Lab 2 - ECE 421L 

Authored by Jovanne Dahan

dahanj1@unlv.nevada.edu

11 September 2018

  

Pre-Lab:

 

After downloading lab2.zip, I defined the lab2 library in cds.lib. Afterwards, I initiated virtuoso to open the ADC to DAC schematic (as seen below).

 

Next, I launched ADE L and loaded the spectre1 session to simulate the schematic. The waveform is as shown below.

 

In this simulation I set Vdd to 4V which resulted in the following waveform. Mathematically, the LSB can be obtained by  ( or ). By setting Vdd to 4V, the maximum output for the DAC became 3.996V (which can be obtained by the formula ).

 

 

Post-lab:

 

Here is the full schematic view of my DAC.

 

 

Closeup view of the DAC.

 

 

·        Solving for output resistance:

 

Calculations

001

 

 

·        Delay:

 

Modified schematic.

 

Notice that Vout goes to 2V instead of 4V. This is because after the circuit is simplified, there is a 2R in parallel with a 2R which creates a voltage divider (in this case halving the input voltage).

 

Calculations

002

 

 

·        Creating a symbol:

 

First, I copied the symbol for Ideal_10-bit_DAC to the cell Mydesign_10-bit_DAC.

 

Similarly, I copied the schematic for my 10-bit DAC (located at the cell named my_10-bit_DAC) to Mydesign_10-bit_DAC.

 

Below is the cell view with the schematic and the symbol for my 10-bit DAC.

 

 

·        Simulations:

 

When simulating without a load only a third of the full waveform is shown; nevertheless, it is nearly identical to the simulation from sim_Ideal_ADC_DAC.

 

The presence of a 10kΩ load cut the maximum value of Vout by half. Additionally, the Voffset is also halved as can be seen by the waveform below. This is because the load resistor creates a voltage divider, dividing the Voffset and Vin (in this case they are halved because I chose a 10kΩ resistor).

 

 

The addition of a 10pF load tremendously alters the waveform. Vout no longer has steps; this is most likely because capacitors have to charge up, meaning that Vout will take about 5RC time constants before actually becoming Vin. This confirms our observation from when we tried to find the delay.

 

 

Connecting the DAC to a parallel load of a 10kΩ resistor and a 10pF capacitor resulted in a waveform which combines the properties of the previous waveforms. To elaborate, the waveform’s amplitude and offset is halved (like when the 10kΩ resistor was the only load) and the waveform no longer has steps (exactly like when the 10pF capacitor was the load).

 

 

 

·        Switches:

In real circuits, the switches at the output of an ADC are implemented with transistors. If the resistances of the switches are not small then they would affect the total output resistance. One consequence would be that the delay is increased.

 

 

Return to EE421 Labs