Roast Meter User's Guide
Roast color correlates to coffee taste, and along with other data provides a way to corroborate the quality of a coffee.
- The Roast Meter is intended to be used by casual coffee roasters and coffee consumers to evaluate the color of their coffee for quality purposes.
- The Roast Meter replaces the use of roast color cards and out-dated roast descriptors with something more quantifiable that can be shared with others.
- While the Roast Meter achieves good alignment with reference meters, it is not intended for use in applications requiring high accuracy and precision, such as in a commercial environment.
- The Roast Meter is not designed to replace a calibrated reference meter, such as a LightTells, Agtron or Roastrite.
- No guarantee is made for the accuracy of the readings from the Roast Meter against reference standards.
The formula used in the Roast Meter to generate an Agtron Gourmet scale reading was created by testing many coffee samples using a number of Roast Meters and a Lighttells CM100 reference meter. Roasts from 50-130 Agtron on the CM100 were tested when creating the formula. A fine grind at 70 microns burr spacing on a Lagom P64 with MP burrs was used in the sample measurements across all the Roast Meters and the CM100 used in the testing.
*Note: the Roast Meter itself requires a fine grind for precision due to the limitations of the sensor hardware. Also note the CM100 manual doesn't specify a grind size to be used for that meter. Thus, the Agtron Gourmet formula used in the Roast Meter does not claim to follow the SCA Agtron Gourmet sample preparation standard, and it doesn't claim to provide completely accurate Agtron Gourmet readings.
Regardless of whether you are calibrating or taking measurements, please follow these general guidelines when using the Roast Meter:
- Power on Roast Meter and allow to stabilize for 3-5 minutes.
- Use it at room temperature with room temperature substances.
- If the sensor glass is dirty, clean with isopropyl alcohol.
- If the substance clumps, consider performing WDT before applying to the sensor.
- Apply enough of the substance to be measured to the sensor well so that the red LED light can no longer be seen
- Use a consistent method of tamping the substance before taking a reading
- After taking a reading, remove the substance with a soft bristled brush, and clean with isopropyl alcohol if needed.
Ideally, calibration of each Roast Meter would be done against reference meters and substances. But since it is an open source and owner-built device this is not possible. So, calibration is instead achieved by proxy using commonly available reference substances and reference range targets for those substances. These targets were derived from initial testing with the CM100 and are updated using crowdsourced data from users.
- Efforts are underway to find consistent, inexpensive and globally available reference substances for the Roast Meter.
- As of April 2023, an additional reference substance is:
- Sodium bicarbonate (e.g. baking soda)
- This is being experimented with due to inconsistencies found in using Hario papers
- Apply a heaping amount to the sensor so that little red light shows through
- The target reading objective is around 230 on the display
- As of March 2023 the current reference substance and target is:
- Hario V60 untabbed natural (unbleached) filter paper
- Cut into 16mm (⅝ inch) discs from a single layer of the filter paper
- 10 discs placed on the sensor and held in place with light pressure
- The target reading objective is 182-188 on the display
- See General Usage Instructions
- Apply the reference substance to the Roast Meter.
- Record the reading on the display.
- Remove the substance and repeat several times to get an average reading.
- Change the variable "ledBrightness" in the "roast_meter.ino" file up or down 1 number based on the averaged reading. E.g. if the average reading is below the target, increase ledBrightness and vice versa.
- Upload the changed code to the Roast Meter
- See the software installation section of the assembly manual for detailed instructions
- Repeat the calibration process until the display shows the reference substance is in the target range.
- TBD
- The color will change with the age of a roast. When possible, try to measure a coffee at the same age.
- For roasters, general best practices are to test within 1-2 hours after roasting. Try to measure at consistent post-roast times.
Consistent grinding to an appropriate paricle size is very important. There can be a difference of 20-25+ Agtron for a coffee simply by using different grinder settings on the same grinder.
- The Roast Meter reads reflected light from tamped coffee placed on the sensor glass.
- The grind of a single coffee may cause different Roast Meter readings. There are multiple factors which can cause this:
- Different grinder burr sets create different particle size distributions, which may lead to different results even if using the same micron size setting
- Coffee origins, roast levels, and age of roast may influence resulting grind sizes and particle distributions within the same burr set.
- When precise results are needed, sieve ground coffee before measuring. Sieving also helps remove chaff that may affect readings.
- Initial results as of April 2023 indicate sieved grind sizes of <500 microns generate repeatable readings on the Roast Meter. This may change as more data comes in.
- For those that lack sieves or other grind size measuring equipment, grind to as fine a grind as possible and use that setting consistently. (e.g. fine espresso or Turkish grind level). But, do not grind so fine that you start getting clumps.
Users can spot check a coffee with one test. But for more precision the following more rigorous procedure is recommended:
- First, see General Usage Instructions
- Obtain several samples of beans from the bag or roast of the coffee to be tested and remove any quakers.
- The amount of coffee per sample needed to cover the sensor will depend on grinder and grind size, but start with 3-5 grams unground coffee.
- Grind each sample, apply it to the sensor and tamp. Record the Agtron reading on the Roast Meter display.
- For each sample remove and reapply the coffee to the Roast Meter several times. The purpose is to ensure each sample gives the same reading when applied more than once.
- Brush the sample off with a soft bristled brush between each reading.
- Repeat this process for each sample and average all readings to arrive at a final result.
As an example, for each coffee to be tested remove (3) separate samples of 5gr each and put into cups. Grind a sample and apply to the Roast Meter and take a reading. Remove the sample back into its cup, reapply and repeat the reading a total of 3 times. Discard this sample and move to the next one, and repeat. This example results in an average Agtron from 9 readings taken for one coffee.
- Questions about the Roast Meter can be addressed on the Discord channel.
- Report bugs and enhancement requests using GitHub
- Luther, M. (2023, March 5). Artisan blog: Understanding Roast Color. Artisan Blog. Retrieved April 9, 2023, from https://artisan-roasterscope.blogspot.com/2023/03/understanding-roast-color.html
- Münchow, M. (n.d.). Variation analysis of roast colour measurement. CoffeeMind. Retrieved April 9, 2023, from https://coffee-mind.com/variation-analysis-of-roast-colour-measurement/
Rev. 1.0 3