News

Want to measure temperature using K-type thermocouples?

We just finished our first tutorial! This tutorial describes everything you need to know to connect a K-Type thermocouple to an ARDUINO and to the Evolved Horizon application. Start sampling and storing temperature data within minutes. The tutorial is completly stand-alone, and it even includes a companion-script for easy setup!

Click here to get you going!

Attach thermo-couples sensors to a supported device

 

Evolved Horizon 1.13 Released

A smaller update to Evolved Horizon has been released, focusing on GUI performance, and DPI-awareness. All of the GUI components should now correclty scale to OS-scaling on smaller or high-resolution screens in Windows.

Added FFT-like time->frequency domain plot in the analytical interface. Gain insights of hidden frequencies in your data is just a matter of choosing the right plot. Works for non-evenly distributed using the Lomb-Scargle  technique, see screen-shot below.

Power Spectral density plot using the Lomb-Scargle periodogram. Transform your data to the frequency (Power) domain to e.g., reveal hidden frequencies.

Evolved Horizon 1.10 Released!

We are proud to release v. 1.10 of Evolved Horizon, with many interesting new features, see change-log below.

Release Notes for Evolved Horizon v.1.10 (“Curie”)

New Features:

  • Added New Analysis plot type: “Box plot”
    • The box-and-whisker plot may be used to visualize the distribution of data through its quartiles. Use the boxplot for easy overview of the distribution of your data per channel. All the boxes are notched at the D50 (median); two medians are significantly different at the 5% significance level if their intervals (notches) do not overlap. Outliers are marked as red. See more at https://en.wikipedia.org/wiki/Box_plot.
  • Added “Show Previous” and “Show Original” checkboxes in Analyze.
    • Extremely useful feature to see the effect the DSP has on your data. The previous is the last applied DSP data (or original if number of DSPs are 1). The original is the data without any DSP. The current DSP overlays the previous/original who is rendered in a grayish color.

Incidentally, all the DSP applied is only temporary…your data is not altered. In fact, to ensure data integrity, you may not alter it post-I/O!

Fixed:

  •  Various Icon fixes in splash screen and other GUIs
  • Added booter_log.txt in Launcher->Configs->Logs. Changed colors of the log to be white/black for background/text. Added “Open in OS” of all logs.
  • Toned the color of the “selected”-background color of the image-text button.
  • Smaller bugs/fixes
See original and previous data to evaluate effect of DSP
New plot – Boxplot

Evolved Horizon 1.09 Released!

Release Notes for Evolved Horizon v.1.09

Added Features:

  •  Normalization DSP
    • Normalize your data for better comparison, and easier viewing. Use any value for lower and upper limits. Perfect in conjunction with integration for better insights,  without data drowning.
  • Outlier removal DSP
    • Filter unwanted outlier data, using moving-median or moving-mean without smoothing the data too much. Powerful when data contains artifacts. Choose outlier fill methods (Linear, nearest, splines) and threshold values for maximum flexibility to your data-set.
  • Added USB port selection on the firmware uploader. Some devices (e.g., MEGA) requires a specific USB port for uploading. Use “Auto” for complete scan. Please unplug all devices except the device scheduled for firmware upload.
  • Doubled the buffer size and added many more buffer configurations in the firmware. Use long and narrow buffer for few channels but high I/O rate. Use short but wide for many channels and lower (requested) I/O rate. OBS: the actual I/O rate is not only determined by the buffers but also the reading of the pins etc.

See an analysis of two K-Type thermo-couple experiments that seems to have some sensor glitches; possibly some sensor-wire problem. Lets try to remove those to not obfuscate the analysis:

 

Analysis of two two-channel K-Type thermo-couples experiments in two containers of hot and Cold water. There seems to be some artifacts possibly related to sensor wiring… Lets try to remove them!
Alas. Moving mean smoothing does not accurately remove the artifacts (glitches)…
Lets try to find the outliers instead. Use the new Outliers DSP. The data is somewhat smoothed but it performs much better than the moving statistics smooth filter… Incidentally, you may increase the threshold value for even less smoothing. The value used here are >3 median absolute deviations.
Lets intigrate and see how much temperature each sensor experienced; #21 T2 almost cought up with #20 T2…

 

Fixes:

Fixed the lag-race associated with a session that has a I/O-rate set at higher Hz than the hardware can handle. Warning: a diff between actual and requested I/O-rate will not be easily noticed.
Fixed basic DPI-awareness for OS-scaled high-resolution screens… All elements (including graphs) are DPI-aware except some data- and setting-tables.
Rewrote the flush/acquire routine of the session to be more responsive to trigger commands and less taxing to the USB communication. Evolved Horizon 1.09 requires Arduino Firmware +1.4.x.y
Improved buffer memory footprint of the Arduino. More memory == more I/O rates. potentially more sensors. 🙂
Made the graphical user interface engine more responsive (hoover of buttons etc.)
Fixed a bug in the numerical calculation DSP (static time stamps no longer assumed)

Evolved Horizon 1.08 released!

Evolved Horizon 1.08 released!

Stardots is pleased to announce that we are now supporting the lovely Adafruit MAX31855 and MAX31865 breakout boards for K-type thermocouples and PT100/1000 elements. Specify the pin location and you are ready for high-speed, high range, or high precision, temperature readings. Any combination of analog input, PT100 or K-type thermocouple is allowed; turn your Arduino into a sensor powerhouse within minutes!

Major release features:

Hardware: Adafruit MAX31855 15 bit K-Type thermocouple amplifier. SPS: ~1000/#channels

Hardware: Adafruit MAX31865 15 bit RTD amplifier (PT100/1000 elements). SPS: ~10/#channels

DSP: Noise reduction/smoothing: Polynomial Smoothing using Savitzky-Golay technique. Could render more accurately noise reduction compared to moving statistics. Possible to remove e.g., 50 Hz noise.

DSP: Differential Calculus: Visualize the gradient (slope, derivative) of your data. Twice to see the second derivative.

See release notes for all new features and bug fixes.

Download Evolved Horizon, and our Azure powered booter today!

Evolved Horizon DSP

Evolved Horizon lets you combine and apply custom smooth-, scale- and transform-DSPs to your I/O data in matter of seconds, for fast data insights. See example below; first we smoothed (twice) the data using a simple symmetric moving mean with a window of 10 points. Then we scaled the data with both respect of the data and time, using an arbitrary function. Then we performed numerical integration. All DSPs are combined simple by clicking a button. Export the final plot and its data to .xlsx or .csv format, for your final analysis or further export/storage.