Zaurus and the DSO-101

For a while now, I've been working on a port to the Zaurus of the driver software for the DSO-101 oscilloscope. The DSO-101 is manufactured by Syscomp Electronic Design Ltd.

To show the state of the project, here are a few screen shots and descriptions.

The above shot from my SL-C3000 shows the scope in normal operation. The input to channel B (blue trace) is a 16.8Hz sine wave. Channel A (red) has the same waveform, but phase-shifted and reduced by a series C-R combination.

The next shot shows the same waveforms, but this time in high-resolution mode (Hi-Res button checked at the bottom of the controls). In high-resolution mode, the input signals are sampled at a rate in the range 20M samples per second to 4.9k samples per second, selectable as powers of two. In the screen shot, the selected rate is 312kS/s. In each hi-res scan, 32768 samples are collected from each input, and the lower part of the display shows all 32k samples. The upper part of the display shows a magnified view of a subset of these samples. The size of the subset is dictated by the timebase setting, and the blue horizontal scroll bar lets you pan across the sample set. The ends of the scroll bar thumb (and the vertical cursor lines dropping down from them) show which portion of the sample set is in view on the main display.

The third screenshot shows the Zaurus in portrait mode. It took some doing to get the controls to rearrange themselves so that everything still fitted...

Display Features

On the main display, there are various draggable cursors:

• A pair of time cursors, T1 and T2. These can be dragged horizontally to measure time intervals. In the examples, the timebase setting is 2ms/div and the cursors are one division apart, so their time interval is 2ms and the corresponding frequency is 500Hz. Those measurements are shown as the TT entry in the control panel.
• There is an X cursor, visible in the first screen shot. This can be dragged horizontally to control the amount of pre- and post-trigger samples on the display. If it's at the left of the display, samples are collected after the trigger condition fires. If it's at the right, samples are collected until the trigger fires. In hi-res mode, there are always 16k samples collected before the trigger event and 16k afterwards, so this cursor then becomes a non-draggable marker.
• The cursors VA1 and VA2 can be dragged vertically, to measure voltage differences in the waveform. The voltage level of VA1 is shown as the red item in the V1 entry, that of VA2 as the red item in the V2 entry, and their difference as the red item in the dV entry. The cursors VB1 and VB2 operate in the same way, with respect to the B channel.
• The T cursor can be dragged vertically, and controls the voltage at which triggering occurs.
• The A and B cursors can be dragged vertically, to shift the corresponding channel waveforms up and down on the display.

Control Panel

The control panel has pretty much the controls you'd expect:

• Three menus: File, View and AF. The File menu has options to save/reload the current waveform, and to save/reload the current control settings. Don't get excited here: the code for these options has yet to be written. The View menu lets you reduce display clutter by removing the time and/or voltage cursors. The AF menu affects the auto-frequency measurement display, which in the screen shots means the measurement entry labelled A. If you look at the first screen shot, it's saying that the channel A waveform is measured at 16.8Hz, and its period is 59.5ms. In the hi-res shots, the main display contains less than a full cycle so the auto measurement does not take place.
• The channel controls A and B let you turn a channel off altogether, or adjust its sensitivity with the left and right arrow buttons. Sensitivity ranges are in 1:2:5 ratios from 50mV/div to 5V/div.
• Trigger mode can be selected as Auto, Normal, Single-Shot or Manual. The trigger source can be selected as channel A or B, and triggering can be done on a rising (+) or falling (-) edge. The trigger voltage level (set with the T cursor) is also displayed.
• The Single-Shot Reset and Manual Trigger buttons do what you'd expect.
• The timebase speed is shown in the examples as 2ms/div. It can be varied from 200ms/div to 50ns/div with the arrow buttons, in the traditional 1:2:5 sequence.
• The Hi-Res button selects high-resolution mode, as already described. In hi-res mode, the button which is showing 312kS/s is actually a menu button, letting you set the hi-res sampling rate.

Download

If you've made it this far, you might want to try out the software, so here's dso101_0.3.0_arm.ipk. There won't be much to see without a scope connected, but you'll get an idea of whether it runs at all on your ROM and kernel. PLEASE be sure to back up your Z before trying it - this is a work in progress, and if it crashes and destroys all your data, I will refuse all responsibility.

Installation is not quite as straightforward as usual with the Zaurus, because there are also two USB drivers to install. Those are ftdi_sio.o and usbserial.o. You can download them through the links.

There is also a PDF document article.pdf. This contains a few background notes on the project, together with detailed installation instructions and a list of known bugs.

Also please note: the software writes a file called dso101.trace, which is the usual idea of debugging output: it reports what happens, in case it's useful when something goes wrong. So don't leave the program running overnight, or your disc partition will overflow. But if you find the software doesn't work for you, I'd appreciate a copy of the trace file with your report.

Vapourware

These items are on my wish list, and one day they may happen - but don't count on it!

• Make the entries in the File menu do something.
• Syscomp manufacture a waveform generator as a companion item to the scope. It's controlled in the same way by host software, and I'd like to add a control panel for it to this project. I've made a start, but there's some way to go yet.

Acknowledgements

For getting the project this far, I'd like to thank:

• The members of the OE Forums, who have helped me to get to grips with the Zaurus both by direct answers to my questions and by the answers they have provided to others
• Syscomp, who have actively supported the project with advice, information and donated hardware. And who, of course, made the project possible in the first place by open-sourcing their software and being happy for me to mangle it.