>>>READER-TO-READER QUESTIONS AND ANSWERS
box, one in the TV control van; and
two technicians in the 1st and Ten van
— who constantly adjust the system as
the lighting and weather change.
TV cameras used to televise the
game are fitted with encoders to
collect XYZ and lens zoom settings so
that the computer can modify the line
to fit the playing field as seen from
that vantage point. The color of the
playing field is separated from the
color of the player's uniforms and
body so that the line is painted on the
field underneath them.
Before the game, a computer
model of the actual field is recorded to
aid in creating the 1st and Ten graphic
yellow line (which must also have the
correct curve to match the playing
The broadcast TV video is passed
through the 1st and Ten system to
have the yellow line graphics added,
and the live sound from the game
(crowd and announcers, etc.) is
delayed electronically to match the
image processing delay caused by the
[#12075 - December 2007]
I need a device with two serial
ports that can capture a line of RS-232
data at 1200 or 2400 baud and then
reorder the information sequence (user
definable) and resend it on demand at
1200 or 2400 baud to a PC. The
purpose is to convert data from a PBX
to a form the PC can recognize; the
program the PC is running cannot be
adjusted to read the original data. I
have looked at various single board
microcontrollers, but most seem to be
designed for 9600 baud.
#1 Check out the offerings from
Rabbit Semiconductor (www.rabbit
semi.com). They have inexpensive
single-board controllers that are
C-programmable and have four or
more serial ports, each of which can
have independent baud rates.
The BL1810 is one I am very
familiar with, and it will easily do what
#2 I wish you'd specified the PBX
and the program on the PC. I maintain
three Nortel PBXs and half a dozen
keysystems, and have faced the
problem of data from the switch being
in an inconvenient format.
Instead of trying to correct the
data between the PBX and PC with
custom hardware, I captured it to file
on the PC. Then the data was parsed
and reformatted with a program I
wrote. If the PBX is a Meridian, and if
the data is LD 2 TFS (sorry about the
machine specific code), I may already
have what you need. If not, I may be
able to help anyway.
If the program on the PC is
Hyperterminal, I should warn you that
Hyperterminal becomes unstable if
"Capture to file" is left on indefinitely.
Hyperterminal is okay, it just wasn't
designed for long-term data logging.
(Yes, I found out the hard way.) :-)
Again, my fix was to write a terminal
program that logs everything, and
detects TFS and saves it to a separate
file. There are too many details missing
to give a more specific answer
here. P.M. me at http://forum.servo
magazine.com/ username dyarker,
and we can swap email addresses.
#3 What you're looking for used to
be called "RS-232 Wedge" which was
software that would take a serial data
stream and convert it to a keyboard
sequence. The last version I used
would take almost any protocol and
data rate, massage the data, and
transfer it to a file or a calling program.
Check for "Serial keyboard software"
via Google and you'll find www.ke
yin-jector.com which sells it.
#4 From your description, it sounds
like you are running a PC-based PBX.
You don't state what OS is on the PC,
but if it is Windows or Linux, it would
be possible to write a "translator"
program that would run in the
The program could receive your
legacy phone data on one serial port
and either send the translated data out
on another serial port or via a virtual
serial port. The serial port you select in
the PBX software could be the virtual
port or a third serial port that "null
modem" connects to the output port
of the translation software. A PC
software solution would require little
additional hardware (possibly addition
serial ports or USB-to-serial adapters)
and easier modification; both are
advantages over hardware solutions.
With this solution or with a hardware
solution (single board computer),
you will have to develop the program
to do the translation. The hardware
solution would be the best bet if your
PC runs a non-standard OS.
[#1081 - January 2008]
Anyone know a rule of thumb for
determining how much energy you can
pump into an LED without destroying
it? I've seen many specs on maximum
voltage, current, and duty cycle.
However, I’m trying to determine how
much current you can actually pump
into one in a short duration pulse
(around one microsecond) without
damaging the LED junction, and at
what maximum repetition rate. I've
heard of strobe applications where 10X
the maximum absolute current could
be pulsed without destroying the LED.
#1 You are correct with the
assumption that having the LED on for
a short period of time allows for an
increase in the intensity of the current
through the LED. Basically, you want
to keep the power dissipated constant.
Let's assume that the manufacturers
specify a value of 20 mA for continuous operation for one LED and we are
going to have it ON for 0.1 seconds
each second. So, this will allow for a
current of 200 mA while the LED is on
because it has enough time to cool off
before another peak of current arrives.
Extrapolating this, we can find the
maximum current as I(max) = I(nom) *
T/Ton with I(nom) being the nomimal
current ( 20 mA) in the previous
example, T is the repetition period
(one second in the previous example),
and Ton the time that the LED is ON
(0.1 seconds in the previous example).
This equation will give a rule of
March 2008 103