I finally did a roast with artisan on a Linux laptop, and I am pleased with how everything worked out. One thing I can't explain is the difference in the ET and BT readings between my two laptops. Here is the first roast on my Windows laptop.


NOTE - my thermocouples read about 25 degre ... st others.

Here is the second roast on my Linux laptop. The background curve for this roast is the prior day's roast on the Windows laptop.

This time, the temp readings seem to be consistently 4 degrees higher. I was afraid that this difference might occur, and now I'm thinking about how to adjust for it when I roast on one laptop to the other. For example, I have various alarms set based on dry end (280F on the Windows laptop) and anticipated first crack.

Not sure why the resized images are chopped off - apparent bug with very long file names?!

I know a lot more about test equipment than roasters.. but most 'utility' grade thermocouples are rated to +/- 2.5 degrees C, or roughly 5 degrees F. So your readings are within the margin of error due to probe tolerance.

I assume you can't get to the probes to do an ice/boiling bath test? If you still have access to both laptops I would do back to back tests to see what each one reads at atmospheric to see if there is a skew between them. Could be something in the libraries in the code. Not sure how Artisan reads from your roaster, if the roaster is providing direct derived temp output or if it is sending thermocouple voltage.

The thermocouples have been mounted in my Hottop for many years, so no way to do the "ice bath" tests. I will be roasting on both laptops going forward, so I will know soon if the differences are consistent.

Is your laptops plugged into the same power strip?? Perhaps one draws more power than the other resulting in more/less heat from your roaster. Just a thought . . .

Many years ago, I noticed that I got different temp readings when my laptop was plugged in, versus running off the battery. To get consistent temp readings, I have subsequently not done roasts with the laptop plugged in. Both of the above roasts were done using battery power.

Assuming the curve settings are the same it appears the Linux laptop has more thermocouple noise based on how the ROR curve looks 'less smooth' compared to the Windows roast. Hope this helps.

You are correct. I have to "strengthen" the smoothing options under Linux to get a curve similar to the Windows laptop roast.

I wonder why Linux introduces more noise? Regardless that RoR line reminds me of when I had my hottop. I used roast logger but the repeatability of that setup is sooooo nice. Roasting on the hottop was quite awesome, just wish it had more power.