Advertisement
Section VIll
SERVICE
GROUP
B
Model 3455A
strument
set
to
the 100
V
range
and
with
Auto-Cal
on,
measure
for
any
readings
at
AtOTPl.
If this
voltage
is
incorrect,
check
the bias
of
the input
FET's
by
shorting the drains
of
Q38
to
the sources
of
Q38.
TPl
should
now
read
zero.
The
drains
of
Q37
should be approximately
+
10
V
and
the
sources
of
Q3
should read between
+
1
V
and
+
2
V.
Troubleshoot
the active attenuator,
if
the
readings
are
bad.
b.
For
the
100
V
range
zero check,
set
the
34S5A
to the
100
V
range with
Auto-Cal
off.
Adjust
R66
for a
zero reading
at
TPl.
If
unable
to adjust for
a zero voltage,
check
for
a leaky
C21, C22, C26,
CR32,
or
Q37.
If
the zero reading
is
good,
the
34S5A
should
display
0
V ±
1
count,
a short
time
after
Auto-Cal
is
turned on.
If
an
offset
remains
on
the display,
Q36 may
be
defective.
c.
For
a 100
V
range
full
scale
check, observe
for a
cal
constant 5
failure,
when
the
34S5A
is
in
the
self-test
mode.
If
cal
constant 5
fails,
troubleshoot
its
circuitry
by
using the
procedure of paragraph
8-B-l
1.
If cal
constant 5
passes, step to cal
constant
5
by
using the
procedure of paragraph
8-B-I
I.
If
cal
constant
5
passes, step to cal
constant
5
by
using the
procedure of paragraph
8-B-l
I.
Measure
TP2
for
approximately
-1
V
and measure
for
approximately
-.1
V
at
the
10:1
divider (pin 9
of
J3).
Check
for
exactly
-i-
10
V
at
the Junction
of
R47
and Q33, and
for
+
9.9
V
at
the
junction
of
R47
and
K6.
If
10
V
is
measured
at
R47
and
K6
instead
of
+
9.9
V
or
R63
may
be open.
8 B-27.
1000 V Range
or
Constant
6
Fails.
a.
For
the
1000
V
range zero check,
remember
that
some
of
the
same
circuits
are
used
in
the 100
V
range.
The
100
V
range
zero
and
gain
should be
working
before troubleshooting the 1000
V
range.
Check
for the
proper
switch closures
used
in cal
constant
6
(use the
procedure of paragraph
3-B-1
1
to
locate
constant
6).
b.
For a
1000
V
gain check, use the
procedure of paragraph
8-B-24b.
c.
Set the
34SSA
to
the
1000
V
range with
Auto-Cal
on,
and
apply
1(XX)
V
to the input terminals.
If
the
reading
on
the display
changes
intermittently
from
100 counts to
200
counts,
A10K5
or
K6
may
be
breaking
down.
KS and K6
can
be
checked by connecting channel
A
of an
oscilloscope
(set
to
20
V/div)
at
the
Junction
of
R47 and
K6.
WARNING
For
safely,
connect
the
scope
probe
lo the
100
K
resistor
R47.
Connect
channel
B
of
the
scope
(2
V/div)
to pin 10
of U24.
Set the
scope
to
the
chop
mode
and
trig-
ger
on
channel
B.
If
channel
B
indicates
a
5
V
spike
when
arcing occurs, as seen
on
channel
A,
then
K6 may
be breaking
down.
If
the indication
on
channel
B
appears
to
be good,
KS may
be
defective.
C21, C23,
or
the input
node
of Q37,
may
also
be
defective.
8 B-2B. Various Other Malfunctions.
8-B-29.
Shorted FET's.
a.
Occasionally
FET's on
the multiplex
node
may
short.
Two
ways
can be used
to
isolate
shorted
FET's.
1.
Measure
the gate
to
on
resistance
with
an
ohmmeter.
2.
Short the input
of
the
34SSA and
turn the high resolution
and Auto-Cal
functions
off.
While monitoring
A10TP3
step
through
all
the
dc
ranges
(1
V,
10 V,
etc.)
and
ac
functions.
Make
sure
Auto-Cal
is
turned off
after
every range
and
function change.
If
TPS
dips to -24 V, a
FET may
be
shorted. Usually, the defective
FET
is
normally
turned off
in
that particular
range
or function.
b.
If
AlOUl
gets
very hot
Ul, Q2, Q13,
or
Q14
could
possibly
have
their
gates
shorted
to
their
can.
8-56
Advertisement
