Lab 03 - ECE 421L 

Authored by Desi Battle,

battled@unlv.nevada.edu

09/20/2017

  

Lab Description: Layout of a 10-bit digital-to-analog converter (DAC)

 

 

 

To begin the layout of our 10-bit DAC designed in lab 2, first we need to determine how to layout the individual 10k resistors

For our process we know that the n-well layer has an approximate resistance of 800ohms/square, and our minimum width is 12*lambda or 3.6um (lambda = 300nm)

 

sheetequ.png

Substituting using R= 10k, W = 4.5u (the width of ntap with 2 contact rows)  , Rsq = 800 and solving for L gets us..

sheetequSolved.png

So an n-well with L = 56.25u and W = 4.5u should measure approximately 10k ohms

nwell_10k.JPG

After laying out the resistor I used rulers to measure my W and L.  The W matched exactly but the L came out to be slightly shorter than 56.25 at 56.1. 

Being 0.15 microns shorter that I originally planned is fine since I was approximating to begin with.


Before going any further, I made sure my resistor will be able to fit neatly on the grid by dividing W and L each by the distance between grid points (0.15), ensuring the result is an 

integer.  Otherwise, DRC will fail due to being off grid.

 
 
extract_10k.JPG

Here the extracted view of the layout above is pictured to confirm the resistor is approximately 10k.  It is a little higher than I would like, but since our R-2R topology is more 

dependent on the resistors all being the same value rather than precisely 10k, I can use exact copies of this cell without too much harm.

  

 

 

layoutzoomed.png

Using the previous layout as a template I connected the 10k n-well resistors in an R-2R topology.  After I connecting B8, I did

a DRC to verify I did not break any rules I was unaware of, deleted B8, and used the F3 option when copying to copy 9 rows of bits at once.

After that was complete I only had to add the bottom 10k resistor, input pins, Vout, and gnd to complete the layout.

 

 


LVS.jpg
Once the layout was completed, it looked similar to the schematic, but to verify the equivalence I extracted my layout, and performed
an LVS. My netlists matched, and the output of the LVS report did not flag any other issues. (Click Here to View  Full Report)

 
 
LVSsim.jpg
Finally, In ADE L I added "extracted" to  Setup -> Environment to simulate my extracted DAC and compared it to the identical sim run on the schematic to show they generate the same waveform.
 
 
extractSim_proof.JPG
Simulation -> Netlist -> Display used to confirm I did indeed simulate the extracted view

 
Click HERE to download a copy of the library used for the DAC design and layout

 

Retrun to my labs