| Author |
Message |
tdf
Joined: 27 Jun 2007
Posts: 6
|
 08 Jul 2008 10:31 Current vs Width of wire. |
|
|
|
Can anyone show me the width of a metal layer that drive a current: 50mA ?
I use tsmc 0.35 process.
Or can show me which material mention about it?
Thx,
tdf.
|
|
| Back to top |
|
 |
needforspeed
Joined: 21 Dec 2005
Posts: 20
Helped: 1
Location: FRANCE
|
| 08 Jul 2008 13:20 Re: Current vs Width of wire. |
|
|
|
normally, you can find current density rule in DRC document.
For metal of CMOS035(not top metal), 1mA/um at 70degree usually. I mean unidirection current.
|
|
| Back to top |
|
|
k_90
Joined: 25 Jan 2006
Posts: 97
Helped: 14
Location: Scotland,GB
|
| 08 Jul 2008 14:35 Re: Current vs Width of wire. |
|
|
|
depend how long the paths is.
work out the number of squares and multiplay it by the sheet resistance.
eg
a 2mm long, 1um wide supply rail carrying a current of 1mA with a sheet resistance of 0.03ohms/square.
Numer of squares 2000 (2mm/1um)
total resistance of interconnect: 60 ohms (2000sq x0.03ohms)
Total voltage drop 60mV (60ohms x 1mA
|
|
| Back to top |
|
|
jecyhale
Joined: 19 Feb 2008
Posts: 402
Helped: 39
Location: China
|
| 09 Jul 2008 5:28 Re: Current vs Width of wire. |
|
|
|
| tdf wrote: |
Can anyone show me the width of a metal layer that drive a current: 50mA ?
I use tsmc 0.35 process.
Or can show me which material mention about it?
Thx,
tdf. |
Usually, the current density is close to 1mA/um, but there should be a margin.
And I suggest you using more metal layer as you can.
|
|
| Back to top |
|
|
tdf
Joined: 27 Jun 2007
Posts: 6
|
| 09 Jul 2008 8:29 Re: Current vs Width of wire. |
|
|
|
Thanks to all your advice.
tdf,
|
|
| Back to top |
|
|
AdvaRes
Joined: 14 Feb 2008
Posts: 561
Helped: 27
|
| 09 Jul 2008 10:54 Re: Current vs Width of wire. |
|
|
|
| k_90 wrote: |
depend how long the paths is.
work out the number of squares and multiplay it by the sheet resistance.
eg
a 2mm long, 1um wide supply rail carrying a current of 1mA with a sheet resistance of 0.03ohms/square.
Numer of squares 2000 (2mm/1um)
total resistance of interconnect: 60 ohms (2000sq x0.03ohms)
Total voltage drop 60mV (60ohms x 1mA |
Provided that the height of the wire is not given I think that method is for a specific technology.
Can you tell us what is the sheet resistance of a 65 nm tech wire ?
|
|
| Back to top |
|
|
k_90
Joined: 25 Jan 2006
Posts: 97
Helped: 14
Location: Scotland,GB
|
| 09 Jul 2008 13:10 Re: Current vs Width of wire. |
|
|
|
the sheet resistance can normally be found in the design rules which takes into account the thickness of the metal/poly.
example 0.35u tech metal1 85m ohms/sq
toplevel metal4 ( thicker metal) 50m ohms/sq
|
|
| Back to top |
|
|
lylnk
Joined: 12 Sep 2005
Posts: 33
Helped: 4
Location: China Beijing
|
| 09 Jul 2008 16:00 Current vs Width of wire. |
|
|
|
tsmc has a document about reliablity.
it inculde EM, HCL...
1mA/um for normal metal, 1.5mA/um for top metal.
You can use several metal layers together.
|
|
| Back to top |
|
|
test_out
Joined: 10 Feb 2006
Posts: 132
Helped: 23
|
| 27 Aug 2008 4:32 Current vs Width of wire. |
|
|
|
| I am not sure for 0.35 tech, but for 0.18 tech, the wide of metal is limited (35u). So, for 0.18 tech, and in the case of carrying 50mA, I think you should divide the current to 2 ways.
|
|
| Back to top |
|
|
leebluer
Joined: 20 Apr 2004
Posts: 59
Helped: 2
|
| 28 Aug 2008 4:40 Re: Current vs Width of wire. |
|
|
|
Firstly, please confirm it is DC current or Transient current.
Then, find the value in the process file (maybe DRC rule). In the rule, only it is labed with DC rating.
|
|
| Back to top |
|
|
