BT lagging compared to other curves

[rowancoffee]

I'm experiencing a delay on my BT curves when the ROR for BT is relatively low. The BT will stagnate at the same temperature, then update by several degrees after about 5 seconds, while the ET will gradually increase or decrease. This mostly impacts the end of our roasts when trying to determine a target drop temperature. This occurs across various sample intervals, and I've attached a reference with a sampling interval of .5sec because it shows the staircase effect of the BT compared to the smooth ET.

We're roasting on a new 15kg Mill City.

[MAKOMO]

Please attach your Artisan settings as artisan-settings.txt file (save via menu Help >> Save Settings and rename to end in .txt before drag-and-drop it here on the GitHub Web page into a answer field of this thread) as well as the recording you show in your image as .alog file (again saved, renamed to end in .txt and drag-and-dropped here on an answer field) to allow us to analyze the data that the machine sends.

[rowancoffee]

Thanks for the reply. Attached are the settings and the recording.
05.19.25 Example.txt
artisan-settings.txt

[MAKOMO]

I don't see anything obviously wrong here. It seems the machine is not delivering fresh data more often than every ~1.5sec on all channels, less often here for BT for unknown reasons. Intermediate readings are approximated by Artisan as you have the "Interpolate Duplicate" filter active. Not much you or Artisan can do to improve this.

Here are the raw readings with all smoothing deactivated (you can see them also in the Roast Properties dialog under Data).

[rowancoffee]

Thanks for looking into this Marko. Is 1.5sec slow in general for the other channels as well or is that about average for what you see on other machines? Is reaching out to Mill City the next best step for me to determine why the BT is delivering data slower than the other channels?

[MAKOMO]

I am myself sampling at 2sec interval on my machine. Note that taking a temperature reading takes time and there is an absolute limit (depending on the electronic you are using and the sensor type) how fast data can be delivered (independent of communication). I would say everything up to 3sec is absolute ok.

Nobody in the team has access to or any information about Mill City machines. The producer of equipment that gives troubles should always be the best source for information on it.