Spinning for distribution? - Page 7

February 7th, 2008, 1:46 am

The_Mighty_Bean wrote:But, if I may flog the dead horse once again, if you get that puck to max tapped density, then that variance is drastically reduced, because the tapped density is supposedly constant for a given substance.

I understand your motivation and what you're conveying as a better solution (I think), but if you're truly talking about tap density (as in transverse axial pressure measurements), I don't believe that the micron resolution of measurement is necessary for the context of preparing espresso, IN MY HUMBLE OPINION.

It's not like we're packaging a paralytic like succinylcholine for intravenous use which requires a resolution less than 0.1 mg.

Or, am I missing something? (and I don't mean that sarcastically)

Matt

#62: Post by **Jepy** » February 7th, 2008, 1:47 am

I built a couple of devices that do this. I wasn't aiming for a better shot(no pun intended), but a more consistent shot, mainly for when I was running tests on various espresso machines I've made.
The design I have now is dead on accurate time after time, with pretty much zero waste.
Does it make the coffee better? I guess it depends on who's taste we're talking about. What I can say is that it makes dialing in a blend easier since you take the inconsistency of manually dosing out of the equation. Maybe some of you real pros out there make a very consistent dose. I can't manually.

It's not a simple device, with a fair amount of machining, and welding involved, and I'm sure most would consider it a Rube Goldberg machine, but it's made me very spoiled. Also, I've tried this on Mazzer Mini, Super Jolly, both size Robur burr sets, Cim Junior, and DRM combo. It makes all more consistent, but it surely won't make a Mini a DRM. As far as I can tell there is no substitute for good burrs.

This pic was taken almost two years ago, it looks similar to a Versalab dosed PF, but it's not . I now have it tuned to give a more level center.

February 7th, 2008, 1:53 am

Matthew Brinski wrote:I understand your motivation and what you're conveying as a better solution (I think), but if you're truly talking about tap density (as in transverse axial pressure measurements), I don't believe that the micron resolution of measurement is necessary for the context of preparing espresso, IN MY HUMBLE OPINION

Matt

LOL I understand. And that's the million-dollar question isn't it? At what point does the error tolerance matter, and should we just go down to max density for consistency's sake? See the post just below yours.

February 7th, 2008, 1:58 am

Jepy wrote:I built a couple of devices that do this. I wasn't aiming for a better shot(no pun intended), but a more consistent shot, mainly for when I was running tests on various espresso machines I've made.
The design I have now is dead on accurate time after time, with pretty much zero waste.
Does it make the coffee better? I guess it depends on who's taste we're talking about. What I can say is that it makes dialing in a blend easier since you take the inconsistency of manually dosing out of the equation. Maybe some of you real pros out there make a very consistent dose. I can't manually.

It's not a simple device, with a fair amount of machining, and welding involved, and I'm sure most would consider it a Rube Goldberg machine, but it's made me very spoiled. Also, I've tried this on Mazzer Mini, Super Jolley, both size Robur burr sets, Cim Junior, and DRM combo. It makes all more consistent, but it surely won't make a Mini a DRM. As far as I can tell there is no substitute for good burrs.

This pic was taken almost two years ago, it looks similar to a Versalab dosed PF, but it's not . I now have it tuned to give a more level center.

<image>

It won't make the Mini a DRM because the particle size distribution won't be as even. --Edit-- Link fixed, sorry!

But your distribution errors are history. It would be fascinating to figure out the effect on flavor. It might next be really fascinating (and useful and awesome) to make this something I can fit on my countertop for 50 bucks or less. Oh, and it's gotta fit into Jon's 25 second routine.

But first things first... can you tell us a little more about what your Rube Goldberg actually DOES to the coffee?

~tMb, slightly vindicated.

February 7th, 2008, 9:47 am

The_Mighty_Bean wrote:Once you have that consistent puck, it's easy to test and adjust for flavor issues.

ok what about variables in the beans in the blend. imagine that its a bowl of colored M&M's. you scoop out 5 samples of 20g of M7M's.. do they have a consistent mix of color? probably not. now these are probably equal weight and sized M&M's. beans are not equal sized or weight. even a Single Origin will have variations in the size of the bean itself. this variation will give a slightly different roast profile on the individual bean itself. you could never get a consistent result in the mix of the beans itself. this variation in the actual beans will probably affect the flavor profile more than the distribution.

i brought up this subject in a thread called "using a blend to test.. a flawed approach?".

jon

February 7th, 2008, 9:51 am

hey matt you misquoted me.. i didnt flog a horse living or dead... or did i? either way i got credit for something i didnt say.. type.

jon

February 7th, 2008, 10:09 am

Sorry... don't know why it translated it like that, but I fixed it.

I know you would never hurt a horse.

Matt

February 7th, 2008, 10:17 am

RegulatorJohnson wrote:ok what about variables in the beans in the blend. imagine that its a bowl of colored M&M's. you scoop out 5 samples of 20g of M7M's.. do they have a consistent mix of color? probably not. now these are probably equal weight and sized M&M's. beans are not equal sized or weight. even a Single Origin will have variations in the size of the bean itself. this variation will give a slightly different roast profile on the individual bean itself. you could never get a consistent result in the mix of the beans itself. this variation in the actual beans will probably affect the flavor profile more than the distribution.

i brought up this subject in a thread called "using a blend to test.. a flawed approach?".

jon

Which is why I have a reservation about blends that have more than two or three components (and there are many respected blends blends out there with four, five, and six components).

The good roasters of the multiple origin/lot blends must know something I don't though, because their blends are proactively developed rather than just haphazardly thrown together.

Matt

#69: Post by **dsc** » March 8th, 2009, 3:02 pm

Hi guys,

I was thinking that centrifugal force might help.

maybe something like this would work:

It's a small DC motor placed inside the funnel and a wire stuck to it, so when it spins the wire throws coffee and forms a small tornado inside the funnel (sort of like a Versalab).

The distribution is still not perfect as I'm getting a blank spot on the left side of the basket, but I think it gives thicker shots, the coffee is more gooey and I'm getting a bit less side channeling.

Still it's not as effective as jepy's method (not sure what he's using), although I have to say it only took an hour to put together, so I can't be too demanding.

Regards,
dsc.