Measuring Total Dissolved Solids: A Refractometer Comparison

posted in: Brewing, Tools | 31
Share on FacebookTweet about this on TwitterShare on RedditShare on TumblrDigg thisShare on Google+Share on LinkedInPin on PinterestShare on StumbleUpon

Abstract

The goal of these two studies was to examine the measurements from two refractometers advertised to measure the total dissolved solids (TDS) content of coffee. Utilizing mixed study designs, the VST LAB Coffee II Coffee & Espresso Refractometer (the newer VST LAB Coffee-Espresso III will be tested soon) was compared against the Atago PAL-COFFEE refractometer in two different experiments. For the first study, distilled/demineralised water was used to zero the refractometer in one condition and brew water was used to zero in the other condition. Our results demonstrated no significant difference in the readings between the VST and Atago refractometers, nor between the two water types. Another experiment was performed to assess the impact of the syringe sample filters supplied by VST. Our results demonstrated a significant difference between refractometers (p = 0.04), a significant effect of filter use on TDS reading (p = 0.02), as well as a significant interaction between filter use and temperature (p = 0.01).

Disclosures

We have no vested interest in either product being tested.

Authors

Jeremy and Joe

 

Introduction

Coffee is most often thought of as a solution, consisting of a solvent (water) and solutes. As described in much greater detail by others (e.g., Flament 2002; Andueza et al. 2003), over 600 potential flavor-impacting compounds have been identified in the seed. While not all compounds contained within the roasted beans are soluble, the resulting complex beverage that is brewed is considered by many a homogenous mixture composed of a single phase with particles that cannot be seen by the naked eye, thus making its soluble content quantifiable through standard means of solution concentration measurement such as refractive index.

In simplistic terms, the index of refraction describes how light passes through a medium. The angle of refraction as light passes through the coffee (i.e., how the light bends when passed through the solution) is often captured by a photodetector and the resultant refractive index calculated. This is a method typically used to assess the concentration of a solute in an aqueous solution, most often called the total dissolved solids (TDS) percentage (for more in-depth explanation on refractive index and its application to coffee, see Clarke 1989).

At least two consumer-grade refractometer product lines have been marketed specifically for coffee solute concentration assessment: a series marketed by VST and a series produced by Atago. Both have offered several models. Early iterations of what would become the VST refractometer were produced by Atago, who has been developing refractometer technology for over 70 years. VST then switched to Reichert Techologies refractometer hardware (Reichert has a 150-year history of producing optical measurement equipment), originally called the R2 mini but now sold as the Brix/RI Chek. The current model tested here is based on the hardware of the Misco Palm Abbe. VST branded equipment has custom-loaded software on its refractometer units to provide a TDS of coffee/espresso directly without the need for correlation tables. Atago has been marketing their own coffee-specific refractometer calibrated with an “industry scale” specifically for coffee, providing a direct readout of either Brix (a scale based simply on the refractive index of sucrose in an aqueous solution) or correlated TDS reading. For anything other than sucrose, the Brix scale only provides an approximate dissolved solid content reading. VST recommends the use of a syringe filter for certain non-paper filtered coffee brews or espresso. Atago does not make any suggestions regarding the use of a sample filter for filter brewed coffee or espresso.

We set out to test the performance of the two devices compared against each other, not assess their individual precision or accuracy against stated manufacturer specifications. To test precision and accuracy within a device would require us to compare against another metric of TDS quantification and a known reference solution.

Methods

The VST device used was the LAB Coffee II (Coffee & Espresso) Refractometer (stated range: 0.00-20.00%; TDS resolution 0.01%; precision +/- 0.03%; accuracy from 0.00-4.99% of +- 0.05% and from 5.00-20.00% of 0.10%; temperature range of 15-40C) and the Atago device was the PAL-COFFEE (stated range: 0.00-25.00% Brix; TDS resolution 0.01%; temperature range of 10-100C). Both were loaded with fresh batteries and warmed to room temperature (25C).

Distilled water, using a conductivity-based TDS meter, contained 0 ppm. The brew water used contained 94 ppm. Both of these registered as 0.00 on the VST and Atago.

Equipment used for the experiments.

u2015-04-25 22.08.172015-04-27 08.38.49

2015-04-24 22.19.592015-04-24 22.23.082015-04-25 23.57.56

The coffee used for the experiment was Toby’s Estate fully washed Rwanda Ruli roasted for filter coffee.

Equipment used:

  1. Kalita stainless steel wave dripper
  2. Kalita filter paper
  3. Kalita carafe
  4. Bonavita 1.0 L digital variable gooseneck kettle
  5. Infrared gun
  6. FLIR thermal camera
  7. AWS scale
  8. HGone Conical Manual grinder
  9. Atago refractometer
  10. VST refractometer
  11. 2 new sets of batteries
  12. Pipettes
  13. Beakers
  14. Alcohol wipes
  15. Ceramic cups
  16. Tissues

Study 1: Which Water to Zero Your Device

Pour over coffee was brewed at 94 degrees Celsius at a ratio of 1g coffee to 17g water. The brew was 15g of coffee to 220ml of water, with a total time of 2 minutes and 45 seconds. No stirring or agitation was applied during the brew. The coffee was set to cool for approximately 1 minute after brewing before the start of the experiment. Two separate beakers and pipettes were set up in place, one containing distilled water and the other tap water (referred to throughout as “brew water”) to zero both refractometers at the start of each round prior to the TDS recording. Each round consisted of 4 TDS readings in this order: VST coffee sample TDS measurement with the device being zeroed on distilled water; Atago coffee sample TDS measurement with the device being zeroed on distilled water; VST coffee sample TDS measurement with the device being zeroed on brew water; Atago coffee sample TDS measurement with the device being zeroed on brew water. Prior to every zeroing and TDS recording for each round (10 total rounds, 40 total readings), the lens for each device was cleaned with alcohol wipes. As a sample of coffee was drawn from the carafe, the brew temperature was taken (and verified with multiple devices). The temperature of the sample was taken again when it was transferred to a ceramic cup before being dispensed on the refractometers.

study 1 design

Testing procedure:

Prior to each round, the brew was stirred in the carafe and the temperature was taken and recorded. For each ceramic cup, enough coffee was drawn for all sampling in the round (4 total). Further, the pipette was rinsed with the coffee solution, as per VST recommendations. Two new pipettes were allocated for each round to siphon the coffee inside the ceramic cup. All samples were rested on the refractometer lens for approximately 20 seconds before measuring the TDS. It is important to note that the Atago refractometer took longer to display the readings.

Study 2: Impact of the VST Syringe Sample Filters

Another pour over coffee was brewed at 94 degrees Celsius at a ratio of 1g coffee to 17g water. The brew was 15g of coffee to 220ml of water, with a total time of 2 minutes and 45 seconds. No stirring or agitation was applied during the brew. The coffee was set to cool for approximately 1 minute after brewing before the start of the experiment. Two groups of sample syringes were set up: one group fitted with the sampling filters; one group not fitted with the filters. Devices were zeroed using the brew water. Each round consisted of 4 TDS readings in this order: VST coffee sample TDS measurement with the sample being passed through a VST sample filter; Atago coffee sample TDS measurement with the sample being passed through a VST sample filter; VST coffee sample TDS measurement with no filter on the sample; Atago coffee sample TDS measurement with no filter on the sample. Prior to every zeroing and TDS recording for each round (10 total rounds, 40 total readings), the lens for each device was cleaned with alcohol wipes. As a sample of coffee was drawn from the carafe, the brew temperature was taken (and verified with multiple devices). The temperature of the sample was taken again when it was transferred to a ceramic cup before being dispensed on the refractometers.

It is worth noting VST states the filters should be used for “clarifying espresso prior to refractometer measurements. Also suitable for use with COFFEE ground for filter brewing using perforated metal or woven mesh filters.” As such, our test beverage brewed with a paper filter should not need the filters, according to the VST guidelines.

study 2 design

Testing procedure:

Prior to each round, the brew was stirred in the carafe and the temperature was taken and recorded. For each ceramic cup, enough coffee was drawn for all sampling in the round (4 total). Further, the syringe was rinsed with the coffee solution, as per VST recommendations. Two new pipettes were allocated for each round to siphon the coffee inside the ceramic cup. All samples were rested on the refractometer lens for approximately 20 seconds before measuring the TDS. It is important to note that the Atago refractometer took longer to display the readings.

Results

Study 1: Which Water to Zero Your Device

In the “instrument zero based on distilled water” condition, the VST average TDS was 1.33 with a standard deviation of 0.06 and the Atago average TDS was 1.36 with a standard deviation of 0.03. In the “instrument zero based on brew water” condition, the VST average TDS was 1.35 with a standard deviation of 0.03 and the Atago average TDS was 1.36 with a standard deviation of 0.03. A generalized linear model was run with the independent variables (2×2 between variables: VST/Atago; Distilled/Brew Water; co-variate: Time/Temperature for each sample). There were no significant findings at the p < 0.05 level (Refractometer: F(1,8) = 4.24, p = 0.07; Water Type: F(1,8) = 0.13, p = 0.73).

VST v Atago_water type
Impact of zero point. Means with standard error shown.
VST v Atago_temp of sample_water type
TDS reading as time elapsed/temperature of coffee cooled.

Study 2: Impact of the VST Syringe Sample Filters

These samples were taken from a new batch of coffee, brewed solely with the same tap water used in the previous study.

In the “samples with syringe filter” condition, the VST average TDS was 1.42 with a standard deviation of 0.06 and the Atago average TDS was 1.42 with a standard deviation of 0.02. In the “samples with no syringe filter” condition, the VST average TDS was 1.46 with a standard deviation of 0.08 and the Atago average TDS was 1.47 with a standard deviation of 0.02. A generalized linear model was run with the independent variables (2×2 between variables: VST/Atago; Filter/No Filter; co-variate: Time/Temperature for each sample). A significant difference between refractometers was found, F(1,8) = 6.32, p = 0.04, as well as a significant difference in TDS with the use of a filter, F(1,8) = 10.3, p = 0.01, and a significant interaction between the use of a filter and the temperature of the sample, F(1,8) = 11.2, p = 0.01). Examining the interaction more closely, as temperature cooled, the impact of using a filter decreased (this appears to be driven mostly through variability in the VST readings at warmer temperatures).

VST v Atago_filter
Impact of VST’s syringe sample filter. Means with standard error shown.
VST v Atago_temp of sample_filter
TDS reading as time elapsed/temperature of coffee cooled.
Exploring the Interaction Effect of Filter and Temperature on TDS
The interaction of filter and temperature on TDS.

(Raw data can be downloaded in a tab delimited text file here. As always, while we offer the data for your personal use, we kindly ask that you send a message to socraticcoffee@gmail.com before posting or presenting it in any public forum and attach appropriate acknowledgement.)

Conclusions

Overall, it appears the VST and Atago refractometers are very comparable for measuring the TDS of filter brewed coffee. While the second study examining the impact of the sample filter demonstrated a large amount of variability in the VST readings compared to the Atago, this was not seen in the first study (distilled/brew water; VST standard deviation 0.04; Atago 0.03).

Within a device, there was not a significant difference using distilled water to zero versus using brew water. However, it is worth noting that the brew water for our experiment was much lower than others may use (94 ppm). The purpose for which a person uses a refractometer with his/her coffee may influence the best approach to zeroing their device. It is possible that brew water could be of high enough mineral content to register on the VST or Atago refractometer and that brew water TDS levels may vary from time-to-time and/or place-to-place. Because this is the case, zeroing based on brew water would allow for a useful, relative metric with TDS (e.g., 1.35% zeroed on brew water would mean 1.35% to anyone else who also zeroed on brew water). It is solely a measure of the soluble content contributed by the coffee. This differs from zeroing based on distilled water, which does not allow for a differentiation between soluble content of initial water and soluble contribution from the coffee itself. Further, as zeroing on brew water does not alter the device hardware but instead changes the software set point for the case-at-hand (and can be returned to “absolute zero” with distilled water at any point), the brew water zeroing approach has a practical justification. The argument that a non-distilled water zero will decrease accuracy of the device is not supported in our data and we welcome open and honest data that shows otherwise. While perhaps only extreme brew water mineral content start points might register on these devices and/or, in psychophysics terms, breach the “just noticeable difference” in human perception, until objective empirical data demonstrates a substantiated issue caused by using anything-other-than-distilled-water to zero your device, we argue a good practice would be to zero the device based on the true starting condition (i.e., brew water).

Using the VST sample filters significantly decreased the TDS for both refractometers, but their use created greater variability in the VST readings. Our data cannot suggest whether a person should use or not use the filters. While their purpose is for beverage preparations other than the one used in this experiment, they do appear to have a significant effect on overall TDS (decrease of soluble content). How that improves the accuracy of the TDS reading is not clear. Again, from a practical perspective, the decision to filter out particulates of the solution being measured will most likely depend on how a person intends to use the refractometer. If the intent is to assess the soluble concentration of a brew as-is, no filter seems logical. If it is to assess the soluble concentration of a brew, cutting off particles of a certain size, the filter would be appropriate. This experiment did not assess readings from espresso, which may be more heavily impacted by the use of a filter. Many methods of coffee preparation, no doubt, include varying degrees of sediment and colloidal material. Whether or not it is “accurate” to remove particles before assessing the TDS of coffee/espresso seems, at this point, to be more of a subjective decision.

Overall, it is important to understand what refractometry does and does not tell you about the coffee. Because the light is manipulated as it passes through a substance at a rate proportional to the substance’s concentration in the solution, refractometry in coffee, which is a mixture of many compounds, can only give a metric for the cumulative effect of the various constituents. Essentially, the refractometer gives you a glimpse into how much stuff is in there, but nothing about what that stuff is composed of. Further, temperature is clearly an important factor for refractometry (i.e., temperature changes the density of water, which changes its refractive index). It is also an important factor for the refractometry of coffee, specifically, since some compounds may be soluble at higher temperatures but precipitate out of the solution as it cools. For now, the use of a refractometer seems best suited as a tool for quantitatively tracking coffee preparation consistency and, whichever methods are chosen (e.g., distilled zeroing, use of sample filter), it is important to maintain them for comparability.

References

Flament I. (2002). Coffee Flavor Chemistry. West Sussex, England: John Wiley & Sons, Ltd.

Andueza S., Maeztu L., Pascual L., Ibanez C., Paz de Pena M., Cid, C. (2003). Influence of extraction temperature on the final quality of espresso coffee. Journal of the Science of Food and Agriculture, 83, 240-248.

Clarke R.J. (1989). Water and Mineral Contents. In R.J. Clarke & R. Macrae (Eds.), Coffee: Chemistry (Volume 1) (pp. 42-82). New York, NY: Elsevier Science Publishers LTD.

31 Responses

  1. coffeecircus

    Thank you very much in carrying out this experiment!

    • It is great pleasure and thank you for taking the time to read the review!

  2. Tom Chips

    Thank you for all the work that went into this! Do you plan on doing a similar comparison with espresso strength extractions?

    • Many thanks for the feedback…hopefully in due course we will do it. In having said that We need to be able to control few variables, like tamping, brewing temperature etc…Challenging but feasible!

  3. Fantastic article. Thanks for the time dedicated to this!

    For Atago, it seems that temp has little impact on TDS. Correct? Or would there be a temp best suited for TDS reading?

    • According to the above data & graph, the TDS (Atago without filter) reacted differently to the temperature, it increased at 45.2 Degrees and then slightly decreased and thereafter maintained consistency in the reading. It appears that lower temepratures yielded more consistency in the TDS reading for the Atago without filter.

  4. cembozkus

    jeremy
    enable permanent links from wordpress settings. so google can index your posts.
    great job. it will be massive.

  5. A clarification: The VST LAB refractometer in this comparison is a Misco Palm Abbe. The previous VST model was manufactured by Reichert (r2 mini).

    • thanks, chris. we’ll make the revision!

      • Just to clarify my own reply, all the VST LAB models (I-III) are Misco Palm Abbes. Before 2011 VST used the Reichert r2 mini, and that was sold at the same time as the LAB model for a short time. The early Atago iteration of the coffee refractometer you mention was developed while VST was consulting for George Howell Coffee Co before they spun off the refractometer/software line to VST.

  6. Very, very cool! Bravo to you and this site!

    I’ve been using an optical Brix refractometer for my espresso measurements, but have also been trying to use a conductivity-based meter for gauging TDS. I realize that many think (haven’t seen the studies) conductivity measurements are less accurate than reflective index (RI) ones….?????

    Wondering of you could do further experiments comparing conductivity vs. RI? For this you may want to make a dilution series, in DI water, for gauging linearity, or lack thereof.

    Mention of lipids was also fascinating. If lipids are dispersed as immiscible globules, they should contribute to the RI measurement, since they transmit light.

    Lastly, I found your findings on how VST filtering affects RI measurements interesting. Not sure what is going on here, but it may be that the VST filters are removing solubles???? Perhaps they stick to the filter, and/or adsorb to coffee particles???? A quick way to check this is to disperse of particulate colloids (carbon black–a.k.a.–soot) as a control, vary the concentration, then measure. Or better yet, add filtered coffee particles back to your coffee for measurement (maybe wash them?). The effect of dispersed particle on RI measurements is both simple and complex. To the first approximation, their presence should have no effect (don’t transmit light), but complications set in due to scattering (affecting the detector scheme; particle size, concentration and wavelength dependence, ….) and surface mediated phenomena.

    I really liked your conclusion about using RI as a way to gauge relative coffee values. Next stop–turbidity measurements? Heck, let’s drink more and better coffee!

    • dave — thanks for the reply! these are great ideas. would you mind shooting me an email (jeremy [at] socraticcoffee.com)? i’d like to bounce potential designs for future studies off of you.

  7. You say that the VST refractometer showed variability in the readings when used with a filter, but the variability decreased at lower temperatures. However, the temperature range in your graph is 63C to 30C. VST clearly states in its instruction manual that the sample must be cooled to 15C to 30C:

    https://cdn.shopify.com/s/files/1/0092/7622/files/Lab_Coffee_User_Guide_fb416a2c-be2e-4bd8-bde2-dbe5d3831a96.pdf?265

    (see Temperature Correction, page 13)

    If I’ve read your chart and your sample protocol correctly, it would seem that the impact of the filter decreased as the temperature approached the top end of the range recommended by VST. If that’s the case, it seems likely that the VST instrument would produce the expected results if it was used properly — i.e., within the recommended temperature range.

    Also, it appears that your protocol doesn’t match the protocol recommended by VST. You allow the brew to cool for 1 minute in the carafe, but you don’t specify whether the carafe is sealed to minimize evaporation (and it doesn’t seem that the coffee would cool much in that short amount of time anyway.) Then you draw four samples with the syringe and place them in a ceramic cup. But you don’t specify the size of the samples and the amount of time they are allowed to cool in the cup. Then you place individual samples in the refractometer wells and wait 20 seconds before taking a reading.

    I would think that some evaporation of the second two samples occurs in the ceramic cup as the first two samples are drawn, placed in the refractometer wells and allowed to cool for 20 seconds. That will affect the TDS readings. Also, the temperatures of the two sets of samples will be different. Finally, VST recommends that the sample be allowed to cool for 30 seconds after being placed in the well, not 20 seconds. The time isn’t so much the issue as ensuring that the sample is well-cooled before being placed in the well and that the temperature of the sample and the well have time to equalize within VST’s recommended range.

    For those who haven’t read the VST instructions, the protocol goes something like this:

    1. Brew coffee
    2. Pour coffee into a small cup and allow to cool for 1-2 minutes
    3. Draw 3ml-5ml of coffee with the syringe and place in a heavy-bottomed shot glass
    4. Allow the coffee to cool for 30 seconds in the shot glass (swirling helps)
    5. Use the pipette to transfer a few drops of coffee to the sample well (just enough to cover the glass — a large sample will take longer to cool)
    6. Allow the sample to cool for 30 seconds
    7. Take the reading

    The objective is to get the sample cooled to the 15C-30C operating range of the refractometer with minimal evaporation.

    As for the filters, VST doesn’t recommend using filters for brewed coffee. It should be harmless to do so, but again you have to observe the recommended temperature range of 15C-30C.

    If I haven’t misunderstood your charts and protocol, I believe you should run your tests again using the recommended VST protocol and temperature ranges. If the results differ, please correct your post.

    • Hi Dick,

      Thank you for the great input. I agree, we should re-run our experiments with some modifications to our methodology. We have the newest VST refractometer on its way and are eager to test it out. For the next study, we will follow the VST recommended guidelines more closely. However, even if that yields different results, this current post describes the methodologies we used, meaning our results are valid within those constraints/caveats. It is a presentation of how two devices manufactured to do similar things performed within the methodologies we describe. Variability in the readings by the VST was observed and did seem to decrease as the temperature cooled. Again, we will do another experiment that more strictly follows the VST recommended guidelines. For example, as you mention, the filters “should be harmless”, so we’d like to take another look at that as they, based on this study, did appear to impact soluble content.

      Appreciate your feedback,
      Jeremy

  8. Your protocols (incorrect zero-ing, not observing temperature correction requirements) in the test somewhat negate the quest for accuracy. There also seems to be confusion on your part as to the significance of non-dissolved solids – extraction yield is always the measurement of dissolved solids only, it always has been (for the last 60yrs).
    Can you re run the test observing the established protocols?

    • Hi Mark,

      Thanks for the comments. We will be running additional experiments, which will include the new VST device. As we stated, the purpose here was to assess precision not accuracy, as that would require us to compare against another metric. Fortunately, our lab may soon acquire a Thermo Scientific Vacuum Oven, so we could potentially address accuracy. Your definition of extraction yield is interesting–I just performed a quick look at some peer-reviewed articles examining extraction yield for a variety of things (e.g., a banana peel in Comim et al., 2010; Origanum vulgare extraction in Rodrigues et al., 2004) and the standard approach appears to be weight-based change. If that’s the case, it’s unclear to me how that method truly only assesses dissolved solids, but I welcome any information you may have that we might incorporate into future experiments.

      Thanks again,
      Jeremy

      • Presence of particles during RI measurements is a bit complex:
        http://www.horiba.com/fileadmin/uploads/Scientific/Documents/PSA/Webinar_Slides/TR019.pdf

        Jeremy, I applaud your efforts here to assess precision of coffee measurements, rather than accuracy. The best way to measure accuracy is to use known standards–e.g., a dilution series using sucrose. This will lets you assess an instrument’s accuracy, without having to worry about what your coffee analyte might be doing. Coffee measurements would follow using a dilution series, with caveats applied to solubilities of coffee components, etc.–centrifugation of coffee prior to measurement might be something to consider, as a way to suss out issues with filtering: spurious adsorption of coffee solubles. (Wow, this gets complicated!) In any case, cross comparisons between the VST and Atago instruments, using a coffee dilution series series would likely prove very informative.

        I’m not a food scientist, but weighing dehydrated coffee seems tricky to get VERY accurate results. The hard part is to achieve constant weight, with knowledge that volatile coffee compounds remain, or that volatiles are released in a constant way (bean to bean, brew to brew, etc.), or that these components are a minor factor here???? Also, absorbed water can prove to be very persistent, so that gravimetric methods may provide different results, depending on how the coffee was dried. Lyophilization may be the best way to dry coffee, as compared to a heated vacuum oven????? Another way to is air dry coffee, then place in a desiccator until constant weight is achieved?????

        http://www.ecs.umass.edu/cee/reckhow/courses/572/572bk15/572BK15.html
        http://zimmer.csufresno.edu/~davidz/Chem105/DandW/DryW2.html

        For these reasons and more, I really like relative analytics (as presented here) and find “absolute” measurements of complex analytes, like coffee, to be very difficult in the absence of sophisticated techniques, or some standard operational protocol, which black boxes errors.

        • Thanks, Dave. Completely agree. This gets more and more complex as you delve into the nuances of solubility assessment and the relative approach presented here seems more practical for our concerns. With this post specifically, we’re not out to verify the accuracy of a device, but rather attempting to quantify how different devices marketed to do similar things respond to the same stimuli.

      • Hi Jeremy,

        Extraction yield protocol for coffee was described by E. E. Lockhart in CBI pamphlet “The Soluble Solids in Beverage Coffee as an Index to Cup Quality” (1957), reprinted by the CBC in 1969 as “Coffee Solubles and Beverage Acceptance”.

        Regards, Mark.

  9. Hi Jeremy. Thanks for taking the time to reply to my comments.

    I’m glad to hear you’ll re-run the tests when you get the Lab III. I’m looking forward to reading about the results.

    One item that I forgot to mention in my post is that VST instructions state that the temperature of the calibration solution and the refractometer must be the same (easily accomplished by storing then together), and that both must be in the recommended temperature range of 15C-30C. The sample has to be in that range, too. The implication is that the instrument won’t be properly calibrated unless the zero reading and sample reading are done at the same temperature, and that the temperature has to be in the range for which the device has been designed to achieve maximum accuracy and precision.

    I understand your point about the results being valid for the conditions used, but would counter that if the test conditions are outside the range of published recommended operating conditions for the device, then the results don’t reflect the accuracy or precision of the device. It’s a little like trying to verify the manufacturer’s highway gas mileage figure when driving in city traffic (there are better examples, but you get the point.) As such, I don’t see how the results are meaningful. Worse, the results are likely to be misleading to people who may base purchasing decisions on your review.

    • Hi Dick — those are valid concerns and we will do our best to properly address them in the next iteration. This first effort may be more akin to how a cafe or individual might actually use these devices (e.g., possibly not following every detailed step provided by VST to control for evaporation, temperature, etc.). We attempted a pragmatic approach here but will do a more controlled one next. It is interesting to note that Atago does not seem to dictate as rigorous of an approach to assess a brew’s solubility content. And maybe the problems with this less-controlled approach will become more clear in other assessments (e.g., espresso). Further, even though the VST may have been tested outside of its recommended temperature settings, I don’t think it performed poorly. In my mind, these experiments seemed to show that both devices are comparable–even at the temperatures we used for the assessments. When statistical significance does arise, it is still useful to consider whether or not that finding has practical implications. For instance, a statistically significant 0.01% TDS reading difference between the two devices should be weighed against the purposes for which you plan to use the refractometer and whether or not that difference is “worth it” (using whatever criteria you deem important). As for the meaningfulness of our results, that is totally up to you! We’re just happy to present some objective data for people to consider (within the appropriate constraints that we laid out) and we’ve tried to do so in an open and honest way.

      Jeremy

  10. Hi Dick,
    About 3 months ago we posted on Instagram (https://instagram.com/p/yO_O4LSuKX/) where we have quoted the following “All starts with cleaning the prism and I am going all the way on this test…the temperature of the distilled water used for zero-setting should be the same as the ambient temperature! (let us eliminate excuses)
    the measurements were taken with Amprobe and verified with my infra red gun and the room temperature is set automatically by the air conditioner in the lab!”
    Be assured that we follow the above procedure prior to every refractometer reading.

    Thanks

  11. […] from our previous post on refractometer comparisons, the goal of this study was to examine the precision of three refractometers advertised to measure […]

  12. […] But it is also helpful to look at findings over multiple studies. Looking at our previous posts (here and here), some significant differences between devices were noted. These differences were seen in […]

  13. Matt Perger

    Hi Gents

    Perhaps a good comparison of precision would be to see how each refractometer performs when subjected to a long sequence of samples with small but linear jumps in TDS.

    Brew 2L of coffee
    Portion into 20 x identical 100g samples
    Dilute each portion with 1.00g more water than the previous.
    Measure every sample with the different refractometers.

    Posit – the atago has a lower resolution sensor than the VST, and will likely display groupings of results at each step of its resolution, rather than a smooth linear response.

    To the boundaries of coffee,

    • Hi Matt,

      Thanks for the comment. That’s a great idea that we might want to consider in the future. I think it really goes back to our intent for the study. If our desire is to assess the validity of a manufacturer’s device specifications, then I think your approach would be the way for us to go. What you propose is, essentially, a dilution series, which we have done using an Atago device (not the same we we typically utilize). Our dilution steps were much bigger than what you are proposing, but, again, our intent was different:

      Brix v. Instant Coffee Dilution
      Brix v. Sucrose Dilution
      TDS v. Instant Coffee Dilution
      TDS v. Sucrose Dilution

      For this particular write-up, our approach was as any end user looking to get an idea of the dissolved solid content of their coffee beverage using a refractometer–do the readings from these devices differ from each other? Would a user utilizing one device get readings that significantly differ if he/she had been using the other device? We analyzed the data from an inferential statistics perspective to determine potential statistical differences in the readings, but then there’s the bigger “practical significance” perspective. This was just to help consumers make more informed purchasing decisions. Some people want/need a device with higher specifications. Nothing wrong with that. Depends on your end goal, your understanding of the metric and its limitations, etc. As for our lab, we just skip these precision concerns and utilize the Atago RX-007α :).

      Best,
      Jeremy

  14. Chris Ptacek

    I am equally grateful for you folks putting in the legwork, as I am for the (absolutely necessary) critiques of the process. You and the commenters, i. Some tiny little crevice of the internet, are pushing coffee forward. Improving best practices benefits us all, from people in the industry, to consumers. So thanks for your curiosity and initiative.

  15. Hey Guys,

    We found some differences in case the Atago and LabII. At the moment we have some trouble to decide which the right one is. Seems the results are same as yours but we haven’t so deep analysis like you guys.
    I like your measurement process, especially you have more samples to validate the precision.
    Next time I recommend to use 5 devices from each type. Or at least 3.
    I know it is not easy, time and cost effective sure, but I am interested if this results are coming from 3 unique equipment (like the 3 you have used) or these are a kind of device type specific measurement results.
    I mean like eg. each Atago device values (mean) 0,1-0,2 lower compared to Lab2 values (mean).

  16. Hello,
    Thank you for your post. I want to buy refractometer and learn how to use it properly and I have had dilemma which one to choose – after lecture your article I tend to buy the Atago stuff. Great job!

    • Thanks! With either refractometer you can’t really go wrong. The important point is to be consistent in your testing procedure.

      Best,
      Jeremy

Leave a Reply to Jeremy Cancel reply