Required level of device physics

Hi

If I want to be an Analog designer (Discrete or IC) how much device physics is needed?

Already done a semi-conductor physics course, but not devices (i.e First 6 chapters of Neamen's book and an intro to PN junction).

Do I have to take a device course in order to be good analog engineer?

Level: Undergraduate

Thanks.

thushianthan15 said:

... Do I have to take a device course in order to be good analog engineer?

Click to expand...

Not necessarily, but it could be helpful, at least if you'd later decide to design analog ICs, and perhaps even create their layout.

If you ever want to do analog design at the chip level (designing IC chips), than a good foundation in device solid-state physics is likely necessary.
Otherwise you can probably get by without it.

crutschow said:

Otherwise you can probably get by without it.

Click to expand...

Right, you probably can, because the DRC rules prevent that you'd violate necessary spacing regulations, determined by device physics. But an understanding of the DRC rules on a device physics base is very helpful in recognizing the reason for their violations' error messages, and thus for repairing those - or how to run them over ;-).

erikl said:

Right, you probably can, because the DRC rules prevent that you'd violate necessary spacing regulations, determined by device physics. But an understanding of the DRC rules on a device physics base is very helpful in recognizing the reason for their violations' error messages, and thus for repairing those - or how to run them over ;-).

Click to expand...

There's much more than DRC rules needed to design analog IC chips.
You also need detailed models and understanding of the active device characteristics and an understanding of device physics is very helpful in that.

There's all kinds of analog design and you might be fine
doing higher level work using only approved PDK elements
if that's your thing. But if you need to cheat, then you
need to "cheat smart" and part of that is being able to
challenge the "No!" Posse (your device and reliability
engineers) when they start mouthing dogma at you.
Someone doing ESD and I/O design, or power management
or trying to perform an analog function with as few devices
as possible, will benefit from a gut understanding of device
behavior and fragility, and what can be done by layout or
by processing to improve what's on the plate. That, or
learn to like the taste of macaroni and cheese and nothing
but.

crutschow said:

You also need detailed models and understanding of the active device characteristics and an understanding of device physics is very helpful in that.

Click to expand...

Yeah, would be awfully nice, if active device characteristics would find their way in cell models - usually it isn't so, unfortunately. (A semiconductor physicist - me).