[MOVED] Can put the inductor into the deep nwell

[incol]

Dear everyone,

I am considering the layout of the LC oscillator, can put the inductor into the deep nwell? For noise isolation, I think it's better to put the inductor in it, but I don't konw if it will lead to any other problems of the inductor.

Thanks a lot.

 

[erikl]

It's the best you can do, I think. Just connect this nwell to the highest vdd potential available, to achieve lowest possible cap parasitics.

Even better would be such one, of course - if only your process would allow for ;-)

 

[dick_freebird]

As a rule you want the inductor over the highest resistivity
material you can get. How deep Nwell compares to regular,
I don't know (but I might assume it's higher to support high
voltage devices, its usual purpose).

You want an idea of thing like, does the deep Nwell abut the
low resistivity handle (if this is an epi process), which might
result in a higher well capacitance to substrate than the
plain Nwell which might sit in the higher resistivity epi layer
(or not - have to know the doping-pairs and look up the
depletion region characteristics that result).

You may also see different oxide stack thicknesses over
the two wells, which affects SRF and stray C and eddy
losses (Q). If you're putting the inductor on an upper
interconnect level this is probably just a few % but at
layer 1, could be significant difference.