Roasting quality control, Tim Wendelboe fanboy edition
- baldheadracing
- Team HB
- Posts: 6272
- Joined: 9 years ago
In Value Of A Bottomless Portafilter In Developing Skill?
Statistical Quality Control is a simple recipe:
1. Measure objective quality indicators from every batch (the "statistical" part happens when some batches are measured, but not applicable)
2. Check how much those indicators deviate from the norm.
3. Remove batches that deviate too much (TW calls these "fails")
4. Investigate every deviation to try to eliminate future deviations.
However, as cupping typically happens after shipment ...
1a. Collect a sample from every batch
2a. Subjectively compare samples (cupping)
3a. (doesn't happen)
4. Investigate every deviation to try to eliminate future deviations.
As cupping is subjective:
5. Regularly cup your roasts against other roasters that roast similarily, and with similar coffees (blinded)
Finally, if the batches are close enough that they don't fail, then tighten up the criteria for "too much" deviation.
Examples of objective measurements (not complete):
- ambient climate
- charge weight and density
- roasted weight and density
- charge temperature and drop temperature
- roast colour
- time in stages/development time/development ratio.
Oldest to newest TW roasting video - if you want the historical context, start from oldest to newest; but if you want the gist, start from newest to oldest. If you want just one, watch the Loring video:
Coffee Roasting QnA (49 minutes)
http://www.youtube.com/watch?v=HoIoLxKIJ_8
Cupping for Roasting QC (35 minutes)
http://www.youtube.com/watch?v=3SM8QNjWgO4
"Ask me anything" about roasting (hour long)
http://www.youtube.com/watch?v=aCur6NqcyPk
Loring roaster workshop (1:06, but there's a bunch about Loring)
http://www.youtube.com/watch?v=Akqx4fJEJfg
Stronghold demo (an hour, not yet on YouTube)
LoL In a nutshell, be consistent, and measure so you know that you are consistent.LBIespresso wrote:Hmmm...Do I sense a new thread in the Roasting section of Home-Barista? "Tim Wendleboe Approach to Quality" by Baldheadracing ?
Pretty please?
Statistical Quality Control is a simple recipe:
1. Measure objective quality indicators from every batch (the "statistical" part happens when some batches are measured, but not applicable)
2. Check how much those indicators deviate from the norm.
3. Remove batches that deviate too much (TW calls these "fails")
4. Investigate every deviation to try to eliminate future deviations.
However, as cupping typically happens after shipment ...
1a. Collect a sample from every batch
2a. Subjectively compare samples (cupping)
3a. (doesn't happen)
4. Investigate every deviation to try to eliminate future deviations.
As cupping is subjective:
5. Regularly cup your roasts against other roasters that roast similarily, and with similar coffees (blinded)
Finally, if the batches are close enough that they don't fail, then tighten up the criteria for "too much" deviation.
Examples of objective measurements (not complete):
- ambient climate
- charge weight and density
- roasted weight and density
- charge temperature and drop temperature
- roast colour
- time in stages/development time/development ratio.
Oldest to newest TW roasting video - if you want the historical context, start from oldest to newest; but if you want the gist, start from newest to oldest. If you want just one, watch the Loring video:
Coffee Roasting QnA (49 minutes)
http://www.youtube.com/watch?v=HoIoLxKIJ_8
Cupping for Roasting QC (35 minutes)
http://www.youtube.com/watch?v=3SM8QNjWgO4
"Ask me anything" about roasting (hour long)
http://www.youtube.com/watch?v=aCur6NqcyPk
Loring roaster workshop (1:06, but there's a bunch about Loring)
http://www.youtube.com/watch?v=Akqx4fJEJfg
Stronghold demo (an hour, not yet on YouTube)
-"Good quality brings happiness as you use it" - Nobuho Miya, Kamasada
- LBIespresso
- Supporter ❤
- Posts: 1245
- Joined: 7 years ago
Niiiiice! A post with homework. I love it!
Thanks Craig.
Thanks Craig.
LMWDP #580