petrsocha / sicak Goto Github PK

random-traces-file vs traces-file conflict

At the moment, the plug-in module requires the setTiming method to be called before the oscilloscope can be run. This method is called by meas when appropriate object in Oscilloscope JSON config file is found.

Without calling this function, the driver expects 0 samples per trace (default init value) and that confuses the measurement scenario, resulting in either no data being downloaded, or an exception being thrown.

This means, that in order to perform a measurement, at least oscilloscope timing settings must be given!

This is however not necessary, since this oscilloscope can be set directly using front panel. Given this, the module should obtain current configuration from the oscilloscope before the measurement begins. This will allow to use the meas utility with oscilloscope plug-in even without any JSON configuration.

The error is raised when reading ieee data block from keysight3000 oscilloscope. The measurements may or may not fail. Wtf kind of error.

The correlation matrix evaluation plugin searches for the maximum absolute value, however, stores the non-absolute value for the future comparison.

and the succesfully measured and downloaded data are thrown away, if an error occurs. Given that Keysight oscilloscopes tend to freeze, this is quite unfortunate.

too strict array boundary

The cipher key (sent in random128co, ttest128co) and preselect constant data (sent in ttest128co, ttest128apdu) should be parameterized (e.g. using --param option)

See also https://community.keysight.com/message/91917

When waiting for acquisition to complete, add sleep(100ms) to loop that queries status work register. Similarly when waiting for AER as mentioned in #3.

This will prevent excessive queries to the oscilloscope, which halt the oscilloscope.

Oscilloscopes with newer Keysight-branded firmware does not suffer from this bug.

Keysight 3000 series oscilloscope doesn't allow user to directly set the sample rate. The sample rate, thus the number of samples per trace, is set automatically by oscilloscope in accordance to the selected timebase.

However, only a subset of power samples (from the beggining of the trace) may be transferred from the oscilloscope.

Given this, the number of samples per trace could be set: the driver may heuristically select larger timebase than requested, and download only the requested part of the acquired power trace.

In Keysight driver, add :AER? query loop after every :SINGle.

In PicoScope, add sleep().

After the power trace is downloaded using IEEE488 data block format (:WAV:DATA?), the next command/query causes -410 oscilloscope error.

This error is produced even by querying for error right after the data are downloaded (the first error query returns +0 No Error, but produces -410 error, which gets discovered by the next error query).

This error may be produced by some data (newline?eom?) remaining in the oscilloscope's output buffer, even after the IEEE block reception. The previously downloaded data seems to be complete.

This error was experienced only in Win32+VISA environment so far.