Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Why are double guard rings used.

Status
Not open for further replies.

krrao

Junior Member level 2
Joined
Feb 13, 2006
Messages
24
Helped
0
Reputation
0
Reaction score
0
Trophy points
1,281
Activity points
1,517
Hi All,
Why are the double guard rings used? For a PMOS driver layout, what if I use a single guard ring? What are the advantages of using double guard ring?
--
warm regards,
krrao.
 

Guard rings are used to prevent injection of charge carriers from the substrate to active devices through latchup.
Since latch up requires two cross-coupled BJTs (i.e. creating a parasitic thyristor), the idea of the guard rings is to provide
an alternative collector or emitter for the parasitic devices to latch up to.
Since there are two parasitic devices involved (NPN and PNP), in situations were both PMOS and NMOS are used,
we want the regular thyristor, made of NPNP (this is NPN and PNP interleaved), to separate to NPNPNPNP and this combats
the mutual beta factor boosting happening in the thyristor. For this we added both a P and N guard ring.

Hope this helps.
 
  • Like
Reactions: krrao

    krrao

    Points: 2
    Helpful Answer Positive Rating
Guard rings are used to prevent injection of charge carriers from the substrate to active devices through latchup.
Since latch up requires two cross-coupled BJTs (i.e. creating a parasitic thyristor), the idea of the guard rings is to provide
an alternative collector or emitter for the parasitic devices to latch up to.
Since there are two parasitic devices involved (NPN and PNP), in situations were both PMOS and NMOS are used,
we want the regular thyristor, made of NPNP (this is NPN and PNP interleaved), to separate to NPNPNPNP and this combats
the mutual beta factor boosting happening in the thyristor. For this we added both a P and N guard ring.

Hope this helps.


Hi eladla,
Can you please elobarate the 2nd part of your reply. "we want the regular thyristor, made of NPNP (this is NPN and PNP interleaved), to separate to NPNPNPNP and this combats
the mutual beta factor boosting happening in the thyristor. For this we added both a P and N guard ring."
 

Ok... here we go :)
A thyristor is a NPN and PNP BJTs cross-couled.

As you can see from the image, this creates a NPNP block, where the base of the NPN is the collector of the PNP and vise versa.
The problem is that when the NPN and PNP BJTs are both conducting they keep each other in saturation.
beta = Ic/Ie and because they are coupled like they are, they keep each other's beta factors up and keep the current flowing.
If we add two guard rings, one N and one P, we get (I wrote this wrong in my first post) NPNNPPNP.
So now we have NPNN and PPNP. The BJTs are not cross-coupled any more and if one of them happens to be forward-biased in some circumstance,
the current diminishes because the beta factor is small and there is no positive feedback.

Hope this is clearer now.
 
  • Like
Reactions: krrao

    krrao

    Points: 2
    Helpful Answer Positive Rating
Ok... here we go :)
A thyristor is a NPN and PNP BJTs cross-couled.

As you can see from the image, this creates a NPNP block, where the base of the NPN is the collector of the PNP and vise versa.
The problem is that when the NPN and PNP BJTs are both conducting they keep each other in saturation.
beta = Ic/Ie and because they are coupled like they are, they keep each other's beta factors up and keep the current flowing.
If we add two guard rings, one N and one P, we get (I wrote this wrong in my first post) NPNNPPNP.
So now we have NPNN and PPNP. The BJTs are not cross-coupled any more and if one of them happens to be forward-biased in some circumstance,
the current diminishes because the beta factor is small and there is no positive feedback.

Hope this is clearer now.

Hi eladla,

It was confusing for me in the previous reply when you said NPNPNPNP. Thanks for the clarification. It answered my question.
 

You are looking to kill the base of both transistors in
the SCR. Because it is the hFE product that matters,
and leaving one weakly tied still holds the possibility
of the other lighting it up.

It also helps to extract any substrate noise currents
as close to their point of origin as possible. Ringing the
well alone, leaves the well as a noise injector (in the
case that the well is tied to source, and source to
supply through a nonzero resistance, nonzero inductance
and hard-switching current, for example).
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top