The Designer's Guide Community Forum

The Designer's Guide Community Forum
https://designers-guide.org/forum/YaBB.pl
Design >> Analog Design >> antenna diode
https://designers-guide.org/forum/YaBB.pl?num=1258286510

Message started by Dipankar on Nov 15^th, 2009, 4:01am

Title: antenna diode
Post by Dipankar on Nov 15^th, 2009, 4:01am

Dear All,
To get rid of antenna violation I use antenna diode (n+ diode - n+ implant in p-sub) in the long metal lines connected to MOSFET's gate and it works. But the reverse break down voltage of the diode is ~11.5 V. So it means if some static charge gets developped in the metal line the voltage can go upto ~11.5 V before the diode provides a short. Now I wonder how the use of diode passes the antenna rule because 11.4 V should be enough to damage the gate-dielectric ( gate thickness 4 nm ) . Please tell where the logic goes wrong.

Title: Re: antenna diode
Post by Mayank on Nov 15^th, 2009, 9:43pm

Hi Dipankar,
As far as i remember, Antenna Diodes work on forward bias voltage...What i mean to say is that antenna diodes prevent Gate voltage to exceed supply voltage by more than Von(0.4~0.7V) of the Antenna Diode...Not the Reverse BreakDown Voltage of the diode....

Consider this :-- You tie the Gate of a MOS at +ve-jnxn and Vdd at -ve-jnxn of the diode....In normal operation, it is reverse biased...In case, charge got accumulated at the gate and Gate Votlage exceeded the supply voltage by more than Von of the Diode, Immediately diode will become ON and would facilitate charge flow from gate bringing it back to <=supply voltage. Sounds LOGICAL, right ??
Maybe some ESD expert can explain better...

--regards,
mayank

Title: Re: antenna diode
Post by Mooraka on Nov 16^th, 2009, 9:02am

Dipankar,
I think you have a valid point here for the inversion region. I see some options in the literature for this case such as using a gated diode instead of simple diffusion diode. See the link below.
http://www.google.com/patents/about?id=ZzMEAAAAEBAJ&dq=Method+and+apparatus+for+protecting+gate+oxide+from+process-induced+charging+effects

Other ways are to maintain safe gate/metal area ratio and use of vias to shorten the single metal lengths.

Ram

Title: Re: antenna diode
Post by loose-electron on Nov 16^th, 2009, 9:13am

If you cant suffer the use of antenna effect diodes in the design the trick of breaking the metal line up into pieces and connectining it together in other layers does work pretty well.

Title: Re: antenna diode
Post by Mooraka on Nov 16^th, 2009, 9:17am

On the same point, use of simple diffusion diode will also suffice some times even though break down voltage is high. The accumulated charge on the gate will transfer to the substrate as diode leakage current in reverse bias.

ram

Title: Re: antenna diode
Post by Dipankar on Nov 16^th, 2009, 9:30am

Mooraka wrote on Nov 16^th, 2009, 9:17am:

Dear Mooraka,

Have you found any material supporting what you are saying ?

Title: Re: antenna diode
Post by Mooraka on Nov 17^th, 2009, 7:46am

None yet. But I think it is true as I still see use of simple antenna diode even at scaled technologies down to 65nm.

Ram