Connecting an OTA to pad without buffer

[hyleeinhit]

Hi, folks,

I want to know the effect of connecting an OTA directly to PAD without buffer. The OTA is folded cascoded structure as shown in attachment. Transistors N2 and N3 provide 500nA bias current, respectively. This OTA output is also connecting to a transistor gate in the following stage.

I need to connect this OTA directly to PAD, but I am worried about the effect from the pad leakage current, pad parasitic capacitance and bonding parasitic capacitance and inductance.

Two options of PADs:
1. PAD with ESD protection. Leakage current exists with such a PAD. I have a PAD protected by a 333um2 area diode. How much is the leakage current of PAD. Can my OTA drive it?
2. PAD without ESD protection. No leakage current exists but having risk of breakdown. Anyone recommend such a PAD? FYI, I am a student and not developing a product.

Effect from pad parasitic capacitance and bonding parasitic capacitance and inductance:
How much is the typical value of these capacitance?

 

[nitishn5]

Pad leakage, especially the diode leakage would not be comparable to the 500nA current that you are biasing the OTA with. It will be of the order of <10nA at high temperatures.
You can try simulations with an additional leakage current of upto 15nA and ensure that the OTA functions properly with every possible leakage at every temperature.

It is a good thing that you are worrying about these issues, so the best thing would be to expect the worst and design for that. Add a few pF of cap and about 1nH of inductance for your cross corner simulations. This, along with the additional leakage, would give a comprehensive result regarding the working of your OTA across corners.

 

[hyleeinhit]

Your reply is very helpful! Thanks a lot for your quick reply!

I am also curious about the Pad without ESD protection. Any comments on this?

Pad leakage, especially the diode leakage would not be comparable to the 500nA current that you are biasing the OTA with. It will be of the order of <10nA at high temperatures.
You can try simulations with an additional leakage current of upto 15nA and ensure that the OTA functions properly with every possible leakage at every temperature.

It is a good thing that you are worrying about these issues, so the best thing would be to expect the worst and design for that. Add a few pF of cap and about 1nH of inductance for your cross corner simulations. This, along with the additional leakage, would give a comprehensive result regarding the working of your OTA across corners.