Spice and Capacitance

[harry456]

I tried to plot the capacitance as calculated with a bsim-model and I know that the spice software calculates the values like Cij = dQi/dVj but I don't know how that works. Esp. because Cdb and Cbd are different.

Could someone explain to me why Cdb and Cbd (resp Csb and Cbs) are different in my case?



I guess that around VSource=-600 mV the source-body diode get conductive but why is there a difference between Cbd and Cdb and with of them is the drain-bluk capacitance?

 

[kemiyun]

Re: Spice and capacitance

Hey, I can't really answer these questions but I think you should refer to the model documents. The thing is other than being different from each other some of them are negative if you noticed. I know that bsim employs negative capacitances for some awesome purpose but I don't really know the basics of it.

Nevertheless if you actually put a measurement setup to measure those capacitances I'm sure that you'd end up with the same number. The numbers that the simulator reports are the numbers that are used inside the model and inside the model the capacitances you mentioned may have different meanings.

I'm not 100% sure on what I said so you should ask someone more experienced but it's a valid question and I realized I should do the model document reading that I've been postponing for billion years. I'll share if I find the reason.

 

[D.A.(Tony)Stewart]

Is there a DC bias difference in computing the AC capacitance?

As you know C is inversely related to conductance which is affected by DC bias and peaks in gate Ciss during DS conductance transition zone.