请教“天线效应”
我本人是做电路的。 对这方面也是不是很清楚,所以请各位大侠指点
简单解释一把:天线效应是说如果Poly在版图中走得过长(几十um)并且被包在场氧(Poly和silicon之间)和Nitride(Poly和Metal 1之间),会增加收集电荷的几率。(氧化层和氮化层中的电荷是由于制造工艺不能保证高纯度造成的)。这些被收集了的电荷被Poly收集后由于获得了更低能级,很难再向高能级(氧化层或氮化层)跃迁。从而Poly的电位就会由于电荷而增加或减小,Gate Poly下的沟道也会被永久开启或者永久关闭甚至是击穿栅氧。于是对芯片造成误触发或者是永久性的伤害。
解决这个问题的办法是找到一个比Poly更低能级的材料,吸收Poly中的电荷。很显然,芯片中比Poly能级更低的只能是金属,所以版图级别的解决方法是尽量画短Poly,不用Poly当做连线;当Poly不得不画得太长的时候(比如说反管),需要在Poly上再借足够长的Metal,任何一层的Metal都可以。当然,如果通过Metal1连接到Metal2,最后连接到Top Metal,这样电荷会跑到芯片最顶端,很难开启氮化层和场氧下面的silicon,这样最安全了。
回答不错
同意二楼的解释
还有一种方法是用二极管消除
谢谢二楼的兄弟 说的好 要是能写一篇 关于天线效应的文章 就最好了
这个解释还是第一次听说,新奇。
天线效应razavi的书上有讲啊,看看就知道了。
还有模拟版图艺术上面也有。
很好很好。
天线效应貌似不是这样的吧
不单是大片POLY,大片金属如果和MOS 栅直接相连,也会引起天线效应
将栅氧化层烧掉
用跳线就行了
同意8楼的,能收集电荷产生天线效应是POLY和与POLY相连的金属啊,与二者面积有关,因为栅的制作是用的干法刻蚀,此工艺中有较高电压,多晶硅栅会收集电荷,造成MOS击穿。
同意8楼的,若金属层面积超过规则,则要向高层跳线才可以,若往下一层跳线的话是没有意义的
呵呵,学到知识了,谢谢!
往下跳也可以吧,天线效应是出现制造过程的,如果将poly断成两断,中间用top metal连起来其实在中间过程是应该没有天线效应的
往下跳是不行地。
你物理学的不错,不过你的解释是天方页谈,根本不正确,理解有很大偏差,简单的东西被你复杂化
往下调等于没跳
2楼的回答不知所云
往下跳线确实等于没跳
学习了、、、
往下跳线为什么没用呢?
原因:大面积单一metal/poly
跳线:多块小面积metal/poly
难道是考虑到制造过程中由下到上的顺序,高层产生的时候低层的已经存在?
还是不清楚……
二楼的,
结果:不用长poly
原因之一:poly电阻
嗯,确切来说用短的高层来跳
理解了
这话比较贴切。
Explanation in English:
(though in English, has all the details you want to know.Please bear with me. ...
During the layout, a jump using higher interconnect layer means to to cut the length of lower (interconnect) layer down to multiple segments.In that case, the interconnect area of lower layer / gate oxide area will be reduced (minimized), therefore no more rule violation.
For instance, the original layout has poly interconnect of 300 um x 0.6 um, and the thin oxide area is 0.18 um x 0.6 um.It is likely to cause rule violation (of antenna effect)in Area_poly / Area_tox.Break the poly interconnect as early as possible, say break it when it only wires out the gate by 10 um long, and jump up to M1, jump back to poly to continue the rest of the 290-um length, then connect it to a circuit output node.(Remember the next 290-um poly connects to diode junction, so does not cause antenna rule violation.So the rest 290-um poly does not cause antennal rule violation.)During the process step of poly, the "antenna" only consists of 10 um x 0.6 um poly area, instead of 300 um x 0.6 um area.This is why jumping up to higher interconnect layer helps resolve the antenna rule violation.In case, Area of (M1 + poly) / area of tox still causes rule violation, break M1 (to multiple segments), and connect the route by M2, M1 and poly.While processing M1, the antenna only consists of short M1 + short poly.Thus the issue of rule violation is resolved.
By induction, the resolution of antenna rule violation can be derived to "jumping to higher interconnect layer (as early as possible), all the way up to the top (metal) and jumping back down to where the antenna rule violation layer is."
天线理论与技术
天线理论与技术
天线理论与技术
天线理论与技术
天线理论与技术
天线理论与技术
天线理论与技术
天线理论与技术
天线理论与技术