Project - EE 420L
By Marco Muņiz,
05/08/2019
Email : munizm1@unlv.nevada.edu
Lab
description
Design a voltage amplifier with a gain of 10 using either the ZVN3306A or ZVP3306A (or both) MOSFETs and as many resistors and capacitors as you need.
Specifications:
- As fast as possible driving a 1K Load
- Input resistance greater than 50K
- As large an output swing as possible
- Can pass a 100Hz Signal
- Can draw under quiescent conditions ( no input signal), no more than 1mA from a 9V supply voltage
________________________________________________________________________________________________________________________________________
Project:
For
this project, I will be using the Push-Pull circuit as the base for our
Voltage Amplifier. I chose this amp topology because it works very well
in sourcing, and syncing, current due to the fact that
at any given time, at least one of the transistors will be on. Also,
the push-pull topology is great for driving loads due to its small
output resistance.
(Figure 1: Push-Pull
Amplifier)
The
first project spec we must consider in this design is the current draw
with no input connected. We must draw no more than 1mA of current from
a 9V power supply. In order to make sure that the current drawn is as
small as possible, we will add resistors to the sources of both
transistors(Nmos and Pmos). Specifically, they will be placed between
the source and VDD for the Pmos and between the source and ground for
the Nmos. This will help in limiting the current draw so we dont pull
large amounts from the power supply.
The calculations for the needed resistance and capacitance can be seen in figure2 below.
(Figure 2: Hand
Calculations for Gain, Rin, Resistance, and Capacitance)
Let Current ID = 0.7mA
Let Rb = 1Gig Ohm
KPn = 0.1233 Vthn = 1.824V
KPp = 0.145 Vthp = 2.875V
The
large capacitors on the sources were used in
order to treat the sources as AC ground for simplification when doing
the gain analysis for the circuit, this gain analysis is pictured
above. On the left image above, we can see the
relation between Vs and Vout which results in the gain equation. Since
our Load Resistance and gain are fixed for this design, the only
components that can be freely adjusted for the gain are (R1||R2).
To find Rin, we find the relation between Vs, A (Gain), and the
Current. We will set Rb to a fairly large number to, say 1Meg Ohm, and
solve for Rin with the needed gain.
We already know the values of KP's and VTH for these devices from the spice model used in Lab 6
, we can set the drain current to a value under 1mA, say 0.7mA, and
solve for the VGS, VSG, Resistors, and finally, the capacitors.
R1 | 6k Ohm |
R2 | 120 Ohm |
Rin | 90k Ohm |
Cin | 1.1 uF |
Cout | 10 uF |
Cs | 28 uF |
VGS | 1.93V |
VSG | 2.97V |
gmp | 14.2 mA/V |
gmn | 13.1 mA/V |
(Figure 3: Tables with calculated component values)
Designed Voltage Amplifier:
For
the final design, all calculated component values were implemented.
Overall, the design did meet the calculation expectations, however, we
did have to separate the Nmos source impedances. This was done to match
the RC time constants on both sources so we wouldn't unintentionally
restrict the signal from passing. In doing this, we now had the added
benefit of being able to micro-adjust the gain value of the amplifier
by making minor changes to R2 and R4 since those resistances dominated the parralels on the sources and controlled the gain.
(Figure 4: Final Voltage Amplifier
design)
Quiescent Current Draw:
As
we can see in the image below, the source resistors helped in limiting
our current draw from the power supply, with no input signal.
(Figure 5: Current draw from 9V Power
Supply under quiescent conditions)
Circuit Gain:
This plot shows a Small Signal AC analysis with Vout/Vs = A(Gain). At the design requirement of 100Hz, we see a gain of 20.9 dB = 11 (Gain), and reaching a maximum gain of 23dB = 14 (Gain) until 1Mhz.
(Figure 6: Circuit Gain Plot)
Our
Vin was set to 100mV, thus our experimental gain = 816m/100m =8.16. We
tried increasing the gain experimentally by increasing the source
resistances but the highest value we were able to attain was 8.16. The
oscilloscopes were also pulling in some noise which was messing with
our read input voltage, which would have affected our experimentally
measured gain.
We estimate our gain from the transient where Vout/Vin => 1.17V/100mV = 11.7 Gain
(Figure 8: Transient showing Vin and Vout)
Input Resistance Simulation:
The
input resistance Rin was found by using a Small Signal AC analysis to
measure the input voltage Vin divided by the current through the input
capacitor Cin. [Rin = V(Vin)/I(Cin)].
From the hand calculations, we found an input resistance of 90k Ohms.
The simulation showed an Input resistance range from 83k ohm at 100Hz
to 65k ohm up to 1MHz. The simulation shows a clear illustration of how
the input resistance in the Push-Pull Amp is difficult to keep exact
since the feedback continuously lowering our input resistance due to Rin = Rb/(1-A)
(Figure 7: Rin Simulation Plot)
Max Output Swing:
The
output swing was found by increasing the amplitude of the input
voltage, which will increase the input current, and looking at the
waveform when the output was clipped to its maximum and minimum values
For this design, the output swing simulation was roughly 4V.
(Figure 9: Output swing simulation plot)
_____________________________________________________________________________________________________________
| Calculated | Simulated | Experimental |
Gain | 10 | 11 | 8.19 |
Quiescent Current | 700 uA | 705 uA | 791 uA |
Output Swing | | 3.77V
| 1.34V |
Power Draw = Iquiescent * VDD = 791uA * 9V = 7.12 mWatts
In conclusion, the amplifier design displayed met all of the
requirements during simulations. However, the experimental results
varied by a larger margin. Specifically, we did not reach our expected
gain value of 11, nor did we see the same output swing, our quiescent
current was fairly spot on.We may be able to
increase this output swing by placing more devices in parallel with
larger resistor values to help account for the larger current demand.
This would require we supply more current which would put us closer to
the 1mA current limit. Another improvement which can be made to this
design is using a common
source amp as the input. This would allow the designer to more
accurately control the input resistance so that we do not have to deal
with the push-pulls feedback decreasing the gain.
Return to My Lab's
Return to Student Directory
Return to Cmos Home