Inline Refractometer: Getting real-time density readings
The idea of this project was to measure the Brix of a wort inline during recirculations and transfer during the various processes of a brew day. For example it could be used as part of the sparging process, so that the sparge can be stopped when the standard gravity (SG) approaches 1.008 (2.1 °Brix) to prevent the extraction of tannins. I can measure the mash’s first running’s to be able to see if the conversion efficiency was higher or lower than expected. Additionally, the principle could also be utilized during fermentation assuming appropriate formulae are utilized to account for the presence of alcohol. That one is still in theoretical phase though.
For this build, I utilized a typical non-digital brewing refractometer along with a Raspberry Pi Zero and camera to image the refractometer output. If you are unfamiliar with Raspberry Pi technology, then I highly recommend looking into it. The capacity of these small boards is incredible. But back to the project at hand . . . the temperature of the liquid pumped through the inline refractometer needs to be measured as the reflective property of water changes with increasing or decreasing temperatures. I’m using a DS18B20 1-wire temperature probe at the outlet of my brewing vessel near the butterfly valve so I can track this and later use it in my correction calculations. I tried to correct for the effect of high temperatures on the Brix scale of monitoring sugar density using a simple calibration process.
Unfortunately for everyone reading, this build is still in beta mode. Further tests will need to be conducted to determine the accuracy of this system at various temperatures using more comparisons to a standard digital refractometer, for example. My system’s mash temperature readings are still a little off when compared to my digital refractometer. Additionally, other improvements would include investigating plastics most suitable for upper-end mash temperatures of around 70–80 °C (158–175 °F) to prevent warping along with better seals. I unfortunately had issues with leaks through the lid at higher pressures when I increased flow rates, so a different type of lid may need to be investigated too. A glass jar could be the solution, but I am not comfortable cutting a glass jar.
For this build, I utilized a typical non-digital brewing refractometer along with a Raspberry Pi Zero and camera to image the refractometer’s output.
Also, darker beers are problematic as the ability to see through the jar is a limiting factor with this design . . . like trying to perform a starch conversion test on a stout. A thinner, squared compartment could possibly allow readings on dark-colored beers.
That aside, I’m excited for this project. There are several advantages to obtaining real time sugar readings. Being able to see that running’s gravity is approaching the danger zone of 1.010 (2.6 °Brix). Or the fact that I can ascertain that my gravity readings are lower than expected during sparge and I will need to make a kettle sugar addition to correct the problem and to check my grain crush for potential inconsistencies.
Tools and Materials
- Transparent plastic jar
- (2) ½-in. bulkhead
- M32 cable gland
- 1-mm (1⁄32-in.) drill bit
- 11-mm (7⁄16-in.) drill bit
- 21-mm (13⁄16-in.) Q-Max hole punch
- 32.5-mm (11⁄4-in.) Q-Max hole punch
- Refractometer
- DS18B20 temperature probe
- Raspberry Pi Zero
- Raspberry Pi camera
- Mini tripod
1. Drilling Pilot Holes
It is necessary to drill three holes (two for the bulkheads on the side of the jar and one for the cable gland used for the refractometer at the bottom), first by drilling pilot holes using the smaller drill bit, then for each hole use the larger drill bit. The larger hole allows a hole punch to be inserted.
2. Punching Holes and Adding Bulkheads
The drill holes are used to insert a hole punch into, to punch the larger hole. I made use of an Allen key to control the rotation of the punch to create the large hole. Once two larger holes have been punched using the 21-mm punch in the sides of plastic, two bulkheads may then be added.
A larger hole needs to be punched on the bottom of the jar using the 32.5-mm punch and then the M32 cable gland added, which is used to hold the refractometer.
3. Refractometer Logistics
It is important to tighten the cable gland once the refractometer has been inserted, using spanner wrenches to ensure there are no leaks through it.
An LED light was placed above the jar to provide illumination for the refractometer. The Raspberry Pi camera is placed below the refractometer to take photos at one second intervals. The angle and position of the light was also very important to get the best image from the refractometer. The camera position itself was also critical. It’s not shown in this photo, but I found also using a piece of paper as a diffuser over the light helped improve the photo quality.
The inline refractometer was suspended in a lab clamp above the camera. You can see silicone hoses were attached to hose barbs on the bulkheads.
4. Inline Logistics
I make use of a Blichmann Riptide pump, which has a linear flow valve to fine-tune the flow rate allowing me to throttle down to try to minimize leaks. The pump pushes the liquid through the inline refractometer and then can go either back to the mash tun or on to the kettle.
You can see the DS18B20 temperature sensor, which was inserted through one of the tri-clamp ports of my brewing vessel. This temperature sensor was connected to the Raspberry Pi through jumper wires.
5. Calibration Time
The following step involves heating plain water from around 20 °C (68 °F) to 70 °C (185 °F), to see the apparent Brix change (first photo below vs. second). This is important as we need to attempt to correct for the higher temperatures that may be present in different brewing processes. A simple bash script was written, which recorded photos of the refractometer using raspistill at one second intervals whilst simultaneously recording the temperature of the water from the DS18B20 temperature probe.
I processed the images I recorded from the camera, using the OpenCV Python library to rotate and crop the image. I used two of the images I recorded from this calibration process to generate the following table manually (using an image editor to obtain the pixels from the top of the scale to the refractometer line):
I entered these values into a spreadsheet program and used it to generate an equation for the trend of the line, which I used later in testing. The equation it created was used to try to correct for different brewing temperatures. Note that I am planning to create a simple program to automatically obtain the position of the line from the refractometer photo at arbitrary temperatures using simple image processing techniques, this would allow me to create a calibration table with many more points than from the two images I picked out of the many photos the Pi captured.
6. Testing
To compare the results from my DIY inline refractometer to reality I added 500 g (1.1 lbs.) of powdered sugar to the brewing vessel’s water at 70.3 °C (158.5 °F). I took a reading of this sugar water by collecting a small sample using a measuring cylinder and pouring a small drop onto a digital refractometer (Photo above).
Using the temperature correction equation created from the calibration process I created the a python snippet, which calculated the Brix from the image I took from the inline refractometer. This gave me a Brix value of 1.323 (using values 1284 pixels from top of scale and a temperature of 70.3 °C/158.5 °F), whereas my digital refractometer gives a value of 1.4 Brix. Close but not perfect. The tinkering continues . . . please check out my website www.anfractuosity.com to follow along. Cheers!
Written by Chris (Anfractuosity)
How nice would it be if you could monitor your wort’s gravity in real time without constantly needing to pull samples? While this build is still a work in progress, its innovative
design is still impressive enough.
