Who is here? 1 guest(s)
 Print Thread
Roasting to the target profile
Just wondering how others are roasting to the targets or in other words replicating profiles? How easy or difficult the process for others is? Are you making a plan/notes before and try to follow them during the roast? Are you loading profiles in background of software and trying to follow them? :/

I want to share with you how I was replicating my previously recorded roasting profile on my Gene Cafe (CBR-101) 🙂

I found very nice taste in a cup of my espresso from one of my previous roasts. Beans were running out, still had same greens, so decided to use same recipe and same profiling (as a Target) to try get same result in a cup.

I wanted to try and test few thing:
  1. How easy is to replicate profile following the target by the graph only. Would it be same taste from same recipe?
  2. Wanted to test microphone HY929. How easy to recognize cracks?
  3. Wanted to test temperature readings. How big difference in readings or sensitivity?

Hardware & software:
Roaster: Gene Cafe CBR-101 + extra Thermocouple K (no mods);
Software: Roasters Pal application;
Temperature logger: Bluetooth logger (Arduino) connected to app and thermocouple;
Microphone: HY929 with stick inserted into the chaff collector;

Blend from 75% Ethiopia Sidamo and 25% Colombia Excelso.

1. Target profile was recorded few weeks before (same roaster with extra thermocouple, same software, but no microphone).
2. In the app, I connected to the Bluetooth logger to receive data from thermocouple, created a new profile and chosen to show a target profile in graph.
3. During the roast I was trying to match ongoing roasting (red colour) with target (gray colour) close as possible.
4. Close to the 1st crack I was using earphones to listen for a first crack. As you can see from the graph image, with the microphone I was able to hear clear 1st crack about 40+ seconds early comparing to target. Higher temperature impacted that a little as well probably.
5. I didn’t expected so early 1st crack so presumed that same was for target roast. So I kept beans in development for the extra 30 seconds.

  1. To replicate profile with the app was easy and convenient. You don’t need to create a plan, remember points, temperatures as it is enough just to follow the graph and adjust roaster to follow line. Would be interesting to try other Gene Cafe owners recorded profiles and compare how easy/difficult to replicate profile due to differences of heating element (230/240), power (120 vs 220), dim mode, etc.
  2. The taste from beans I got is the same or are very close to my target. My tongue don’t feel any difference, so I am happy to think that replica was successful 🙂
  3. Microphone HY929 was very handy and easy to use. Definitely worth having for any Gene Cafe owners to clearly heard all the cracks without a delays. And cost just around £25-30 on Ebay.
  4. Thermocouples. I found that stock thermocouple on Gene is more sensitive but not sure if the values shown are correct. As soon as you start heating the temperature shown on Gene increases very fast. Drops quick as well once heater off. Thermocouple K readings increases/decreases more slowly and it looks more accurate comparing to Gene readings. Difference somewhere in area 40-60 C, but values match in room temperature 😕 Maybe my Gene is buggy or has defect, but I don’t think that it would be easy job to replicate profiles relaying on the Gene temperatures only. Probably this is why I was not able to do that early.


Video of the process:

CB600, CBR-101 (modifying)
Roasters Pal app (developing)
Even better is to take control of the heater and roast using power profiles.
Much more accurate airflow temperature than the on-off pulsing used by Gene.
This way you can even automate the roast process.

The FC approaching can be guessed also even by hearing impaired persons, from a sudden rise of RoR preceding by 15-20 seconds the very first cracks.
Possible only if your measurement system computes and displays rate of change.
Not with Gene own display.


renatoa wrote:

The FC approaching can be guessed also even by hearing impaired persons, from a sudden rise of RoR preceding by 15-20 seconds the very first cracks.

It's interesting. I'll take a look, just out of curiosity. But that would probably still be too complicated for me smile
CB600, CBR-101 (modifying)
Roasters Pal app (developing)
Not so complicate, actually, if using a measurement system showing rate of change.
I mean a real time rate of change, computed by measurement system, not a user computed manually based on observed temperatures and time, this is not fast and precise enough.
Check this post for an example: https://homeroast...post_77057

Personally I am careful more with rate of change evolution than temperatures figures, which completely different from machine to machine, and even for same machine depends on the probe placement.
But the rate of change tell me the same for any machine.
Jump to Forum: