Doser Accuracy

[cafeIKE]

How repeatable are grinder dosers when used as designed?

[another_jim]

I've heard around 1/3 gram; never tested it

[Fullsack]

Not exactly on point, but not worth its own thread.

I put 6 grams of coffee through my Macap this AM, brushed it out afterwards and weighed the grinds that were left in the grinder. A little over 1 gram remained. I clean the grinder after each session, so in the first grind of a session I am producing only 5 grams of coffee and in subsequent grindings, I am getting 6 grams.

[cafeIKE]

By "as designed" I meant grind enough coffee to fill the doser, catch several single pulls of the doser and weigh each dose.

An interesting side test would be to refill the doser and let it sit for 30 minutes, then repeat the dose weighing test.

I don't have a doser or I would do the test myself. Any Angelenos willing to supply the grinder, I'll supply the coffee and scale.

[HB]

How repeatable are grinder dosers when used as designed?

If you remind me after next week when I'm back in town, I'll test it with old beans. From what I've read, they are accurate to less than a gram when used as designed (i.e., filled well above the wedges). Some grinders like the Cimbali Max and Mazzer Kony have auto-dosers, i.e., the motor runs every 12 pulls to keep the doser full. The grinder's main switch can override this behavior.

BTW, is this a question of idle curiosity or is there a deeper meaning?

[cafeIKE]

BTW, is this a question of idle curiosity or is there a deeper meaning?

Much has been written of late of the need for a 0.1g scale.

For the past couple of days, I've been grinding as normal into the PF, then weighing before tamping. Doses of about 14.5g±0.5g are pretty consistent.

In bars where I've had the best and most consistent shots, the barista simply pulls the doser, tamps a bit, locks and pulls. Nary a scale to found. If dosers have a reasonable repeatability of say ±0.5g, then that more or less establishes the bench mark for consistency.

[HB]

If dosers have a reasonable repeatability of say ±0.5g, then that more or less establishes the bench mark for consistency.

That's certainly accurate enough for me.

In Exercises for tuning your barista techniques, I advocated within 0.5 grams, though I would not fuss too much if it's within 0.7 grams. Sometimes I'll intentionally shift the dose to change the pour time; my rough rule of thumb for most blends is that one gram adds/subtracts about four seconds of pour time.

[Psyd]

In bars where I've had the best and most consistent shots, the barista simply pulls the doser, tamps a bit, locks and pulls. Nary a scale to found.

No leveling or distribution involved? Most pros will dose by volume, using a version of Stockfleth's, the Chicago Chop, etc. Once you get the 'fill 'er up' method down, the doses will be fairly similar.

[cannonfodder]

0.5 gram my be fine for day to day use but when you are trying to compare shots between grinders, or machines, you need to be as accurate as possible. I have a 0.1g scale and only use it when I am doing bench testing or blending.

[cafeIKE]

No leveling or distribution involved? Most pros will dose by volume, using a version of Stockfleth's, the Chicago Chop, etc. Once you get the 'fill 'er up' method down, the doses will be fairly similar.

Asking for a 'Chicago Chop' in Italy will probably get you directions to the steakhouse in the nearest chain American hotel.

[Psyd]

Asking for a 'Chicago Chop' in Italy will probably get you directions to the steakhouse in the nearest chain American hotel.

Does that mean that they don't use any sort of leveling technique (what, the Pompei Push?) or that they do?

[HB]

Luigi Lupi was working the Elektra booth two SCAA conferences ago and kindly gave me a mini lesson. He chided me in broken English about fussing with distribution. To paraphrase: "Pull the handle twice, no toucha anything [slapping back my hand]. Now a little tamp using this [upward pull on grinder tamp]. Now putta the portafilter in the machine." That's what I observed when in Italy too.

[Psyd]

Luigi Lupi was working the Elektra booth two SCAA conferences ago and kindly gave me a mini lesson. He chided me in broken English about fussing with distribution. To paraphrase: "Pull the handle twice, no toucha anything [slapping back my hand]. Now a little tamp using this [upward pull on grinder tamp]. Now putta the portafilter in the machine." That's what I observed when in Italy too.

While I'm sure that this works in a crowded cafe with a full doser, how does it translate to the HB that is grinding per dose and thwacking till the doser's dry? I'm not sure that it doesn't, but my recent attempts at such things are leaving me with shots that blonde earlier than I'm used to, and spritze generously, and a very soupy puck. Granted, I'm not just pulling twice, but that's because my doser vanes aren't full. Is anyone having great results with the HB version of this? No distribution, no nothing? What do you do if your second pull puts a mound on one side of the basket? Does it just take care of itself?
I need a day with a coupla pounds to just sit and play and get really wide...

[cafeIKE]

Bump

I'm quite surprised that none have responded with any doser accuracy tests in light of the penchant here for 0.1 g/bar/° measurements.
What's a couple a bucks worth of coffee in the name of science

FWIW, in a 15g basket a measured 15.0±1.0g amounts to very little more taste difference in the cup than normal shot to shot, day to day variation, if one stops the shot on color, not time. Light tamp, no taps, twists or twirls Variance is far less than that of different baristas pulling the same shot in a good commercial shop.

Baseline established by a week of 15.0±0.1g doses followed by a week of 15.0±1.0g doses, randomly selected.

Coffee : Supreme Bean Caffe di Norte

[another_jim]

:arrow: Bump

I'm quite surprised that none have responded with any doser accuracy tests in light of the penchant here for 0.1 g/bar/° measurements.

I promised, but I'm away at the moment.

Otherwise, there's absolutely nobody on this or any other coffee board, amateur or professional, who actually uses a doser as it is designed, that is, kept filled with at least 7 grams times 12 doses, or 84 grams of ground coffee at all times.

[cafeIKE]

I promised, but I'm away at the moment.

Comment was directed at the other 1981 registered users. HB also promised

Otherwise, there's absolutely nobody on this or any other coffee board, amateur or professional, who actually uses a doser as it is designed, that is, kept filled with at least 7 grams times 12 doses, or 84 grams of ground coffee at all times.

That's not the point.

Dosers were designed to dispense a unit of coffee with each pull. Presumably they were engineered to a consistency to give the barista a fair chance of the customer returning. What is that consistency?

At a good bar in Italy, a pull of the handle and presto, an eminently drinkable shot.

I'm fed up with paddle-thwacking baristas in shops that can't build one decent shot, never mind two in row.

If a shop has a sufficient trade volume, I'd settle for 90% best shots to the miserable chance currently prevalent.

Additionally, with new grinders like the Macap Electronic Doserless becoming available, what should be our expectations

[psycho_supreme]

I'm fed up with paddle-thwacking baristas in shops that can't build one decent shot, never mind two in row.

Dosing by volume allows for "paddle-thwacking". This "paddle-thwacking" also does a very good job preventing clumping of the coffee grinds. When used with the stockfleths move it therefore betters the dosing and distribution and results in a better more consistent shot. So I'm not sure where you were going with your statement......

- Matt

[cafeIKE]

Dosing by volume allows for "paddle-thwacking". This "paddle-thwacking" also does a very good job preventing clumping of the coffee grinds. When used with the stockfleths move it therefore betters the dosing and distribution and results in a better more consistent shot. So I'm not sure where you were going with your statement......

- Matt

Clump breaking, Stockfleth's move, 30# nutating tamps, PF tapping, puck polishing, etc. are pretentions. Espresso survived a very long time without them. Their introduction has done nothing to improve consistency, especially in shops where far too much emphasis is placed on the 'show'. Perhaps paddle-thwacking begat them all?

When a doser is used as designed, a volume of coffee is dropped in the basket, reasonably well centered. A gentle pressure on the coffee spreads it out in the basket. When paddle-thwacked, the coffee is loaded up on one side and needs moving. A heavy tamp on an uneven distribution results in severe density gradients.

When a basket is overdosed coffee expands to the shower screen. Grind must be coarser to allow water to pass due to the extra screen compression on the puck. If distribution is poor and heavily tamped, the water will force its way through the less dense areas of the puck that are not even more heavily compressed by the shower screen, i.e. channel.

With a proper finely ground dose, the puck may barely touch the shower screen in a small area. As coffee is removed, any resulting pressure is relieved. Only water pressure is controlling the flow through the puck. The puck is more fluid and less likely to channel. This paper on Coffee Percolation is most interesting. It's a bit heavy on the math, but the concepts are reasonably well explained.

Which brings us back to :
Dosers are "by volume" devices.
The issue at hand is how repeatedly a given doser volume matches a dose by weight.

[RegulatorJohnson]

Which brings us back to :
Dosers are "by volume" devices.
The issue at hand is how repeatedly a given doser volume matches a dose by weight.

ill do it. ive got some older beans to mess with. ill grind a doser full of coffee, then pull the handle and dose each time and weigh it.

do you have any requests? or suggestions?

i can tell you that on the mazzer mini. if you put the timer on just enough to make it stay on, no more. mazzer owners can get a feel for it. it will consistently grind enough for a double shot, the timer will grind out accurate amounts. i can fill the hopper and turn the dial and let the grinder run out the time. i can pour anough for a double and turn the timer to the same time and it will stop just after the beans run out. its pretty consistent. close enough to be able to ignore weighing if you felt like it. i think this is the same mentality with the dosers design. its close enough to the same weight each time. its there to help reduce waste, not necessarily make a better shot. if it was there to make a better shot it wouldn't need 84 grams to work properly. or they would make a doserless grinder.

--
on the "no toucha anything..."

i have been using a minimalist approach lately. not as extreme as "no toucha" but it seems to work well. here is a shot of a recent shot. i have been really into these macro shots of the initial beads. it seems like the less i do the better these look. i let the coffee fall and then bump it around a bit then tamp. no re-distribution, trust gravity.





thanks for the time.

jon

[HB]

I'm quite surprised that none have responded with any doser accuracy tests... HB also promised.

FWIW, in a 15g basket a measured 15.0±1.0g amounts to very little more taste difference in the cup than normal shot to shot, day to day variation, if one stops the shot on color, not time.

Sorry, I've been delayed by a lingering cold/sore throat. Coffee is simply off the menu while I focus on primary concerns like breathing and sleeping. That said, I'm confident the volume measures are repeatable; level cutting a basket is certainly repeatable and a doser probably does it even better. Keep in mind, however, that volume to weight ratios can change dramatically between blends. I remember single origin Yemen packed way down in the basket at the same weight as other coffees reaching the retainer ridge.

As I promised, I will measure doser consistency as a matter of curiosity when I'm healthy again, but I don't expect it will reveal anything beyond barista trivia.