Hmm, good question. I always assumed it was 14g. Maybe Eric could shed some light on the issue, or I could weigh out a basket with 14 grams in the puck, pull a 2 ounce shot then weigh the basket again to see how much weight it gained, minus the dissolved solids.

Well, I just weighed out a 14 gram dose on my 0.1g scale. I pulled a 2 ounce shot by weight then weighed the basket after the shot. It gained 17.2 grams which equates to 0.6 ounces. So my 14 gram puck that had a 2 ounce shot pulled through it gained 0.6 ounces in water. Now that was only one shot and one measurement which is statistically insignificant and does not take the extracted solids into account.

AndyS wrote:1. It is MUCH easier for brewed coffee, if one has the right TDS meter! I don't, mine lacks the required sensitivity by a factor of 10.

Excuses; excuses! Just brew ten times as much coffee, then boil to reduce the volume to a tenth ;P

Great work, as usual, Andy.

Cheers,

Luca

cannonfodder wrote:I have a Eutech instruments/Oakton instruments TDSTester11 which has a resolution of 10ppm or 0.10ppt with a range of 2000 ppm and 10.00ppt with a +- or 1% of scale. Unfortunately the scale you need is ppm which leaves you with 10.0ppm resolution. Don't know if that is any better than what you have but you are welcome to use it if you want.?

You are very generous to make that offer, thank you. I'm gonna buy one myself because you never know when your life might depend on an accurate TDS reading.

cannonfodder wrote:In your test, did you subtract the water hardness from the TDS to get actual TDS from the extraction or is that extraction TDS plus water hardness?

I didn't subtract it because the crummy TDS meter I used read zero when dipped in plain water.

King Seven wrote:
The maths for brewed coffee would go something like this:

<snip>

the 2ml/g retained seems to have been well documented since the CBC.

<snip>

With espresso I guess the maths is similar - not sure about water retained by the coffee off the top of my head though.

James, thanks very much for the clear explanation. What I'm fuzzy on is the relationship between "solubles" and "insolubles." Last year when Jim Schulman and I did our studies on solids extraction in espresso, we measured the total amount of coffee material removed from the puck. This presumably included both solubles and insolubles (I guess insolubles might include suspended solids and oils).

On the other hand, the method you present appears to consider only soluble matter, which I presume is what a TDS meter measures exclusively.

I suppose it's possible that in filtered, drip coffee the amount of insolubles is very small, but as you raise the brew pressure and/or use coarser filters the amount might increase to a significant amount.

Am I making any sense? It's late and I'm going to bed now (more excuses).

BTW, I just measured an Aeropress puck before and after: 6.5 g before, 15 g after.

I think the problem with the espresso puck and the amount it absorbs is the sucking effect of the solenoid valve.

I think with espresso drying pucks is the only way to get an accurate read on % extracted. That said, with a TDS you could reverse engineer the amount of water used with some basic maths and then maybe link it in to an accurate volumetric pad and see if you can set up a constant water amount, at which point you could use that and TDS and know your extractions.

Does that make sense?

(still struggling with my Chemex and some TDS measurements)

As for undissolved stuff - probably more important with espresso as I'd guy paper filtration removes a lot of the undissolved. That is pretty much guesswork though...

King Seven wrote:I think with espresso drying pucks is the only way to get an accurate read on % extracted.

I think you're right. With espresso, meters are quick, easy, and probably not very accurate. On the other hand, drying and weighing pucks is accurate but really tedious.

In other words, the tedious method is more accurate than the TDS method.

AndyS wrote: several group members were extremely skeptical that coffee could be properly brewed in the Clover's abbreviated time cycle.

The thing with the Clover is that people who taste it in coffee shops, like Grumpy, have not tried the same coffee brewed traditionally using a press pot (with no press) such as the Eva Solo or even straight filter brew.

On a Visit to George Howel's Terroir, Jim Schulman, George Howel, Peter Lynagh & I did that experiment. We brewed the same coffee using the Clover against the Eva Solo and the Technivorm. The Eva Solo came first, the Technivorm second and the Clover came last by a large margin. Now, does that prove that the Clover will always finish last? No. But it is the first time I had an opportunity to do any impartial comparison. With so little independent research on that machine, it will always be treated with suspicion.

Nice Work.

AndyS wrote:Sometimes the hardest step in doing an experiment is deciding WHAT IT IS that you're going to test.

Well, that's science. And the science of coffee is certainly in its infancy. What I'd like to ask in relation to this (and its probably been asked before) is WHAT PARTS of the extraction make up WHAT PARTS of the brew (espresso in my case). I.e., are the early moments of the extraction all sweet and rosy, and the later moments bitter? When is the caffeine extracted - is it linear or curvilinear (negative, positive, sigmoid)? How about the inner vs. outer surfaces of a coffee bean? The point is a very fine grind might be extracted all the way through, whereas a coarser grinder would not.

PC

cannonfodder wrote:With the Scace device, you pull 3 ounces of water to simulate a 2 ounce shot assuming that the coffee cake will retain one ounce of liquid. I have not done any kind of test to verify but if Greg says it is so, I am inclined to believe him.

Recommendation is 75ml to simulate 60ml shot volume. This was based on a series of extractions I did during the development of the USBC testing protocol where I weighed baskets before and after extraction. I don't recall the dose weight offhand, but I'm pretty sure the retention was approximately 1ml/g of grounds, so the dose would have been in the 14g to 15g range.


Hey! I found the data!

The test was done on 12/17/04. Dose was approximately 18g and water retained was approximately 17.5g. We normalized the amount to 15ml to be in line with the 7g - 30ml single and 14g - 60ml double, also taking into consideration the test measuring vessels were most likely to be marked in 5ml increments rather than 1ml increments.