334 CHAPTER 7. TEMPORAL RESPONSE OF DIODES AND BIPOLAR TRANSISTORS
We assume solutions of the form
n(x, t)=ndc(x)+nω(x)ejωt (7.5.31)
wherendc(x)is the dc component of the current calculated above andnω(x, t)=nω(x)ejωt.If
we insert the ac part of equation 7.5.30 back into equation 7.5.29, the result can be written in the
form
d^2 nω(x)
dx^2
=
nω(x)
λ^2 e
(7.5.32)
whereλeis thefrequencydependent diffusion length and is given by
1
λe
=
√
jω
Dn
=(1+j)
√
ω
2 Dn
(7.5.33)
Assumingnω(wB)=0(Shockley boundary conditions), the solution fornω(x)in equation
7.5.31 is given as
nω(x)=nω(0)
sinh [(wB−x)/λe]
sinh [wB/λe]
(7.5.34)
wherenω(0)is the amplitude of the the ac portion of the electron concentration atx=0.
The value ofnω(0)obviously depends on the magnitude of the ac voltage, which we have
calledvω. Again assuming Shockley boundary conditions,n(0,t)can be written as
n(0,t)=np 0 exp
[
eVBE(t)
kBT
]
=np 0 exp
(
eVdc
kBT
)
exp
(
evωejωt
kBT
)
(7.5.35)
Assumingvωis small, we can linearize the second exponential in this equation, such that
exp
(
evωejωt
kBT
)
1+
evω
kBT
ejωt (7.5.36)
We may then write equation 7.5.35 as
n(0,t)=ndc(0) +nω(0)ejωt (7.5.37)
where
ndc(0) =np 0 exp
(
eVdc
kBT
)
(7.5.38)
and
nω(0) =np 0 exp
(
eVdc
kBT
)
·
evω
kBT
=ndc(0)
evω
kBT
(7.5.39)
Now that we know our ac charge distributionnω(x), we can calculate the amplitudes of the
ac currentsiω(0)andiω(wB).Ifweinsertnω(x)from equation 7.5.34 into equation 7.5.29 and
evaluate the derivative atx=0,weget
iω(0) =−
eAEDnnω(0)
λe
coth
[
wB
λe