BJT

Simulation
ECE 417
George Vartanov

Introduction

BJTs consist of two PN junctions constructed in a special way in series, back
to back. In this case we are looking at an NPN transistor. We have looked at all of the
material parameters of the device and used those parameters to calculate the
important values in the device. We are using three different types of software to
visualize the different possible interpretations and presentations of the BJT.

Part 1: Matlab
Material Parameters:

uE
88.299
The values to the left are the parameters that are
uC
1281.7
unique to this BJT device. They were calculated by using
uB
287
tE
4.1663e-11 the provided equations in the assignment as well as the
equations provided in lecture. The mobility values in all
tB
1e-7
my regions seem very reasonable, as I have seen similar
tC
4.386e-7
values in the past. The minority carrier lifetimes seem
pE0
71.108
like they are on the correct order of magnitude because I
nC0
500
would expect them to be quite small. My minority
pB0
14286
carrier concentrations are somewhat concerning to me,
De
2.2826
but the amount of doping that is present makes the
Db
7.4191
values seem a little more reasonable. The diffusivity
Dc
33.132
values I have are reasonable as well because the amount
Le
9.752e-6
of diffusion occurring in the collector region should be
Lb
4.77e-4
higher due to the amount of doping in that region.
Lc
1e-3

Table 1


The graph at the
left is showing the
characteristics of the
current as the junction
voltages increase in both
the collector and base
region. In the first
portion of the graph
while Vce is below .1v
the transistor is in
saturation mode. After
the collector-base
voltage increases past
.1v, the transistor is in
saturation mode. At no
point on this graph does
the transistor reach a
Figure 1
breakdown voltage.


Results:
Variable Name
Is
BetaF
BetaR
Re
Rb
Rc
Vjc
Vje
Cje0
Cjc0
TF
RF
VAF
VAR
IKF
IKR
Gm
Gpi
Go
Gmu

Values
1.1973e-18
44.785
1.5798
.1149
5.1957e3
6.9617e4
.784
.9998
1.2937e-15
2.6869e-16
8.4038e-12
3.1273e-11
5.9186
1.2293
1.1214e-4
9.612
6.6857e-5
1.4925e-6
2.9257e-7
6.5313e-9

Table 2

The values at the left are vital for
understanding the operation of the BJT. Is is the
saturation current of the transistor. This shows
the small amount of current that is always
present in the transistor. This is a characteristic
that determines BetaF and BetaR and is unique
to each device. BetaF represents the forward
current gain and should be as close to 100 as
possible. BetaR is the reverse current gain and
should be smaller than BetaF. The forward
Early voltage is larger than the reverse Early
voltage which makes sense. When comparing
the corner current values to the graph the
graph from lecture, they made sense regarding
the orders of magnitude. Go is greater than
Gmu, Gpi is greater than Gmu and Gm is greater
than Gm so these values check out with the
expectations. The transit times are relatively
small which makes for good switching times.




Part 2: HSPICE

Complex plot using all variables:

Figure 2

The graph at the

left shows the HSPICE
representation of the I-V
curves for the BJT. This is
very similar to what I
acquired through the
Matlab simulation, but it
has some visible
differences. Although we
were able to use a large
number of variables in the
Matlab plot, we were not
able to account for the
early voltage that we can
see in the plot at the left.
In addition, it is visible

that at the minimum value of Vbe, the transistor falls into a cutoff mode that
produces a current that is virtually zero.
Plot Using only Is, Bf, Br:

The plot to the left
is much more relatable to
the Matlab representation.
This depicts the current as
Vbe increases. This is the
same magnitude as the
Matlab graph, which is
encouraging. The spacing
between the current
curves is almost identical
to the Matlab graph of the
same values.




Figure 3

Gummel Plot:


The plot at
the right depicts
the collector
current and the
base current at
various values of
biasing in the base-
emitter region. The
collector current is
barely impacted
until about .3 volts
where it increases
rapidly. The ratio
between the two
currents will
determine the gain
value. The dip in
Figure 4
the base current is
likely due to some incorrect values so I am slightly disappointed that I was not able
to get the correct depiction for this plot. Also the high current roll off is not visible in
this range of voltages. My estimate for the maximum CE current gain would be
around 78, but due to the lack of precision in the graph, it is hard for me to estimate
the exact value. This is the ratio between the collector current and the base current.

Transconductance Values:
Gm
Gpi
Gmu
Go
Table 3

1.458e-5
7.06e-7
6.2e-9
1.36e-7

Gummel Plot:

The small table at the left depicts the various

transconductance values in the BJT. Once again, Gmu is
smaller than the other values. In comparison to the values
that were calculated in Matlab: Gm is the same order of
magnitude, Gpi is slightly smaller, Gmu is almost identical,
and Go is slightly smaller.
Part 3: ATLAS

Figure 5

The above plot is a secondary representation of the Gummel plot. This plot
begins at the active region and has a clear representation of which base voltage will
produce the largest beta value. Although the variables are still being carried over
from the previous calculations, this Gummel plot seems more reliable than the
previous Hspice generated plot. There is no unexpected behavior from what I
understand from the Gummel plot that I have seen in class.

Net doping:

Figure 6

The above image is a depiction of the levels of doping within a semiconductor

exactly as it would appear in a real life version of the BJT. This matches my
expectations seeing as the base is very heavily p-type away from the emitter and the
emitter being very n-type and being almost unaffected by the base region. The n-
type collector region makes up most of the BJT and definitely is affected by the very
p-type base due to diffusion across the boundary region.

Mobile charge distribution:

The image at the left is a
representation of the mobile charges
within the BJT. There seems to be very
minimum mobile charges within the
transistor other than within the base
region. This is likely due to the fact
that there is transfer of charge
through the base region. The reason
the majority of the mobile charges are
close to the emitter region is due to
the proportional amount of n-types on
the right side of that region.
Figure 7

Donor and acceptor concentration/Net doping concentration

The plots at the
right are likely the most
interesting to me out of
all of the ATLAS
components. There is a
clear connection
between the donor,
acceptor and net
doping concentrations.
The first thing worth
analyzing is the visible
donor doping along the
green line that dips
down as the emitter
region turns into the
base region and
increases again in the
collector region. The
acceptor doping is only
Figure 8
visible between the
emitter and the base
region.

The most
exciting part that I was
able to notice was that
the intersection points
between the acceptors
and the donors match
up with the dips in the
overall net doping
concentration drops.
This makes perfect
sense because there is
recombination
occurring at those two
points due to the equal
numbers of acceptors
and donors. Also the
net doping looks
similar to the donor
doping because of how Figure 9
much more impact the
electrons have on the doping.

Energy bands/Electric field/Electric potential:

Figure 10

Figure 11

Figure 12

The plots at the left can
be viewed as a progression.
Starting with the energy band
diagram, it is clear that there is
a large amount of energy for the
mobile carriers to overcome
when moving from the emitter
to the base region. This point
where the energy increases
dramatically is the same point
on the E-field graph that has a
spike in the electric field.
The electric field also
jumps up when the slope of the
energy band changes from
positive to negative. This
induces a negative electrical
field and then levels out as the
energy band moves into the
collector region. At the same
spike of electrical field, there is
a huge voltage induced in order
to compensate for the amount
of energy at that point in the
band diagram.
This is an interesting
visualization as the voltage is
similar to a reflection of the
energy band diagram across the
x-axis. The electric field is only
induced when there is a change
in energy so the graph is
accurate in depicting the
locations where the E-field
increases.


Equilibrium/Active/Saturation Modes:



The plots at the right are the
depictions of the energy band
diagrams three different BJT states.
The initial picture is the equilibrium
plot with 0 bias. This is identical to the
plot that is visible on the previous
page.
The second graph is more
interesting in displaying the energy
bands while in active mode. This
means that the B-E region is forward
biased while the B-C region is reverse
biased. This means that the majority
Figure 13
of the mobile carriers will be found in
the base region due to the increase in
energy in the collector region. It is
clear that the energy between the
emitter and the base region is greatly
reduced which is a proper
characteristic of a forward biased
region.
The third and final graph is the
depiction of the energy bands in a
state of saturation. In this case, there
is almost no energy difference
between the base and the emitter
allowing even more diffusion to occur.
The collector will be collecting a
large amount of mobile carriers due to Figure 14
the energy drop that is experienced
from the base to the collector region.

Figure 15