Sources of Noise
Table of Contents
Overview
This article will introduce the different sources of noise that cause temperature fluctuations.
Instructions
Noise is a common issue in PC-based temperature logging systems and can originate from different sources:
Sensor noise
The temperature sensor itself can produce noise, e.g., when the probe is grounded or worn off.
Grounded thermocouples
Grounded probes are too sensitive for roast temperature monitoring and prone to inducing noise from ground loops, which results in a less accurate reading.
You cannot distinguish between a grounded and an ungrounded thermocouple simply by visual inspection. You will need a multimeter to perform this test, as shown in this video by Omega.
Worn off probes
Over time, your probe will become worn out due to constant vibrations caused by the coffee beans. On worn-off probes, the sensor within the probe will come in contact with the sheath.
This connection can create electrical noise known as "ground loops," which causes irregular spikes in your temperature curve. If this occurs, you will need to replace your probe.
Poorly isolated probe
When there is constant noise while the machine is running, it may refer to a poorly isolated probe. You can try moving the sensor a bit.
Tips and Recommendations
If moving the sensor does not do the trick, you may need to unground the thermocouple from the machine by isolating the probe.
Signal wire noise
Signal wires are sensitive to electrical interference from nearby motors or power cables. These disturbances (EMI and RFI) distort the signal, resulting in a "noisy" or jagged roast curve and Rate of Rise (RoR).
Tips and Recommendations
Signal wire noise can be eliminated by rerouting the sensor wires or using ferrite beads.
Ground loops
A ground loop happens when two connected devices (like your roaster and your computer) have different paths to the ground. This creates a "circle" where unwanted electrical currents can flow, causing jagged spikes in your temperature graphs.
Your roasting machine is grounded through its power plug. When you connect your laptop via a USB cable or a Phidget, you create a second path to the ground.
If your temperature probe is grounded (touching the metal of the roaster) or simply worn out, it closes the loop between the two machines, letting "leaking currents" interfere with your data.
The "Battery Test"
You can easily confirm if a ground loop is the problem:
- Unplug your laptop from its power adapter.
- Run it in battery mode.
If the noise on your graph disappears, you have a ground loop. By unplugging the laptop, you broke the second path to the ground, leaving only one.
How to Fix It Permanently
If you don't want to run your laptop on battery every time you roast, you have two main options:
- Replace the probe: Use an ungrounded thermocouple.
- Isolate the probe: Ensure the sensor is physically insulated so it doesn't touch the metal frame of the roaster. If you are using a Cropster Connector (or Phidget), you may want to add a USB isolator between the Phidget and the computer.
Sources
The information in this article is based on:
- https://www.omega.co.uk/techref/das/noise.html
- https://phidgets.wordpress.com/2014/06/02/how-to-avoid-grounding-a-thermocouple/
- https://www.cropster.com/news/article/what-probe-type-is-it/
- https://en.wikipedia.org/wiki/Ground_loop_(electricity)
If you have further questions, feedback, or concerns, please feel free to submit a support ticket directly from support.cropster.com.