I use volume unless I am working on a new coffee or doing testing where I need super accuracy. On very dry days I get more static in the grind which will make it fluffier but I know that as soon as I start thwacking the doser and compensate for it. I will occasional get out the scale as a reality check, especially if something is a bit off. As a general rule I fall within 0.2 grams of my target pretty consistently which is good enough for me for my daily shots. You eventually learn the coffee and grinder and recognize when the grind is a denser or fluffier for a given blend. I dont think humidity has an effect on the weight of the coffee in the manner you are suggesting. As Greg points out it is a very minuscule amount. I do however think that the ambient humidity will effect the static in the grinds which will make them fluffier or denser. So volume dosing can and will vary now and then, but a gram is a gram no matter how fluffy the grinds are.

I weigh every single shot when I'm using a doser, so I have a fairly decent idea of how volume corresponds to weight. I still get variations of up to +/- 0.5-1g, and I am sure most professional baristas working in a cafe aren't much better. I recall a post by Jim about visiting a cafe and noting how inconsistent the shots were from pull to pull, consistent with variations in dose weight.

I've found with timed grinders, the amount of time you wait after grinding the previous shot affects the weight of the next shot. If I base my timing on a shot ground directly after the previous one (while dialing in the grinder), then it will overgrind for shots pulled several minutes apart.

When using a timed dose, a short purge grind can yield very consistent results. I've measured ±0.1g over 15 shots over 5 days on the MC4. Typical tolerance is about ±0.2g on both the MC4 and MXK when conditions are relatively invariant.

You'd think that someone would build a scheme by which a portafilter was installed into a stand attached to a load cell, with the load cell getting tared when the grind switch was activated, and shutting off the grinder when X-amount had been dosed.

-Greg

another_jim wrote:Timed grinding is a very bad way to dose consistently. Grind rate is affected by the bean type, how full the hopper is, the grind setting, and accumulating operator errors (every time a previous error changes the flow rate, the operator changes grind setting or grind time with no real clue why the flow changed in the first place).

Volume dosing is much more consistent, good to about 1/3rd gram, when the grounds are compressed in a standardized way before dosing, as Greg describes.

There is a small and a big drawback to volume dosing.

• The small drawback is well known, you get the volume of the basket, and to change it you need another basket or a set of curved leveling tools.
• The big drawback is that the combination of fixed volume dosing and grind setting changes is an unstable equilibrium. If you make the grind finer, you also increase the packing efficiency and effective density of the coffee powder. This means tiny grind changes lead to excessively large changes in flow, and make adjustments far more difficult than they need to be. I'm pretty postive that in Italy, where people use full dosers, they compensate for flow changes by making a dose adjustment turning the doisng screw, not a grind adjustment. If you use a sculpted dosing tool set, going one size up or down, may be better than adjusting the grind

Of course, all these complications go away as soon as you weigh doses. Espresso dosing is basically in the dark ages, what other commodity is not dispensed by weight? I dislike the extra work required by weight dosing, and I regret that it cannot be used commercially. How difficult is it to build a grinder that single dose grinds and properly dispenses a pre-weighed dose of whole beans quickly and reliably? The Versalab M3 is proof of concept, a viable design needs to be more reliable and faster, but should be quite doable.

... with a diverter to discard the first n msec... and a spreader apparatus for a perfectly distributed dose

gscace wrote:You'd think that someone would build a scheme by which a portafilter was installed into a stand attached to a load cell, with the load cell getting tared when the grind switch was activated, and shutting off the grinder when X-amount had been dosed.

-Greg

I can see that happening. Although, it would have to be a little more "computerized" than just a tared scale attached the the PF holder. There would have to be a program that would sense the rate of change in weight and stop the grind so that the grounds still falling would finish the dose. Or there could be an offset much like PID's where if you wanted 18gm the grinder would stop at say 2gm less.

Sounds like a new addition to your already awesome Scace devices! The Scace grinder scale!

If you really think about what is being hypothesised here, one can certainly come up with a few variations on how to achieve good results of outputting the desired amount of grounds into the PF, through time or weighted dose.
I just want you all to ponder what industry standards achieve on a regular basis for the food or liquor industry. I work for a large company that does exactly what you are pondering, putting loose granular substance of precise size into packages, but, at say 100 times a minute, sometimes 400, and on a few select machines, its 800 a minute, into pouches or bottles. Dosage is exact, dialed in so that it hits a target, by weight. Overfill costs millions and I'll stop there as the analogy goes downhill.

It might be nice to contemplate a few group ideas that could eventually lead to a paring down to some diy solutions.

Here's one idea- my non digital grinder switch-
Using a see-saw type device, with the PF basket at one end, small weights on the other. The coffee exiting the grinder fills the basket until the weight of it tips it down (near 14g), whereby a switch is triggered and it shuts off the grinder. Tweaking the weight amount would allow for precision on the grinder kill. Viola.

Another -
electric eye cup-
The grinder fills a clear cylinder, that has an electrical eye passing through it, like on garage doors, only more precise. When the ground level is high enough in the tube, it triggers the eye and shuts off the grinder. Weighing the result and then raising or lowering the eye, yeilds the desired dosage. Of course, the cone shape of the fill would need a vibration pad under it, to keep the grounds somewhat horizontal.

enjoy-

...or you could use a Swift.