[MWJB]

Samples will suffer from evaporation, sample straight after brewing & record results. You want to correlate that reading with a change in taste for the coffee/change in method you just made, whilst drinking the coffee at normal temp, so doing it live would be an advantage?

 

[Zephyp]

Oh, yeah. I'm using the same recipe, so EY will be pretty much the same. Small changes in beverage weight can affect the result. With 1.41 and 209 BEV, I get 21.24%. I'll try even finer and see what happens. The recommended range for V60 with Comandante is 20-25 iirc, and I'm now at 20 or something.

Maybe I'll try AP on Kalita then. I got lots of filters, but it doesn't get much use nowadays.

Good point on comparing taste and EY at the same time. Could of course note taste while drinking and compare later. Maybe I'll try one day to test before drinking and leave a sample in a closed container for testing later, to see if I get the same result.

 

[MWJB]

Oh, yeah. I'm using the same recipe, so EY will be pretty much the same. Small changes in beverage weight can affect the result. With 1.41 and 209 BEV, I get 21.24%.

Even using the same recipe & grind, for a 200g average brew, you'll see maybe a 5g span of brewed weights. That's about 0.5%EY's worth at typical drip ratio. Weigh each brew.

Maybe someone can pour to the gram each time on brew water, I can't. Beverage mass (for drip) will vary slightly more than water input. Basically, when repeating brews you're going to get some variance and smallest to largest it looks like this:

Dose - no reason why this shouldn't be the same to 0.1g

Brew water stdev of ~1g for a 230g pour

Bev mass stdev of ~2-3g for 200g in cup

EY is the next smallest variance (no change in grind)

Brew water absorbed

Brew time

Draw down time after the last pour interval ends is the largest variance.

 

[Zephyp]

I noticed that even the infusion of the paper made a difference since I weighed the server after doing so, and the few drops that remained in the server added a gram or two. Last brew today I infused it over a cup and not the server I brewed in. Then I can use the same server weight each time and just subtract from BEV+server.

Last brew of the day: 4 clicks finer and I knew this one would make a difference. 208 BEV, 4:34 drawdown (at 4:10 it seemed to nearly come to a complete stop having a little water left). With this fine grind, I had to really stir the bloom. I saw when stirring that a few groups of dry grounds popped up. 1.47x4 TDS and 22.07 EY. I thought it had more fruitiness and distinction, but also introduced more bitterness and drying mouthfeel. I suspect the bitterness will increase if I push it further, and the brewing might come to almost a complete stop during drawdown.

In your experience, is there an EY range where certain origins taste better? I often see people going finer on Africans, which I suppose increases EY. In your log you seem to hit averages from 19 to 21 across different origins and methods. I started out at 20.7 and now tried 22. Something between there is probably ok, but I'd like to see what below 20 taste like.

 

[MWJB]

I noticed that even the infusion of the paper made a difference since I weighed the server after doing so, and the few drops that remained in the server added a gram or two. Last brew today I infused it over a cup and not the server I brewed in. Then I can use the same server weight each time and just subtract from BEV+server.

Last brew of the day: 4 clicks finer and I knew this one would make a difference. 208 BEV, 4:34 drawdown (at 4:10 it seemed to nearly come to a complete stop having a little water left). With this fine grind, I had to really stir the bloom. I saw when stirring that a few groups of dry grounds popped up. 1.47x4 TDS and 22.07 EY. I thought it had more fruitiness and distinction, but also introduced more bitterness and drying mouthfeel. I suspect the bitterness will increase if I push it further, and the brewing might come to almost a complete stop during drawdown.

In your experience, is there an EY range where certain origins taste better? I often see people going finer on Africans, which I suppose increases EY. In your log you seem to hit averages from 19 to 21 across different origins and methods.

OK so it looks like you have found a limit, back off a tad to drop EY & make wetting at the bloom easier.

'Africans' covers the whole gamut really. Kenyans & Rwandans may end up in the higher end, Ethiopians can fall anywhere. I don't change grind setting based on origin, only when dialling in a new grinder/brewer, or if my average EY is so high/low to have anumber of brews over/under extract. I change grind setting very infrequently, months between changes rather than days.

 

[Zephyp]

Should I adjust LRR to the ratio I usually get between BW nand BEV? Does it matter much what LRR is at if I don't change it? Default was 2.1, which gave BW of 235, when I pour 240. If I adjust LRR to 2.55, BEV and BW matches my brews. Changing it doesn't produce a significant change in EY, so I assume it doesn't matter much.

 

[MWJB]

Should I adjust LRR to the ratio I usually get between BW nand BEV? Does it matter much what LRR is at if I don't change it? Default was 2.1, which gave BW of 235, when I pour 240. If I adjust LRR to 2.55, BEV and BW matches my brews. Changing it doesn't produce a significant change in EY, so I assume it doesn't matter much.

If you have scales handy, weigh the brew & subtract server, don't use the LRR presets.

LRR is something I would use if I was looking for ball-park, or if weighing the beverage was impractical every brew (then I'd measure 10 brews & use the average). You might even find that changing the grinder changes LRR, as does changing grind setting/brew weights for the same brewer.

The important things you need to calculate the drip EY are dose, BEV & %TDS. I weigh each brew (to the g) & ignore BW & LRR, inputting BEV & %TDS.

 

[Zephyp]

Maybe I should probe the sample with a thermometer before testing. Yesterday I went 10 clicks coarser and today 4 more. First test today read 1.27, but the next was 1.31. The following tests climbed up to 1.34. Then I removed the sample, wiped the prism and added a second sample from the same cup I used to bring it down in temperature. That sample read 1.38.

Could you even be waiting too long, so that parts of the sample evaporate and the TDS effectively increase? Maybe I should draw a larger initial sample into the cup for cooldown.

Drawdown was 3:54, which was considerably longer than yesterday's 3:36, with today's grind being coarser. I wonder if the way I stir the bloom and amount of shake after the last pour cause these differences. I see what you mean by the drawdown having the largest variance.

Does the amount of sample I put on prism matter? Could different amounts provide different results?

 

[MWJB]

Maybe I should probe the sample with a thermometer before testing. Yesterday I went 10 clicks coarser and today 4 more. First test today read 1.27, but the next was 1.31. The following tests climbed up to 1.34. Then I removed the sample, wiped the prism and added a second sample from the same cup I used to bring it down in temperature. That sample read 1.38.

Could you even be waiting too long, so that parts of the sample evaporate and the TDS effectively increase? Maybe I should draw a larger initial sample into the cup for cooldown.

Drawdown was 3:54, which was considerably longer than yesterday's 3:36, with today's grind being coarser. I wonder if the way I stir the bloom and amount of shake after the last pour cause these differences. I see what you mean by the drawdown having the largest variance.

Does the amount of sample I put on prism matter? Could different amounts provide different results?

No idea why you are seeing such swings in the readings. I trust you are stirring the cup well before taking a sample for cooling?

0.3ml (0.2 to 0.4ml) is the ideal sample on the prism, but try and keep these the same for each sample. Look at the width of the band of coffee around the prism, on the steel dish.

 

[Zephyp]

Poorly executed I think. The next brew was more consistent, giving readings of 1.29, 1.29, 1.30, 1.29. I poured a bit larger amount into a bit larger cup, swirled it more and waited a bit longer. In the cup it was 33C, and after pipetting and dropping on the prism and waiting more, it probably got closer to equilibrium.

210 BEV, 14.45 dose, 1.29 TDS = 19.53 EY.

I thought the cup had less of the unpleasant characteristics, which brought forth more clarity, but was also a bit weaker. The grind is quite coarse, 34 on the Comandante, but still taking 3:35 to draw down. I guess it's the nature of some Africans to brew slower. It would've been interesting to have a different bean to brew every other cup with and see how it turned out. Something from South America. I'll buy two bags of different origin sometime and compare them as I go.

So far, I think the refractometer is an interesting gadget. Seeing the TDS compared to brew time and grind demonstrates that it can be difficult to compare two brews where you see differences, but don't taste too much of it. As you said, variance can be pretty high. I made two brews, one with setting 27 and one with 34, which had the same drawdown time, one with 20.6 EY and one with 19.6. Previously, I would've expected the difference in grind to show more on the brew time, but I see now that I shouldn't pay it too much attention.

 

[Nod]

Hi just got our new refractometer - exciting! I wonder if anyone has tried the Chinese VST equivalent filters? Can you use these cheap alternative to the VST originals? Thanks

 

[berthadee]

nice info...

 

[Zephyp]

A quick question about calculating EY for pour over.

I put the following data into the VST app:

Dose: 14.4g

BEV: 211g

TDS: 1.44

EY: 21.98%

How does it calculate that value and is it something I can put into my Google sheet? I tried BEV*TDS/Dose, but that gives 21.1 Is there some kind of factor in there that I'm missing? Since I only input those three variables, I was thinking the missing number might be a constant, or does it change depeding on the data I put in? It's not a big problem, but if I can avoid having to use the app to calculate EY every time, that would be nice.

 

[MWJB]

A quick question about calculating EY for pour over.

I put the following data into the VST app:

Dose: 14.4g

BEV: 211g

TDS: 1.44

EY: 21.98%

How does it calculate that value and is it something I can put into my Google sheet? I tried BEV*TDS/Dose, but that gives 21.1 Is there some kind of factor in there that I'm missing? Since I only input those three variables, I was thinking the missing number might be a constant, or does it change depeding on the data I put in? It's not a big problem, but if I can avoid having to use the app to calculate EY every time, that would be nice.

Just use the Bev*TDS/dose 21.1% calculation for your google sheet, if the figures are all the same convention then you are just looking at relative extractions.

 

[the_partisan]

A quick question about calculating EY for pour over.

I put the following data into the VST app:

Dose: 14.4g

BEV: 211g

TDS: 1.44

EY: 21.98%

How does it calculate that value and is it something I can put into my Google sheet? I tried BEV*TDS/Dose, but that gives 21.1 Is there some kind of factor in there that I'm missing? Since I only input those three variables, I was thinking the missing number might be a constant, or does it change depeding on the data I put in? It's not a big problem, but if I can avoid having to use the app to calculate EY every time, that would be nice.

It's because of CO2 and Moist variables, they're non-zero by default in the VST App.

The full formula is:

=BEV*TDS/DOSE/(1-%H2O-%CO2)

 

[Zephyp]

Thanks both of you. Since I will use it primarily for myself I'll just keep it at 0 and 0.

Been trying a few things and starting to see how big change the pour regime can create. Made a few cups the last days that I wasn't quite happy with, but today, when I of course didn't measure anything it was very good. Made a small adjustment on the grinder which I think is still within the 10-13%

 

[madrian]

Sorry for replying to this old topic, but this seem to be best place to ask about this refractometer. It's a 100$ refractometer from China. Anyone using it? How does this compare to the well known VST and others? I need it only for home use.

 

[MWJB]

Sorry for replying to this old topic, but this seem to be best place to ask about this refractometer. It's a 100$ refractometer from China. Anyone using it? How does this compare to the well known VST and others? I need it only for home use.

I have one, and an Atago and a VST.

The VST is really the only one with enough precision to meaningful with brewed coffee.

All 3 need syringe filtering for espresso/coffee with suspended solids.

If you want to go ahead with the Amtast for espresso, fine but just don't get too worked up if your shots seem less consistent than you are expecting. I found the Amtast sometimes read half a % low, then half a % high, then the same as the VST. Say you get a reading of 15/16%EY then you may want to act if a malfunction is confirmed by taste, but if your taste is good and it gives a reading of half a %EY to a whole %EY outside of what you expect, I wouldn't fret, not take it too literally and think you are hitting some hitherto unknown region of interest.

An aspect not often mentioned is that it's a lot quicker to take readings with the VST, but whether that is enough incentive for you to spend so much more is up to you.

With any EY readings, once you have got past any severe malfunctions in brewing, then keeping an eye on your consistency (correlated to taste scoring) is probably the most useful use of it, look at averages built up over time and standard deviations. Compare like with like (same brew ratio, same method re brewer, espresso, drip, immersion etc.)

 

[madrian]

@MWJB awesome, thanks for this write up.

 

[madrian]

@MWJBone more question: I calibrated using distilled water, reading 0.00. Then I measured regular tap water and also got reading 0.00 - is this malfunction? For coffee -espresso- I got pretty good reading, around 21%.

Is there any solution against which I can test? I mean XY gramm of sugar to XY gramm water should be read XY value.