# technology current for inversion coefficent

1. ## technology current for inversion coefficent

Hi, I am using TSMC 180nm process for opamp design in CADENCE. I would like to know how can I get the Technology current value in order to get the Inversion Coefficient. Can anybody provide the solution..

•

2. ## Re: technology current for inversion coefficent

The technology current (t.c.) depends slightly on temperature, s. e.g. this curve from D. Binkley's book, p. 59:
technology-current_vs_temp.pdf
This t.c. is given for a similar 180nm process (fab not named). At room temperature the t.c.≈0.6µA for NMOS transistors, for PMOS it is a factor of 3..4 less.

t.c. = (Id / (W/L)) / IC , where IC = Inversion Coefficient

Following Binkley's definition of the (fixed–normalized) IC, IC=1 for Veff = Vgs - Vgs0 = 40mV

For an estimation, you could measure Vgs0 , resulting in a very low saturated drain current, of, say 1nA , then measure the output characteristic for Veff=40mV , i.e. IC=1 . Measure the saturated Id (say at Vds > 200mV) , then you can estimate t.c. = (Id / (W/L)) .

•

3. ## Re: technology current for inversion coefficent

Thanx erikl for your valuable response. I have one doubt that inversion coefficient remain same for 180nm process whether the foundry is TSMC or UMC etc. Please reply...

•

4. ## Re: technology current for inversion coefficent

Originally Posted by rishabh_31ec
I have one doubt that inversion coefficient remain same for 180nm process whether the foundry is TSMC or UMC etc.
Sure, it could be a little bit different, but not so much: the (fixed–normalized) IC has a fixed relationship to the technology current:
t.c. = I0 = 2n0µ0C'oxUT2
As all these physical parameters shouldn't differ too much for the same process node, also the t.c.'s shouldn't differ so much. And the same is valid for the IC, which then only depends on Id, Veff, and W/L .

The ratio of the NMOS/PMOS t.c.'s, however, seems to depend on different fab processes (it depends on the relative µ0 values of NMOS & PMOS resulting from the process): usually this factor is about 2.5 , but in Binkley's book it's > 4 for his 180nm process.

Above I suggested a simple t.c. measure / estimation approach for your process, by simulation or measurement on real silicon.

--[[ ]]--