For a while now, I have struggled to pull very good shots with light roasts with the Robot. Most of the shots I've pulled were overall quite decent with no biting acidity or astringent notes. They also tasted of their origin. So far, so good.

I've experimented with various ratios anywhere from 1:2 to 1:5. I've also pulled various types of shots, straight 6 bar, straight 8, declining pressure, blooming shots, Rao allongés, turbo shots. I've done various preheating techniques, double pour, pour and overflow, pre-heating the piston. The latter was the most impactful but also the most demanding with regards to workflow, even then, the gain in taste was measured.

What was lacking overall was a very strong sense of origin.

I've noticed over time a slightly concave shape for the top of the puck, as well as a dark ring around the bottom where there are no holes (Puckology can be dubious on a Robot, but if you push all the water out the puck survives intact).

The dark ring is typical of an under-extraction ring where there are still a lot of dark compounds remaining inside the grounds rather than expelled into the cup. And if we think of extraction in terms of compounds being washed away, then higher, thicker sides mean that fewer compounds are washed away in those areas. We have a large case of side under-extraction.

(btw I was pointed towards Jonathan Gagné / Coffee ad Astra who has an excellent article on the subject) .

I believe this is one of the keys to the taste of the Robot. The center of the puck is essentially extracted at normale ratio, while the sides of the puck are extracted at ristretto ratio. Especially if you pull darker roasts or/and shorter, this adds complexity to the shot and when you dial it well it makes a normale shot taste a bit like a ristretto.

As far as I can tell, this is a function of the shape of the Robot's basket, the slanted walls and the holes not reaching the edges radially skew the extraction levels. For reference, a VST basket has holes that cover a 49 mm diameter circle, a Pullman basket has holes that cover a 51 mm diameter circle, and a Robot basket has holes that cover a 40 mm diameter circle.

For a light roast, I seek even extraction throughout the puck. And here's the good news, there's a solution, it's been used for other baskets, but in this case where there are slanted walls and where the holes don't cover a lot of the bottom area, the improvement is extremely pronounced.

You can add a wet paper filter to the bottom of the basket. Pour your grinds on top. WDT. Tamp. Add top filter and pull your shot. The bottom paper filter will ensure that there is radial flow from the very edges of the basket to the first holes it meets. Edge under-extraction will be greatly reduced.

Okay, that's the theory, but what are the results?

1. Taste-wise, it's obvious. There's an added intensity to each shot that's almost hitting me in the face. There's simply more coffee, more origin characteristics, it's thicker at the same ratio. It's absolutely delicious.

2. Measurement-wise. I've measured higher TDS with a digital refractometer translating to extraction yields anywhere from 2% to 4-5% higher depending on the type of shots I pulled. With some beans, 21% EY 1:2 shots or 24% 1:3 shots that taste excellent are achievable with a small grinder.

3. Without the bottom paper filter, I have uniform beading across the whole surface of the basket. That sounded great until I realized that there's a large edge ring beneath which there are no holes, where is that liquid going? With a bottom paper filter, that liquid gets to the edge holes more easily and I can easily see it, there's some doughnuting, in this case, this is a very good thing.

4. Puck-wise, there is now a total absence of dark spots anywhere on the puck, it's all uniform in colour. The top of the puck is now nearly perfectly flat, nearly no concavity, there's still more to be gained, but all obvious signs of less edge extraction are gone.

For light roasts and nordic roasts, this is the biggest improvement I've had by far. To the extent where I don't feel the need to preheat the piston.

I would love to see everyone with a Robot who likes light roast espresso try it, it's trivial. Wash an aeropress paper filter or a Robot paper filter under cold water, smush it against the bottom of the basket, wdt and pull your shot as normal. There's more texture, more taste, more of everything and it's great. If you don't like it, that's okay, the taste profile might not be for everyone, but it could also be a huge improvement.

Still, this is an inconvenient step, and not everybody wants paper-filtered espresso. For the future, I can only hope for a basket with straight walls and holes covering a circle that goes to the edges, that would make the whole paper thing entirely unnecessary but for a most minimal increase of extraction.

Meanwhile, it's something interesting to try. Cheers.

I'll definitely try it, using light roasts the past month for the first time. I don't have the knowledge that you have, but i am curious to give it a try.

I alway push all the water out of the basket and I don't think I have ever had a puck that was in good shape. Still comes up slightly crooked and they have some cracks. I think the upward pressure of air does not go through the solid pick too well. I will try to lift a little slower.

Always wondered why the bottom was tapered. I saw a video a while ago that stated the single dose baskets with tapers actually create more pressure coming out of the basket. Wonder if this was done to reduce the pressure needed on the handles?

Hm. Do you preinfuse? I would think paper slows down exit of coffee, acting as a kind of back pressure preinfusion. You're also presumably cutting coffee oils from the output.

Ken5 wrote:I'll definitely try it, using light roasts the past month for the first time. I don't have the knowledge that you have, but i am curious to give it a try.

I alway push all the water out of the basket and I don't think I have ever had a puck that was in good shape. Still comes up slightly crooked and they have some cracks. I think the upward pressure of air does not go through the solid pick too well. I will try to lift a little slower.

Always wondered why the bottom was tapered. I saw a video a while ago that stated the single dose baskets with tapers actually create more pressure coming out of the basket. Wonder if this was done to reduce the pressure needed on the handles?

I don't know why that shape was chosen but I've only read praise from people who like darker roasts, I've speculated that it is a good shape for this.

To get an intact puck, I push the piston all the way down until all the water inside is expelled.

happycat wrote:Hm. Do you preinfuse? I would think paper slows down exit of coffee, acting as a kind of back pressure preinfusion. You're also presumably cutting coffee oils from the output.

I do five to ten seconds of pre-infusion at low pressure before full pressure. Longer and the temperature drops more than I'm comfortable with.

I have not noticed any slow down or additional pressure from the added paper filter, if anything it helps coffee flow out of the basket.

I'm glade to see this post in here. I was got the same result before, and post in cafelat robot owners group of facebook.

do you find that the shot still has body? i've tried an aeropress filter below the puck on my Decent and got a pretty intense shot, but lacking texture. must have something to do with fines no longer getting out.

If you owned suitable grinder you can try to grind into basket and be a mound then tamp straightly with no paper filter at bottom. In this way, you still can get over 20%EY. If you have no suitable grinder(like hand grinder), you can do WDT in catch cup first and through drip(or something else) fall into basket of robot to be a mound then tamp with no paper filter, and you still can get good result of EY and no channeling. I think you may try this way and compare difference in cups.

Thank you zefkir! Did anyone try using yet the puck screen from Bplus? Not tried but I heard about some good results with a big lever machine.

I've been doing this for a while, but with the puck sandwiched between two filters - the bottom one cut using a 2" circle punch.

Nice results with a blooming profile. I've also measured significantly higher extraction. It is a hassle though, the filters on top can be reused but the bottom ones need to be replaced every time in my experience. For some roasts it feels more like a crutch because using more suitable burrs like the 64mm SSP unimodal can yield better results even without the bottom filter.

SorinL wrote:Thank you zefkir! Did anyone try using yet the puck screen from Bplus? Not tried but I heard about some good results with a big lever machine.

Cafelat Robot User Experience