Cafelat Robot Temperature Tests - Page 25

#241: Post by K7 » May 18th, 2021, 3:54 pm

jpender wrote: The probes could be misreporting for other reasons. But I'm not certain that it is impossible that the exit stream -- coming from brew water that was hotter -- might actually exceed the temperature of the current, colder brew water.

I don't think that's impossible either. I was just offering a possible explanation in case the upper probe is much faster than the bottom probe. Since you say they are the same, my comment is a moot point.

Btw, what was the dose and output weight? Thanks for running the test and sharing!

May 18th, 2021, 6:56 pm

18g dose. A medium dark coffee that tastes best pulled short of 1:2. First shot was 30g out in 35s total (10s preinfusion). The second shot was from a new bag of the same kind of beans and it pulled much more slowly. I guess I should have loosened the grind. I stopped it at 25g after a total of 50s (15s preinfusion).

This was pretty easy to do this morning. I'll have to try it some more and see what I get.

May 19th, 2021, 11:55 am

Four shots, all without any preheating:

The brew water traces are pretty much the same. But the stream temperature is a lot less reliable. I wonder if it's just a question of how exactly the stream forms over the sensor? Whatever is going on it means that for my setup exit temperature isn't a particularly valuable measure.

#244: Post by **drgary** » May 26th, 2021, 3:40 pm

This post is to alert all of you experimenting with preheating to a new thread where you can discuss how the taste of the coffees you're trying are affected by these techniques, or in some cases not preheating at all.

Does preheating the Cafelat Robot affect taste? Please post your experiences here!

This thread is focused on the measurement and engineering of pre-heating with the Robot. That doesn't address application of those techniques to improving shot quality. This quote from John from earlier in this thread illustrates his experiences of variations in taste with pre-heating:

jpender wrote:...sometimes preheating the piston seems to really improve the coffee, depending on the coffee. Other times it makes it worse. And there are also times when it's neither better nor worse, just different. That's how I experienced preheating with the particular coffee I pulled in this recent test. The roaster suggests a pretty high temperature, 95-96°C, which would be quite hard to maintain for an entire shot with a Robot. I like the coffee without any preheating but it tastes different when I preheat. So, I don't know what to say except I think it's worth having the tool in the box, so to speak.

#245: Post by **Paul_Pratt** » May 26th, 2021, 10:42 pm

I have more complete data but a teaser is as below. I posted this on a FB group weeks ago so.

The 1st graph is measured using a probe from the piston. The probe was approx 10mm from the face of the piston. My kettle stops at 96.5 deg c and from there I walk over and pour onto the coffee. The first "from cold" is exactly that, 1st shot of the day using coffee, the ambient temp at the time was just under 20 deg c

The 2nd shot was done after the 1st in real time. Knock out the puck, dose, grind, weigh etc...

For the 3rd shot I quickly removed the basket and put it into my extra naked portafilter and the basket was wrapped in silicone sheets in an attempt to insulate it. It did not make much noticeable difference.

For this 2nd pretty picture again from dead "cold" room temp just under 20 deg c. I decided to make a blank puck out of an old tamper seat. This sits at the bottom of the pro basket. I addedd boiling water to the basket, inserted into the machine, let the arms fall down and I left it. You could apply pressure if you wanted to. This had the desired effect of "pre-heating" the piston and basket. I did this for about a minute twice. I then removed the plug and made coffee. The effect was pretty good.

So that is what I am working on now. Making a silicone pre heat plug. It works and produces a consistent repeatable bump up in temp if you so desired.

The large data set I have involved a Scace style thermofilter. I then had 2 data points, the probe from the piston and then another in the simulated coffee puck. I still need to go through it all. But what was very clear was that after a few shots the temp was very repeatable.

#246: Post by K7 » May 27th, 2021, 2:24 pm

This 58mm backflush silicone disk (Amazon $10) sealed the basket bottom pretty well when I tried it out for a few days recently.

May 28th, 2021, 4:10 pm

cap2 wrote: Not so nice, really. I am just eager for results. :D

Patience... I'll get there, soon.

#248: Post by **cap2** » May 28th, 2021, 7:44 pm

jpender wrote:Patience... I'll get there, soon.

looks like you ARE there!

June 5th, 2021, 1:50 pm

cap2 wrote:looks like you ARE there!

I wish that had been true. I mismeasured and drilled the hole 3mm high. It worked but was harder to lock in. So I decided to drill another hole and that's when I really came to appreciate that the drill press I had was a piece of junk. It took a lot of effort, and a total of three holes, to get a decent one. After filling the other two, inserting a new nut, drilling out plastic inside to prevent the Robot struts from binding, a bunch more hand sanding, soldering and epoxying in a sensor, gluing on a piece of silicone to the face to minimize heating of the PVC, and finally coating the exposed PVC with FDA silicone adhesive, I'm finally there, a week later.

I have pulled one shot, nothing preheated, and the temperature decline over 35 seconds was the same that I have gotten with a fully preheated S/S piston. It's just one shot and could be an outlier but it's encouraging.

EDIT: I pulled a couple of more shots, one without any preheat and one with a preheat of the PF and basket. It looks like the PVC piston does a pretty good job of imitating a fully preheated S/S piston.

#250: Post by **cap2** » June 5th, 2021, 8:51 pm

I have pulled one shot, nothing preheated, and the temperature decline over 35 seconds was the same that I have gotten with a fully preheated S/S piston. It's just one shot and could be an outlier but it's encouraging.

Nice and clear preliminary results! Sorry it was so much work.