METAL STRESS RELEASE RULE - TSMC18RF

M

michael_skoufis

Guest
In the tsmc18rf technology, slots must be placed for metals larger (at
least one dimension) than 35 um. Does anyone know how these cuts are
defined on the metal? I tried different layers but I cannot find one that
defines an open slot. Does nayone have a clue? Thanks in advance.

Mike
 
On Sat, 21 Jan 2006 15:57:10 -0500, "michael_skoufis"
<skouf1974@nospam.yahoo.com> wrote:

In the tsmc18rf technology, slots must be placed for metals larger (at
least one dimension) than 35 um. Does anyone know how these cuts are
defined on the metal? I tried different layers but I cannot find one that
defines an open slot. Does nayone have a clue? Thanks in advance.
Click to expand...
The cuts are not represented on another layer, you have to make a hole
in the metal polygon itself.
 
Thanks for the help. I suspected that's how you do it after trying every
single layer.
 
According to the original TSMC document
T-018-LO-LE-004 this should be METALn GDS Layer Number; GDS Data Type '2',
Layer Name METALn slot.

TSMC does a logical operation then with these two layers before mask
making.

As far as I know the RF option at TSMC is always a merge of the logic
process plus the additional rf options, so refer to the TSMC logic documents
If you can't find any useful information in the RF ones.

Are you using an original TSMC kit or some modified one like MOSIS or
maybe an in-house kit, then there can be differences.

Bernd


michael_skoufis wrote:
In the tsmc18rf technology, slots must be placed for metals larger (at
least one dimension) than 35 um. Does anyone know how these cuts are
defined on the metal? I tried different layers but I cannot find one that
defines an open slot. Does nayone have a clue? Thanks in advance.

Mike
Click to expand...
 
I am using the original TSMC kit. Thanks for the suggestions. Since there
is an assigned GDS layer name and type, the METALn slot should be included
in the overall layer list. Problem is that no layer works. The closest
one is "METALn st". However it didn't generate any cuts on the metal
itself. I'll continue to look into it. I appreciate it.

Mike
 
"METALn st" should be the right one, stream it out to GDSII and look at the
log file, usualy PIPO.log, if you find the GDS layer datatype information of
"METALn st" there.

The layer itself or the layout tool does not generate any slots if you layout
metal and metal slot layers over each other. This is done by your silicon vendor
TSMC in a logical operation of METALn layer datatype 0 and METALn layer datatype
2 before mask making, if you GDS database contains both.

Bernd

michael_skoufis wrote:
I am using the original TSMC kit. Thanks for the suggestions. Since there
is an assigned GDS layer name and type, the METALn slot should be included
in the overall layer list. Problem is that no layer works. The closest
one is "METALn st". However it didn't generate any cuts on the metal
itself. I'll continue to look into it. I appreciate it.

Mike
Click to expand...
 
I'll check PIPO.log after streaming it to GDSII and see what I get. The
problem is that DRC is not free even when I use a few "METALn st"
rectangles on a big chunk of metal. Since I was getting DRC errors I
thought I should fix that first before streaming out to GDSII. Thanks a
lot for your help.

Mike
 
Indeed METALn slot should be know and covered by the
DRC command rule file, can you check this as well.
If not I recommend to slot the metal manualy as 'mk'
already suggested.

Bernd

michael_skoufis wrote:
I'll check PIPO.log after streaming it to GDSII and see what I get. The
problem is that DRC is not free even when I use a few "METALn st"
rectangles on a big chunk of metal. Since I was getting DRC errors I
thought I should fix that first before streaming out to GDSII. Thanks a
lot for your help.

Mike
Click to expand...
 
You must log in or register to reply here.

Welcome to EDABoard.com

Sponsor

Back
Top