=0.0
Laminate Layer 1 12.00 4.20 64.3% 3.2/6.0
Signal Layer 1 1.35 4.20 W=13.0 Etchback=0.00
Laminate Layer 2 24.00 4.20 64.3% 3.2/6.0
Copper Plane Bottom 1.30 3.20 Open w=0.0 off=0.0

Rdc_trace_1= 38.67 milliohms/in 20C
DC resistance by pixel count:
Rdc_trace_1= 37.291 milliohm/in.
C = 3.462 pF/in L = 8.709 nH/in.
Er = 4.200 Delay = 173.626 ps/in.
Zo = 50.158 Ohms

Friendly,
Robert Lacoste
www.alciom.com

3. ### john jardineGuest

Good link!. Looks like fun. Thanks.

4. ### John LarkinGuest

Thanks. I'd used Er of 4.5 for the dielectric, so my 12 mils guess is
pretty close, maybe even a bit below 50 ohms. This is for a test trace
that I'm piggy-backing onto a production board; I'll TDR it and post
the results. It will include various traces and vias, and do some
"reference plane" switching.

That program looks like a fancier version of ATLC. It still appears to
need an external bitmap generator. But there seems to be no HELP.

John

5. ### Tom BruhnsGuest

My recommendation: contact your board house. They (should!!) have
good calculation capabilities for the specific board material and
stackup you are dealing with, and can even tell you tolerances. I've
had very good service on this sort of thing from the couple of board
houses we use.

I do have a nice spreadsheet that will come close on things like this,
and on differential traces and various other things, but when push
comes to shove, the board house is my backup.

Cheers,
Tom