Well a milligram is 0.001g, a tablespoon is 15g.
At 60g/l, each tablespoon of beverage could conceivably represent 1g of dose. So 0.19g of dissolved solids = 19% extraction yield...and you still have another decimal place to work with.

MWJB wrote:...So 0.19g of dissolved solids = 19% extraction yield...and you still have another decimal place to work with.

+/-.5%

You can find .001g scales with a tolerance of +/-0.003, even 50g @ +/-.001g if you want to spend a reasonable amount.

Certainly slower & more hassle than a VST refractometer.

MWJB wrote:You can find .001g scales with a tolerance of +/-0.003, even 50g @ +/-.001g if you want to spend a reasonable amount.

Certainly slower & more hassle than a VST refractometer, but not a whole lot less accurate...assuming brew water/beverage/dose weights are accurate to the same degree?

I have a Ohaus 100gx0.01g scale. and yes, boiling the coffee is a pain and not fast.
I can recommend the bottom half of an aluminum can to hold the sample, though.
VST on the way.

Apologies, yes I'm with you now, sorry I thought we were discussing JPender's milligram scale.

endlesscycles wrote:I also don't know what kind of precision scale you have, but I don't think a tablespoon is enough. I'd think with a .01g +/-.005g scale, you'd need to start with 60g (four tablespoons, ish) to get anything meaningful.

I have a very inexpensive milligram pocket scale (20g x .001g). You can buy one on amazon for about $22. It is flimsy and requires a simple modification straight out of the box in order to function, but when operated with appropriate care it has a precision of about ±2mg. There is a writeup on amazon by someone who describes both the required modification and his testing of the scale against a laboratory balance. I also tested mine against a 0.1 mg resolution Mettler balance and found that it performed similarly to what the guy on amazon reported.

Once other factors are controlled reasonably well the precision of the coffee strength is determined primarily by two scale measurements: (1) the dried coffee + weighing pan; (2) the weighing pan. With a 13g sample size and a ±2mg scale precision the worst case variation would be ±4/13000 = ±0.03% TDS.

As a test I brewed a batch of coffee and took ten samples ranging from 12.8g to 13.9g. The measured strengths (in %TDS) were 1.62, 1.61, 1.61, 1.62, 1.61, 1.63, 1.63, 1.61, 1.61, 1.61.

To test accuracy I arranged with someone who was willing to measure samples for me with a VST Lab refractometer. I prepared samples from seven different coffee batches and shipped them priority mail. Meanwhile I dehydrated samples of each of the seven brews. Six of these matched the refractometer to within 0.02% TDS. The seventh (at about ~4.5% TDS) was contaminated by a haze of unknown origin and differed by about 0.1%.

So I think a tablespoon is enough. In fact Vince Fedele recommended the scale precision and sample size that I am using for this purpose.

I would research refractometers before making any purchasing decisions. Consider ease of use and accuracy.

gt wrote:I'd be interested in the "additional details you've worked out".

When I've done drying in the past, I was stumped by which end weight to use - like right out of the oven or cooled to room temperature.

Right out of the oven isn't a good idea because a hot sample will tend to read low due to induced air currents. If your sample is in a sealed container there will be an additional problem with buoyancy.

Moisture absorption: When I was attempting to determine extraction by drying spent grounds this was a bigger problem. I found that exposed grounds would increase about ~2% in weight after only 10 minutes out of the oven. Lacking a dessicator I started partially sealing the grounds by pinching closed the foil trays I was using. This reduced the initial weight gain by a factor of 10. The partially sealed grounds also attained a relatively stable weight for as long as 90 minutes whereas the exposed grounds continued to gain weight with no apparent end in sight.

Dried liquid coffee is lot less hygroscopic. And the smaller samples cool faster. I have found that by weighing them promptly after they cool there is negligible weight gain. Overnight I've seen them increase by ~5% net weight but in the minutes after cooling it is usually not a problem. Your mileage may vary.

As a side note, drying spent grounds wasn't a very satisfactory method for me for a number of reasons. First, there is the hygroscopic issue. Second, there is some difficulty in collecting all of the grounds. Third, they stink like garbage in the oven. Fourth, it is a measure of total solids, not dissolved solids. Fifth, only one measurement per brew is possible. And finally, since it is a determination of extraction by difference, the moisture in the freshly ground coffee introduces a very large error, one that varies with coffee and time. In order to obtain a meaningful result a sample of fresh grounds also has to be oven dried.

Drying 100g of liquid coffee works with a less precise scale but it takes a very long time. It also consumes a substantial fraction of the beverage that I'd rather consume myself. And for those of us without a centrifuge it is less practical to filter out undissolved solids in a large sample.

What I do:

1. Extract ~15ml of hot coffee from the cup into a large syringe
2. Cap syringe and cool (to limit evaporation during weighing)
3. Attach syringe filter if deemed necessary
4. Press 12-15g of coffee from syringe into a small plastic (~4g) cup
5. Weigh cup+sample (total has to be less than 20g for my scale)
6. Pour sample into handmade, pre-weighed foil tray (typ: 5x10cm, 0.8g)
7. Weigh emptied plastic cup
8. Place sample (on cookie sheet) in gas fired kitchen oven set to 215°F
9. After about 60 min check to see if it appears dry
10. Take out sample, weigh it, and return it to the oven
11. Repeat step 11 until weight doesn't change (90 min is usually adequate)
12. Pinch foil closed 5 min prior to final removal from oven
13. Remove from oven, place under small bowl to cool for a few minutes
14. Weigh it
15. strength = (dried sample in foil - foil) / (full cup - emptied cup)

Notes:

I use an inverted clear plastic weighing bowl as a draft shield. Since the scale sometimes "misbehaves", I frequently test it by weighing a 1000 mg calibration weight. As long as it reads between 999 and 1001 I'm happy. If it is off by 2 mg or more I tare the scale, let it settle, and try again.

Caveats: I use my oven (GE XL44 gas) in my kitchen environment (typ: 60-70°F, 50-80% rel hum). There is no guarantee that this method will work for someone else without modification. The basic idea is pretty straightforward though.

Thanks jpender for supplying the details.

I looked a my old notes and in round numbers I was getting a 4.5% increase in weight from right out of a 180F oven to several hours at room temp. I never felt comfortable with which weight to use.

Your sample size of 12-15g renews my interest in doing this again. I was evaporating large full cup samples and it was a real pain as anyone knows who has done it.

Edit:

In looking at my notes some more, the 4.5% is what I found when drying the grounds and paper filter.