在线咨询
eetop公众号 创芯大讲堂 创芯人才网
切换到宽版

EETOP 创芯网论坛 (原名:电子顶级开发网)

手机号码,快捷登录

手机号码,快捷登录

找回密码

  登录   注册  

快捷导航
搜帖子
查看: 5217|回复: 7

[转贴] Tie-hi/Tie-low cells for ESD protection

[复制链接]
发表于 2011-10-24 15:37:49 | 显示全部楼层 |阅读模式

马上注册,结交更多好友,享用更多功能,让你轻松玩转社区。

您需要 登录 才可以下载或查看,没有账号?注册

x
本帖最后由 littlej 于 2011-10-24 15:48 编辑

http://www.edaboard.com/thread63856.html

The historical use of Tie-Lo and Tie-Hi cells is almost as much emotional as it is secular.

The reason I say this is that some people use them religiously without  thought as to why.  Depending on your process and your design they may  or may not be necessary. Most people do not know why they are used.

ESD and Reliability are the right answer.  In some processes, the gate  oxide is very delicate and sensitive relative to the voltage levels of  the chip.  That means for any node with a gate tied to a low impedance,  such as a GND or VDD, the voltage on the gate is fixed...but what  happens if the voltage on the drain or source experienced a surge, over a  short period of time, well after enough surges, your oxide reliability  fails.  Generally these surges are fast impulses, either ESD or ground  bounce or some other fast transient impules, because if it was DC...then  the chip would be operating outside the limits of the process.

So how does the tie-lo/Hi work, it works by creating a DC level path but  a high impedance AC path on the gate oxide, this allows the voltage  level on the gate to spike up or down, with voltage surges on its drain  and or source, and even though these voltage spikes are capacitively  divided between all the nodes, because the gate voltage is allowed to  follow or track surges on drains/sources, than the voltage across the  delicate oxides are kept within more tolerant levels than if the gate  had been hard tied to a low impedance GND/PWR.

This is particularly critical on CDM (charge Device Model) ESD type events for IC's.

For almost this exact reason, you see a lot of 65nm and 45nm (even some  130 and 90nm) process that do not allow LVT decoupling caps with oxides  tied directly to a power or ground terminal (gate leakage problems  aside...though that is also a factor).

They are not always needed and do tend to take up more area.  Know your  process, your design and the conditions for your design to determine if  you need them or not...when in doubt however, I would recommend using  them.

As for how they are designed, there are many formats, the most common  being a large resistor in series with the gate, others involving diodes  or secondary transistors, etc.

SRFTech
发表于 2011-10-24 18:03:28 | 显示全部楼层
good paper , illustrate the internal machenism of tieoff cells ,

normally tieoff cells insertion is a good practise to protect your ESD ,

it is a must for design under 90nm ,

if not , erc checks will show 'tie to gnd/vdd directly' errors
发表于 2011-11-1 09:09:47 | 显示全部楼层
It seems new
发表于 2011-11-1 09:21:37 | 显示全部楼层
So, the Tie cell is used for ESD and Reliability.
发表于 2012-1-4 12:48:13 | 显示全部楼层
Thank your information.
发表于 2012-1-4 17:27:30 | 显示全部楼层
Thanks very much.

And I have one question about tie cell:
  if we have do EDS protection on IO power/ground cells, is it necessary to insert tie cells in core area?
发表于 2016-9-7 09:16:01 | 显示全部楼层
Tt is clear ,ths very much
发表于 2017-3-9 15:09:25 | 显示全部楼层
it's useful
您需要登录后才可以回帖 登录 | 注册

本版积分规则

关闭

站长推荐 上一条 /2 下一条

小黑屋| 关于我们| 联系我们| 在线咨询| 隐私声明| EETOP 创芯网
( 京ICP备:10050787号 京公网安备:11010502037710 )

GMT+8, 2024-4-26 21:01 , Processed in 0.035994 second(s), 9 queries , Gzip On, Redis On.

eetop公众号 创芯大讲堂 创芯人才网
快速回复 返回顶部 返回列表