Colin McDonald writes:
>Just curious, were you totally blind at the time? And if so, what sort of
>ways did you use to work with circuit boards and IC chips and such?
Yes, for all practical purposes. I get data sheets to any
chip I want to use and either have somebody read it to me or copy it
down in Braille from the computer if I get it that way and put the
information in a notebook. When I build the circuit, I generally use
perforated board and wire wrap sockets to connect the chip or chips
together. I use one of those Wahl cordless soldering guns to solder
wire wrap wires to the leads of any capacitors, resistors, crystals or
other parts needed because wire wrap connections need square pins in
order to seat properly. If you just use the wire wrapping tool alone
on the round leads of most components, the wire will loosen up and
come right off almost immediately. Even worse, it can sort of come
off and leave bare wire flopping around which can touch neighboring
components or wires and intermittently short out which, of course, can
cause damage to the circuit or make it work poorly.
I used to use one of those multimeters that AFB sold in the
seventies that has a pointer dial with a raised scale on it that you
turn until the tone nulls out.
One of the things that a technician who is blind needs to be
able to do is to determine the polarity of electrolytic capacitors.
If you get it wrong, they will act as a short across whatever DC
voltage they are filtering and do one of a couple of bad things. They
will either damage something else in the circuit by acting as a short,
prevent the circuit from working because they are shorting some bias
voltage to ground, or blow up with a spectacular bang and lots of
shrapnel which could hurt somebody not to mention make a big mess if
the capacitor is a big one.
One of the things I do is to use a little circuit I built
using a VCO or Voltage Controlled Oscillator chip. The idea is to
bias the control pin on the VCO to VCC which could be 5 or 12 volts
depending on what circuit you are using. You want to use some
relatively high value of resistor to do the biasing so there isn't
much current. A 20 to 50-thousand-ohm resistor is good for this
purpose because you want to be able to safely short it without
anything getting hot.
If you put an electrolytic capacitor between that
current-limited voltage and ground, you will hear the oscillator pitch
reflect the short and then slowly creep back to whatever pitch it had
before you shorted it.
If you connect the electrolytic capacitor backwards, the short
will stay and the tone will never slowly return to the un-shorted
condition.
If you build a circuit like this and test an electrolytic
capacitor, be sure to discharge the capacitor after determining the
correct polarity. If it is a filter capacitor, it has more than
enough juice in it to destroy semiconductors if you happen to connect
it to your circuit while it still is charged. Just short the leads
and hear the spark and be glad that wasn't your circuit that just went
snap.
As for what VCO to use, there are several. The little circuit
I built up uses a NE566 function generator which is a very old chip
and virtually worthless for anything useful, but it works for this
purpose.
Its output is inversely per portional to the voltage on the
control pin so you hear a low tone when the voltage is high and the
highest tone when the pin is shorted to 0. Other VCO's have a
proportional output so the tone is high if the voltage is high.
Either idea works as long as you know what is expected.
I haven't tried this, but you might even be able to use a
NE555 timer chip in a multivibrator mode just like what works in a
light probe. Where the resistor would go, put the AC input to a diode
bridge and then connect your test capacitor to the DC output terminals
of the bridge. The oscillator will start out high and drop like a
fire siren if the capacitor is large. You might want to put at least
a 1000 ohm resistor in series with the bridge to limit current.
I hope this kind of answered your questions. Obviously, there
are lots of tricks for building circuits that a technician who is
blind can use. The one thing I have no answer for at all is what to
do about surface-mount technology. The parts are too tiny and the
techniques for assembling them too dependent on microscopes and
photographic technology.
The nice thing about PIC's and microcontrollers in general is
that there is less grunt work involved in building stuff because more
of the circuitry is simply programmed in to the chip or chips so there
isn't as much to hook up.
You can take a 8-pin PIC such as the 12C509 or 12F675 which
has 8 pins like a NE555 and tell it what polarity of input and output
you want and end up with that one chip being your entire circuit for
some simple applications. There are now programming tools and
software emulators for the PIC that run under Linux.
Those of you who use Windows and JAWS have a free program
called mplab which is made available by the Microchip company which is
the same company that produces PIC's. I have no idea how
speech-friendly mplab is. I am a Linux user myself so I use the UNIX
tools when programming PIC's and it can make you feel like a rich kid
on Christmas morning. I have only actually programmed one PIC, but
that's just because I hadn't made up my mind what project to do next.
The one PIC I did program is a 16F84 and it does just what I told it
to do in the program so they really do work.
For the last twelve years or so, I have had one of those Radio
Shack talking multimeters which I plan to use until I blow it up or it
dies. After that, I will use a meter that has a serial port. You
certainly need good test equipment to know what is happening with your
circuit because very little of this stuff ever works perfectly the
first time.
I think this is enough to digest for now so I will return all
of you to your regularly-scheduled hum.
Martin McCormick WB5AGZ Stillwater, OK
OSU Information Technology Division Network Operations Group
|
