The best way to find out is for you to perform some tests using your particular hardware and software arrangement. You don't need UPNLTs to perform these tests. You do need the two computers concerned to be next to one another.
These tests are designed to confirm that
(1) your arrangement will work using a serial port interconnection
(2) either your arrangement will work without serial port status signals,
or you can fake any necessary signals by hardwiring inside the serial port
plugs
(3) you can send a configuration string to the UPNLT
(4) your arrangement can be set up to establish the network connection
after the configuration string has been sent.
If all these are true then you should be able to use UPNLTs for the
link.
Test 1
(If you don't have a null modem cable, you can go straight to Test
2, but identifying the cause of any problem may be a little more difficult.)
Connect a serial port on one computer to a serial port on the other
computer using a null modem cable.
Configure your network software to run the port at 115,200 bps, 8 data
bits, no stop bit, no parity bit. Attempt to establish the connection.
From minor to much fiddling with settings etc. may be necessary, depending
on your arrangement.
Get this direct connection working to your satisfaction.
Test 2
Make a three-wire null modem cable, using a length of three-core wire
and two DB9(F) connectors.
Using the three-core wire:
- connect pin 5 on connector A to pin 5 on connector B. This is the
earth-earth interconnection.
- connect pin 2 on connector A to pin 3 on connector B. This is one
TXD-RXD interconnection.
- connect pin 3 on connector A to pin 2 on connector B. This is the
other TXD-RXD interconnection.
Replace the null modem cable used in Test 1 with this three-wire null
modem cable.
Attempt to establish the network connection. If the connection can not
be established, look for options provided by your software and attempt
to set the comms port to
- no hardware flow control, also called no hardware handshake or no
CTS/RTS
- no status signals, also called no DTR, no DSR, no DCD.
If you can establish and use this connection, transfer some large files
each way between the computers to confirm your arrangement can run at a
sustained 115200 bps. If you get errors, try lower a speed. For a sustained
115200 bps, 16550 UARTS and comms drivers which utilise the FIFOs are generally
necessary. Skip the remainder of Test 2 if all works correctly.
If it all worked in Test 1, but you now can't get the connection to
work despite much fiddling with the software settings, you'll need to fake
the control and status signals by doing a bit of hardwiring in the DB9
connectors.
On each connector (but not between connectors)
- connect pin 7 (RTS) to pin 8 (CTS)
- connect pin 6 (DSR) and pin 4 (DTR)
Attempt to establish the connection. Get this three-wire direct connection
working to your satisfaction.
Transfer some large files each way between the computers to confirm
your arrangement can run at a sustained 115200 bps. If you get errors,
try lower a speed.
Test 3
Connect the computers using the three-wire null modem cable. Start
a terminal program on one of the computers. Set it to 9600 baud, 8 data
bits, 1 stop bit, no parity.
On the second computer, run the initialisation
software you intend using to send the configuration string. If you
are using the Windows program for initialising, unplug the null modem cable
from the second computer when the program says to depower the UPNLT, and
plug it back in when the program says to power up the UPNLT.
You should see some evidence of the receipt of the configuration string
on the screen of the first computer. Parts of the configuration string
may be unprintable characters, so you may not see it all.
If you are using the Windows program for initialising, it will halt
at this point with a message saying the configuration string was not echoed.
Ignore this message, and close the program.
If convenient, change the configuration string in the initialisation
software to printable characters, run it again, and confirm that the correct
characters are displayed on the first computer.
Get this working correctly.
Test 4.
Power down the computers. Disconnect the three-wire null modem cable
from one of the computers.
Start the computers, and wait until they've completed all their startup
procedures.
Perform Test 3. Shut down the terminal program.
Establish the network connection using the network software.
Repeat, exchanging the roles of the two computers in Test 3.
If you can get all this working correctly in this order, then you should
be able to use UPNLTs for the link.
Can I use a laser diode from a laser pointer, and if so, how do I set the driver trimpots?
We have successfully used laser diodes from laser pointer modules supplied by Oatley Electronics. Here's how.
Measure and note the current drawn by the laser diode module. Call this current Il.
Build a rudimentary laser power meter thus:
Wire a BPW34 photodiode (or another with at least 1 mm square
sensitive area) , reversed biased (ie. cathode to +), in series with a
1 K resistor and a meter set to read a few milliamps. Apply 12 volts or
so DC.
Power up your laser pointer module and shine the spot onto the sensitive area of the BPW34. Adjust distance and laser focus to ensure essentially all the beam falls on the sensitive area. Read and note the current flowing through the BPW34 as indicated on the meter. Call this current Ip.
Remove the laser from its driver board.
Follow the procedure given in "Laser Transceiver Setup and Adjustment", but in section 3. "Laser Diode current limit adjustment" set the limit to about 1.2 times Il.
Set VR1 fully clockwise, but otherwise omit the adjustment procedure detailed in section 4. "Automatic Power Control adjustment ." Resume procedure at "Disconnect the power to the UPNLT".
In the procedure described in "Laser
Transceiver Adjustment and Testing", "Laser power adjustment", and
with the spot from the laser on the BPW34, set VR1so the current through
the BPW34 is Ip/5. DONT OVERSHOOT!!!
(The IrDA duty cycle is about 20%, so this sets the APC for the same
peak laser output as did the original driver )
Is it possible to fit the SMD version of the CS8130 to the PCB?
Yes, but with difficulty. See this
page.
The leads to the photodiode and to the bypass caps must be short. Even
a few millimeters introduces noise into the receiver, which reduces useful
range. Adapter boards tend to have longer conductors and be too noisy.
(If you intend adapting our PCB design to suit the SSOP CS8130, then
please see the CS8130 Datasheet first regarding layout. It's critical -
feel free to email us. If you do create such a PCB design, please make
it available to others either via us or on another web site.)