Getting started

You should read this section if you would like to install TFLA-01 on your system and get it working. Don't ask me any questions per E-Mail if you didn't read this chapter of the documentation completely!

Installing the software

Primary installation

Linux RPM

If you have a RPM based Linux distribution which is mostly up to date (which means glibc 2.3, Qt 3.3 and libieee1284 installed), then just running

$ su
# rpm -Uvh tfla-01-0.1.1-0.i586.rpm

should do it.

Start the logic analyzer via command line by running

$ tfla-01

or add a shortcut to your desktop's start menu. In /usr/share/pixmaps you find some icons which you can use for that.

For displaying the help from TFLA-01, you must have installed Qt Assistant and the binary assistant must be in PATH environment. For example on SuSE, just install the package qt3-devel-doc.

Unix via source code

The normal way to install TFLA-01 on Unix is to compile the source code since it is impossible for me to provide binaries for all various plattforms. If you've built a binary (preferrable a package) for your favourite Unix flavour and you plan to maintain it for future TFLA-01 releases, just drop me a note and I'll put a link on the TFLA-01 homepage or upload the binary. The users will thank you!

However, compiling TFLA-01 is not difficult since the dependencies are small. Building on Microsoft Windows is different, so this chapter only describes the build procedure on Unix. At first, make sure that you have following software installed and that the preconditions are met:

  • gcc or another C++ compiler

  • Qt 3.3 including the developement package. Verify this by running qmake on command line. Also, you have to set QTDIR and QMAKESPEC environment properly, see the Qt documentation for more information. On my Linux box QTDIR is set to /usr/lib/qt3 and QMAKESPEC to /usr/lib/qt3/mkspecs/linux-g++.

  • libieee1284 including the developement package (which includes the header files).

Just build the software now with

$ qmake PREFIX=/usr tfla-01.pro
$ make
$ lrelease tfla-01.pro

The first command creates the Makefile from the qmake project file using the specified PREFIX. This specifies the installation location (and nothing more, i. e. the pathname is not compiled in the binary). If you don't specify PREFIX, /usr/local/ is used as a default value. The second command builds the software and the third command compiles the translation files.

To install the software, just run

$ qmake PREFIX=/usr tfla-01.pro
$ su
# make install

You have to repeat the qmake step because the binary doesn't get installed if it is not present while running qmake. Don't ask me if this is a bug or a feature, it just is so and it's no real problem if you know it. Then get root privilegues and install the software. Yes, that's all.

Windows

To install TFLA-01 on windows you need the zip-file supplied in the download section. After downloading this file extract it in the directory you want.

If you want to use a not parallel port that has an address unlike 0x278, 0x378 or 0x3bc you have to create the config file ieee1284.conf in the directory \dokumente und einstellungen\all users\Anwendungsdaten\ (or \documents and settings\all users\application data\). In this file you have to insert something like that:

parport_address 0xFFF0

where 0xFFF0 is the address of the parallel port. To get more user defined parallel ports you can add multiple lines like this. To find out the address of your parallel port you can use Craig's PCI Software.

Permissions

Because you want to run the application as normal user and not only as administrator, you have to deal with permissions of the operating system. This depends really on the system (i. e. there are differences between Unix flavours and even Windows versions) and does not depend whether you compile the program by hand or use a binary to install, so I made an extra section.

Linux

On recent Linux versions (Kernel 2.4 and higher), the applications uses internally the ppdev kernel module. It should be loaded automatically. The only thing you have to make sure is that the user which runs the TFLA-01 program has read and write permissions to the /dev/paraportX device file where X means the number of the parallelport you want to access.

The classical approach is to create a group parport, set the owning group of the parallel port device to parport, change the parallel port to crw-rw---- and add the users to the group as necessary. Read the manual of your Linux distribution for details. SuSE users may use YaST to create the group and add the users to the group, just add

/dev/parport0                    root.parport   0660

to /etc/permissions.local and finally run SuSEconfig to apply the permissions. Of course, you have to logout and login again to apply the new group for the current user.

Windows 98/ME

This is a single user operating system where the current user has all permissions so nothing needs do be done. However, we don't recommend it for using with the TFLA-01, it's instable, insecure, we didn't test it and we will not test it. In short: it's unsupported but it may work, though.

Windows 2000/XP

If you want to run it in user mode you need to use the giveio driver that is supplied with avrdude. The installation of giveio is described in the documentation of avrdude.

Building the hardware

A schematic in Eagle and PDF format is distributed with the program. In source Tarball, you find the schematic in doc/schematic and it gets installed in PREFIX/share/doc/schematic on any plattform. You can also view it online.

It should be easy for you to get and solder the components. The components are cheap and should be available at any electronic part store. In Germany, for example at Reichelt [1]. If anything is unclear, feel free to send me an email at and hope to get a reply. If anyone made a board layout (preferrable with Eagle, too) it would be nice if you could sent it to me so that I can it include in the TFLA-01 distribution or just link on the TFLA-01 homepage.

Figure 1. Image of my hand-made Logic Analyzer

Image of my hand-made Logic Analyzer

To connect it to the parallel port of the PC, you have to solder a small adapter because of the different connectors. You can also use a SUB-D plug on the board and a standard cable.

Now try it out

After you installed the program and built the hardware, it's time to test it. Insert a battery in the logic analyzer and power it on by connecting the JP-POWER jumper. A green LED should show that the power supply works. Now use a wire to connect pin 9 and 8 of the ANALYZE jack. Start the TFLA-01 application by typing tfla-01 on Unix or by double-clicking the TFLA-01 icon in the Windows explorer.

After the TFLA-01 window appears, choose a parallel port in the Settings+Ports menu. All other default settings should work, so just click on the Start icon, use the Analyze->Start menu item or press F5. You should see something like the screenshot below:

Figure 2. Screenshot of the TFLA-01

Image of my hand-made Logic Analyzer

Because there's no changing signal, you see nothing special. The pin 8 is connected to GND while all other pins are unused which means high level here. The grid on the screen represents the samples. Just before the line is drawn the application displays the number of collected samples in the status line. This may give you a hint how fast your parallel port is. You can work out the sample frequency now, it's 1/(number of samples).

Congratulation, your logic analyzer is working. Play a bit now and come back to this documentation to discover all features of the software (which are, of couse, not very much).



[1] No, I don't get any money from this company.