This page is devoted to the SpiceAlyzer SPICE signal viewer project.
Linear Technology has made possible a very powerful SPICE simulator called "LTSpice" or "scad". It has been written by Mike Engelhardt (aka "Panama Mike"). He has done an incredible job in bringing LTSpice out, and Linear Tech has done a great job in making it available so reasonably.
All of that said, I have long stuggled with the display of the output data and the inability to display relationships other than the most basic. While LTSpice's "waveform arithmetic" adds to the power, you still cannot readily show quadrature constellation diagrams or logic signal eye charts, to name but two problem areas. Some of the display limitations I see include:
As a way around these limitations, I have begun the development of an independent waveform viewer. And I am seeking help from other LTSpice users. The first picture is a screen shot of a telephone ring-tone as encoded in a wav file. When the waveform is first opened, the waveforms are shown auto-scaled on both axes. The traces are listed on the left side by trace color and trace name. The corresponding sensitivity and offset are shown for the currently selected trace (in the trace list). The equivalent sweep speed is shown below the display. All of the sensitivities and sweep speeds are shown as if the signals were displayed on an oscilloscope as "units per division". This display has only one trace and the automatic sweep speed is 50ms per division.
In the picture, above, I could see that the display is not very accurate because the data samples are displayed too close together (that is, the sweep speed is too slow). I look at the box just below the display and see that the automatically determined sweep speed is 50ms per division. So, I go down to the Sweep Speed selector in Horizontal section and select something a little faster, say 10ms per division. The second picture, below, shows the same waveform with a 10ms per division sweep speed. It works just like you would expect an oscilloscope to work. In this mode, it is single shot with infinite persistance, with the effective trigger at the beginning of the data; the controls and readouts don't entirely show that correctly, yet.
Since this project is in its very early stages, the arrangement of controls, menus, and so forth, are subject to change. This will be particularly true as I get feedback. But, in general, I would like to maintain the general look and feel of familiar products such as oscilloscopes and spectrum analyzers. Many SPICE users are more familiar with, and used to interpreting, information in this environment and I would like to fill this need which LTSpice does not do now.
But, with this great start, I need some help. I am looking for a few collaborators who can evaluate and provide feedback. And, I am looking for a few who can provide some data export files (.raw files) to use for development. In return, you will get the unbounded adulation of a few LTSpice users. If you would like to contribute, please e-mail me.
Where do I want to take this project? The next step is to get the trigger function to work like an oscilloscope. Then, I will add a few simple special functions such as Peak Detect and RMS. Polynomial interpolation is needed when the sample density becomes too low. After that, there will be a "zoom screen" (in a second window, with scroll bars) with much higher resolution. Then, I want to add movable vertical and horizontal cursors with time and signal readouts. Then, it needs to be able to metamorphose into a Spectrum Analyzer or a Network Analyzer. Somewhere, down the line, printable displays will be created. I would also like to add some logic analyzer like functions, such as triggering on combinations of signal states (across multiple traces). And, X-Y displays (one non-time/frequency value against another) are needed.
The development is being done in RealBasic which gives a product that can run on Windows, MacOS, or Linux. The pictures, above, were generated in a MacOS9 environment ("Classic") in MacOS10.3.9. At the present time, this is NOT an open-source project, but someone with the interest and the RealBasic tools could convince me otherwise.
Thanks for stopping by.
James Wagner, Chief Engineer, Oregon Research Electronics
Updated April, 2008