>>>READER-TO-READER QUESTIONS AND ANSWERS
they take 10 ma on the input. This
would mean if I reset all lanes at the
same time, I would need to sink 80
ma/eight pins. Am I going about this
the wrong way or is there a better way
to interface between the Stamp and
the pin setters?
www.elektor.com) shows the design
of a very elegant and useful curve
tracer. It will test and compare NPN
and PNP transistors, N and P JFETs and
MOSFETs. Control and data via USB
to a PC. Software available — free —
from ELEKTOR. They also can, for a
reasonable fee, supply the PC boards.
You also can get the whole thing as a
kit or completely built.
Since David expects to test
transistors for quite a while, this is a
very good solution. Alternatively, he
can try to find an old Tektronix curve
tracer, but if it fails he may not find any
If David cannot find a copy of the
aforesaid magazine, he can email me
and I'll send him a copy.
Paul J. Weijers
view, is Semiconductor Device
Measurements by John Mulvey in the
Tektronix Measurement Concepts
Series (free as a PDF on the Internet).
I would like a circuit to build an
electronic automobile curb sensor.
#7095 Lynn White
[#3093 - March 2009]
Matching Transistor Gains
I am looking for a practical vs.
exotic bench setup to determine the
gain of some power TO- 3 transistors.
I'm working with an old piece of
equipment that requires matched
transistor sets. It was going to be
replaced in the near future, but “near”
is beginning to look like "someday." So,
for the indefinite future, we need to
keep the existing units alive and the
dead ones restored.
[#3094 - March 2009]
I have designed and built a USB
dumb data terminal for time and attendance tracking. Now that the device is
in production, static discharge resets
the MCU. It's a very simple design
using only an LCD, PIC chip,
membrane switch (with grounding
layer), a plastic enclosure, and a USB
cable to connect to the computer. I
have a grounding layer on the PCB and
the cable's shield and the membrane
grounding layer are connected to it.
Routing the membrane grounding
layer to the cable shield bypassing
the PCB ground improves ESD
performance, but not completely.
#1 A web search for "transistor
gain test" returned testers from simple
to complex that should work for your
jameco.com has some
inexpensive DMMs that have gain
tests built in. A TO- 3 test socket would
also be a good item to have.
#2 The magazine ELEKTOR, vol.
35, February 2009 #386, page 24
#3 Most DMMs have an h func-
tion, but measure at a few µA base
current. If you need to match at higher currents, see Figure 1. Keep heat
input — especially from your fingers —
to a minimum to avoid measuring at
different conditions and leading to
incorrect results. Temperature increases h dramatically. Drafts and sunlight
are other problematic sources of heat.
The suggested setup will have an 0.1%
duty cycle and therefore a rather small
heat input, even at momentarily high
power. R2, R4, and the power supply
voltage can be adjusted to fit your
needs. You may be able to use just
one pulse with a storage scope, but
the pulse(s) may need to be lengthened for some slow devices. The 50
ohm pulse generator
can deliver as much
as 187 mA to the
bases of the transistors. Most function
generators with 50
ohm output should
work also. VBE can
be matched by
connecting to the
bases. If you want to
read more about this
and related subjects,
the best source in my
#1 Reset issues due to ESD can
be solved by the following methods:
1) Add a 1K to 2K ohms resistor in
series with the /MCLR pin on the PIC.
In PIC microcontrollers, the /MCLR
pin doubles as a programming voltage
pin when used in In-Circuit Serial
Program (ICSP) mode. Typical ESD
voltage can exceed the ICSP program
voltage and reset the part as part of
the beginning of the program cycle.
2) If #1 does not solve the issue, add
a 0.1 µF, 50V ceramic capacitor across
the Vdd and Vss pins. This will act like
a decoupling capacitor and take
the ESD energy away from the microcontroller power pins. If the micro you
are using has more than one pair of
Vdd-Vss pins, a capacitor is needed on
every Vdd and Vss pair of pins.
3) If #1 and #2 do not solve the
issue, add a 10 µF 16V electrolytic
capacitor across the power lines (
Vdd-Vss) to smooth out the ripples caused
by ESD voltage.
#2 First of all, make sure that the
RESET input to your PIC is not floating.
It should have a pullup/pulldown
resistor to keep it in the un-RESET
July 2009 95