More testing done, better example template

Final update in 2015

After this it's going to be quiet for a while as I won't have any sensors at hand to play with.

The example template has grown a fair bit to include

• calibrating Ramses spectra
• logging (raw/calibrated) data to files
  Although there is a fair bit of code involved in this, it stands on its own outside the pytrios library because its focus is on sensor communication, and not what you'd do with the measurement data afterwards.Feel free to use, copy and share to your liking.

minor tweaks

• Turns out I had the pixel order wrong (again) so this is now proper with pixel 0 the shortest wavelength (for UV sensors this will be the wrong way around. As I've never seen one nor it's calibration file I can't add a detection for this in the calibration routine = where it should go).
• Added some more robust error catching in various places

Test results

• I've managed to run with auto integration times for 1000 spectra in a row without pause, without crashing. Sometimes spectra don't get returned from the sensor, this will not give trouble. Sometimes a returned dataframe (1/8th of a spectrum) is rubbish/noise, also nothing to do with the library and this will be difficult to catch. Incomplete spectra are automatically thrown out.. so this seems pretty solid
• When running with set integration times, you can get things going faster - with more communication errors as a result. This was causing crashes ONLY when the sensor addressing was affected, because we'd be looking for a tchannel that doesn't exist. Catching this specific instance of a KeyError seems to have solved this.

Outlook

• The best way to speed things up would be to start increasing the baud rate, I don't really get why this is set so low by default. However I'd need to figure out how to do this on an IPS4 box and fetch a direct serial-to-sensor cable to make sure I've got it right for the individual sensors. Sometime in the future, or if anyone else feels tempted I'm looking forward to the updates..
• I'd like to start testing a minimalised setup on a small linux bos or raspberry pi. Get power consumption of the whole system for environmental monitoring as low as possible. Then move the sister project Rflex out of labview and into python running on a small linux box.

So that's it - end of xmas break, back to work :)

Stefan

Revamped PyTrios, partially tested

Record raw data from TriOS sensors, particularly spectral data from RAMSES SAM modules.

Includes an example for running a set of SAM sensors on an IPS box and recording 1000 synchronized spectra.

Raw to calibrated data conversion is possible for MicroFlu sensors with stored calibration parameters.
For SAM sensors, conversion of raw values against the manufacturer calibration files will be included soon.

Copyright Stefan Simis (PML). Released under GPL v3. Buy him a beer!